DRAFT-DO NOT CITE OR QUOTE EPA/635/R-10/002A
www.epa.gov/iris
vvEPA
TOXICOLOGICAL REVIEW OF
FORMALDEHYDE -
INHALATION ASSESSMENT
(CAS No. 50-00-0)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
VOLUME I of IV
Introduction, Background,
and Toxicokinetics
June 2, 2010
NOTICE
This document is an External 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.
I-ii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS IN BRIEF
VOLUME I: INTRODUCTION, BACKGROUND, AND TOXICOKINETICS
Chapter 1: Introduction
Chapter 2: B ackground
Chapter 3: Toxicokinetics
VOLUME II: HAZARD CHARACTERIZATION
Chapter 4: Hazard Characterization
VOLUME III: QUANTITATIVE ASSESSMENT, MAJOR CONCLUSIONS IN THE
CHARACTERIZATION OF HAZARD AND DOSE RESPONSE, AND
REFERENCES
Chapter 5: Quantitative Assessment: Inhalation Exposure
Chapter 6: Major Conclusions in the Characterization of Hazard and Dose-Response
References
VOLUME IV: APPENDICES
Appendix A: Summary of External Peer Review and Public Comments and Dispositions
Appendix B: Simulations of Interindividual and Adult-to-Child Variability in Reactive
Gas Uptake in a Small Sample of People (Garcia et al., 2009)
Appendix C: Lifetable Analysis
Appendix D: Model Structure & Calibration in Conolly et al. (2003, 2004)
Appendix E: Evaluation of BBDR Modeling of Nasal Cancer in the F344 Rat: Conolly
et al. (2003) and Alternative Implementations
Appendix F: Sensitivity Analysis of BBDR Model for Formaldehyde Induced
Respiratory Cancer in Humans
Appendix G: Evaluation of the Cancer Dose-Response Modeling of Genomic Data for
Formaldehyde Risk Assessment
Appendix H: Expert Panel Consultation on Quantitative Evaluation of Animal
Toxicology Data for Analyzing Cancer Risk Due to Inhaled Formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
I-iii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS—TOXICOLOGICAL REVIEW OF FORMALDEHYDE
(CAS No. 50-00-0)
VOLUME I
LIST OF TABLES I-xiv
LIST OF FIGURES I-xxiv
LIST OF ABBREVIATIONS AND ACRONYMS I-xxxi
FOREWORD I-xxxvi
AUTHORS, CONTRIBUTORS, AND REVIEWERS I-xxxvii
1. INTRODUCTION 1-1
2. BACKGROUND 2-1
2.1. PHYSICOCHEMICAL PROPERTIES OF FORMALDEHYDE 2-1
2.2. PRODUCTION, USES, AND SOURCES OF FORMALDEHYDE 2-1
2.3. ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE 2-4
2.3.1. Inhalation 2-5
2.3.2. Ingestion 2-11
2.3.3. Dermal Contact 2-12
3. TOXICOKINETICS 3-1
3.1. CHEMICAL PROPERTIES AND REACTIVITY 3-1
3.1.1. Hydration of Formaldehyde 3-1
3.1.2. Binding of Formaldehyde to Proteins 3-2
3.2. ABSORPTION 3-4
3.2.1. Oral 3-4
3.2.2. Dermal 3-4
3.2.3. Inhalation 3-4
3.2.3.1. Formaldehyde Uptake Can Be Affected by Effects at the Portal
of Entry 3-5
3.3. DISTRIBUTION 3-7
3.3.1. Transport of Methylene Glycol 3-7
3.3.2. Formaldehyde-GSH Conjugate as a Method of Systemic Distribution 3-8
3.3.3. Levels in Blood 3-8
3.3.4. Levels in Various Tissues 3-10
3.3.4.1. Disposition of Formaldehyde: Differentiating Covalent Binding
and Metabolic Incorporation 3-12
3.4. METABOLISM 3-15
3.4.1. In Vitro and In Vivo Characterization of Formaldehyde Metabolism 3-15
3.4.2. Formaldehyde Exposure and Perturbation of Metabolic Pathways 3-19
3.4.3. Evidence for Susceptibility in Formaldehyde Metabolism 3-20
3.5. ENDOGENOUS SOURCES OF FORMALDEHYDE 3-21
3.5.1.1. Normal Cellular Metabolism (Enzymatic) 3-21
3.5.1.2. Normal Metabolism (Nonenzymatic) 3-23
This document is a draft for review purposes only and does not constitute Agency policy.
I-iv DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
3.5.1.3. Exogenous Sources of Formaldehyde Production 3-23
3.5.1.4. Metabolic Products of Formaldehyde Metabolism 3-24
3.5.1.5. Levels of Endogenous Formaldehyde in Animal and Human
Tissues 3-24
3.6. EXCRETION 3-26
3.6.1. Formaldehyde Excretion in Rodents 3-27
3.6.2. Formaldehyde Excretion in Exhaled Human Breath 3-28
3.6.3. Formaldehyde Excretion in Human Urine 3-34
3.7. MODELING THE TOXICOKINETICS OF FORMALDEHYDE AND DPX 3-35
3.7.1. Motivation 3-35
3.7.2. Species Differences in Anatomy: Consequences for Gas Transport and Risk.. 3-35
3.7.3. Modeling Formaldehyde Uptake in Nasal Passages 3-41
3.7.3.1. Flux Bins 3-42
3.7.3.2. Flux Estimates 3-42
3.7.3.3. Mass Balance Errors 3-43
3.7.4. Modeling Formaldehyde Uptake in the Lower Respiratory Tract 3-43
3.7.5. Uncertainties in Formaldehyde Dosimetry Modeling 3-45
3.7.5.1. Verification of Predicted Flow Profiles 3-45
3.7.5.2. Level of Confidence in Formaldehyde Uptake Simulations 3-46
3.7.6. PBPK Modeling of DNA Protein Cross-Links (DPXs) Formed by
Formaldehyde 3-49
3.7.6.1. PBPK Models for DPXs 3-49
3.7.6.2. A PBPK Model for DPXs in the F344 Rat and Rhesus Monkey
that Uses Local Tissue Dose of Formaldehyde 3-51
3.7.6.3. Uncertainties in Modeling the Rat and Rhesus DPX Data 3-52
3.7.7. Uncertainty in Prediction of Human DPX Concentrations 3-54
3.7.8. Modeling Interindividual Variability in the Nasal Dosimetry of Reactive
and Soluble Gases 3-56
VOLUME II
4. HAZARD CHARACTERIZATION 4-1
4.1. HUMAN STUDIES 4-1
4.1.1. Noncancer Health Effects 4-1
4.1.1.1. Sensory Irritation (Eye, Nose, Throat Irritation) 4-1
4.1.1.2. Pulmonary Function 4-11
4.1.1.3. Asthma 4-50
4.1.1.4. Respiratory Tract Pathology 4-65
4.1.1.5. Immunologic Effects 4-69
4.1.1.6. Neurological/Behavioral 4-82
4.1.1.7. Developmental and Reproductive Toxicity 4-85
This document is a draft for review purposes only and does not constitute Agency policy.
I-v DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.1.1.8. Oral Exposure Effects on the Gastrointestinal Tract 4-96
4.1.1.9. Summary: Noncarcinogenic Hazard in Humans 4-96
4.1.2. Cancer Health Effects 4-97
4.1.2.1. Respiratory Tract Cancer 4-98
4.1.2.2. Nonrespiratory Tract Cancer 4-134
4.1.2.3. Summary: Carcinogenic Hazard in Humans 4-188
4.2. ANIMAL STUDIES 4-191
4.2.1. Reflex Bradypnea 4-192
4.2.1.1. Tolerance 4-195
4.2.1.2. Cross-Species Differences in Inhaled Dose 4-195
4.2.1.3. Cross-Tolerance 4-198
4.2.1.4. Formaldehyde Binding and Activation of Trigeminal Nerve
Afferent Activity 4-199
4.2.2. Respiratory Tract Pathology 4-201
4.2.2.1. Mucociliary Clearance 4-202
4.2.2.2. Cell Proliferation 4-209
4.2.2.3. Short-Term Studies 4-226
4.2.2.4. Subchronic Studies 4-240
4.2.2.5. Chronic Inhalation Bioassays 4-260
4.2.2.6. Summary of Respiratory Pathology and Carcinogenic
Potential 4-286
4.2.3. Gastrointestinal Tract Pathology 4-291
4.2.3.1. Short-Term and Subchronic Ingestion Studies 4-291
4.2.3.2. Chronic Ingestion Studies 4-294
4.2.3.3. Summary of Gastrointestinal Effects and Evaluation of
Carcinogenic Potential 4-300
4.2.4. Immune Function 4-300
4.2.5. Hypersensitivity and Atopic Reactions 4-309
4.2.5.1. Inhalation Studies in Experimental Animals 4-311
4.2.5.2. Dermal Sensitization 4-331
4.2.5.3. Summary of Sensitization Studies 4-333
4.2.6. Neurological and Neurobehavioral Function 4-338
4.2.6.1. Inhalation Exposure 4-338
4.2.6.2. Oral Exposure 4-367
4.2.6.3. Summary 4-369
4.2.6.4. Other Considerations 4-369
4.2.7. Reproductive and Developmental Toxicity 4-370
4.2.7.1. Inhalation Studies Addressing Developmental and
Reproductive Toxicity 4-370
4.2.7.2. Oral Exposure Studies Addressing Developmental and
Reproductive Toxicity 4-387
This document is a draft for review purposes only and does not constitute Agency policy.
I-vi DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.2.7.3. Intraperitoneal Studies Addressing Developmental and
Reproductive Toxicity 4-390
4.2.7.4. Dermal Exposure Studies Addressing Developmental and
Reproductive Toxicity 4-392
4.2.7.5. Summary of Reproductive and Developmental Effects 4-393
4.3. GENOTOXICITY 4-406
4.3.1. Formaldehyde-DNAReactions 4-406
4.3.1.1. DNA-Protein Cross-Links (DPXs) 4-409
4.3.1.2. DNAAdducts 4-415
4.3.1.3. DNA-DNA Cross-Links (DDXs) 4-418
4.3.1.4. Single Strand Breaks 4-418
4.3.1.5. Other Genetic Effects of Formaldehyde in Mammalian Cells 4-420
4.3.2. In Vitro Clastogenicity 4-421
4.3.3. In Vitro Mutagenicity 4-424
4.3.3.1. Mutagenicity in Bacterial Systems 4-424
4.3.3.2. Mutagenicity in Other Nonmammalian Cell Systems 4-429
4.3.3.3. Mutagenicity in Mammalian Cell Systems 4-429
4.3.4. In Vivo Mammalian Genotoxicity 4-435
4.3.4.1. Genotoxicity in Laboratory Animals 4-435
4.3.4.2. Genotoxicity in Humans 4-438
4.3.5. Summary of Genotoxicity 4-448
4.4. SYNTHESIS AND MAJOR EVALUATION OF NONCARCINOGENIC
EFFECTS 4-449
4.4.1. Sensory Irritation 4-449
4.4.2. Pulmonary Function 4-456
4.4.3. Hypersensitivity and Atopic Reactions 4-460
4.4.4. Upper Respiratory Tract Histopathology 4-465
4.4.5. Toxicogenomic and Molecular Data that may Inform MOAs 4-467
4.4.6. Noncancer Modes of Actions 4-468
4.4.7. Immunotoxicity 4-471
4.4.8. Effects on the Nervous System 4-472
4.4.8.1. Irritant Threshold Detection 4-472
4.4.8.2. Behavioral Effects 4-472
4.4.8.3. Neurochemistry, Neuropathology, and Mechanistic Studies 4-473
4.4.8.4. Summary 4-474
4.4.8.5. Data Gaps 4-474
4.4.9. Reproductive and Developmental Toxicity 4-475
4.4.9.1. Spontaneous Abortion and Fetal Death 4-475
4.4.9.2. Congenital Malformations 4-478
4.4.9.3. Low Birth Weight and Growth Retardation 4-478
4.4.9.4. Functional Developmental Outcomes 4-479
4.4.9.5. Male Reproductive Toxicity 4-479
This document is a draft for review purposes only and does not constitute Agency policy.
I-vii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.4.9.6. Female Reproductive Toxicity 4-481
4.4.9.7. Mode of Action 4-482
4.4.9.8. Data Gaps 4-484
4.5. SYNTHESIS AND EVALUATION OF CARCINOGENICITY 4-484
4.5.1. Cancers of the Respiratory Tract 4-484
4.5.2. Lymphohematopoietic Malignancies 4-489
4.5.2.1. Background 4-489
4.5.2.2. All LHP Malignancies 4-490
4.5.2.3. All Leukemia 4-494
4.5.2.4. Subtype Analysis 4-496
4.5.2.5. Myeloid Leukemia 4-497
4.5.2.6. Solid Tumors of Lymphoid Origin 4-500
4.5.2.7. Supporting Evidence from Animal Bio-Assays for
Formaldehyde-Induced Lymphohematopoietic Malignancies 4-504
4.5.3. Carcinogenic Mode(s) of Action 4-508
4.5.3.1. Mechanistic Data for Formaldehyde 4-510
4.5.3.2. Mode of Action Evaluation for Upper Respiratory Tract
Cancer (Nasopharyngeal Cancer, Sino-nasal) 4-521
4.5.3.3. Mode(s) of Action for Lymphohematopoietic Malignancies 4-527
4.5.4. Hazard Characterization for Formaldehyde Carcinogenicity 4-533
4.6. SUSCEPTIBLE POPULATIONS 4-535
4.6.1. Life Stages 4-535
4.6.1.1. Early Life Stages 4-536
4.6.1.2. Later Life Stages 4-540
4.6.1.3. Conclusions on Life-Stage Susceptibility 4-540
4.6.2. Health/Disease Status 4-541
4.6.3. Nutritional Status 4-542
4.6.4. Gender Differences 4-543
4.6.5. Genetic Differences 4-543
4.6.6. Coexposures 4-545
4.6.6.1. Cumulative Risk 4-545
4.6.6.2. Aggregate Exposure 4-546
4.6.7. Uncertainties of Database 4-546
4.6.7.1. Uncertainties of Exposure 4-546
4.6.7.2. Uncertainties of Effect 4-547
4.6.8. Summary of Potential Susceptibility 4-548
This document is a draft for review purposes only and does not constitute Agency policy.
I-viii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
VOLUME III
5. QUANTITATIVE ASSESSMENT: INHALATION EXPOSURE 5-1
5.1. INHALATION REFERENCE CONCENTRATION (RfC) 5-2
5.1.1. Candidate Critical Effects by Health Effect Category 5-3
5.1.1.1. Sensory Irritation of the Eyes, Nose, and Throat 5-3
5.1.1.2. Upper Respiratory Tract Pathology 5-5
5.1.1.3. Pulmonary Function Effects 5-6
5.1.1.4. Asthma and Allergic Sensitization (Atopy) 5-10
5.1.1.5. Immune Function 5-17
5.1.1.6. Neurological and Behavioral Toxicity 5-17
5.1.1.7. Developmental and Reproductive Toxicity 5-26
5.1.2. Summary of Critical Effects and Candidate RfCs 5-34
5.1.2.1. Selection of Studies for Candidate RfC Derivation 5-34
5.1.2.2. Derivation of Candidate RfCs from Key Studies 5-40
5.1.2.3. Evaluation of the Study-Specific Candidate RfCs 5-66
5.1.3. Database Uncertainties in the RfC Derivation 5-70
5.1.4. Uncertainties in the RfC Derivation 5-72
5.1.4.1. Point of Departure 5-73
5.1.4.2. Extrapolation from Laboratory Animal Data to Humans 5-74
5.1.4.3. Human Variation 5-74
5.1.4.4. Subchronic-to-Chronic Extrapolation 5-74
5.1.5. Previous Inhalation Assessment 5-75
5.2. QUANTITATIVE CANCER ASSESSMENT BASED ON THE NATIONAL
CANCER INSTITUTE COHORT STUDY 5-75
5.2.1. Choice of Epidemiology Study 5-76
5.2.2. Nasopharyngeal Cancer 5-78
5.2.2.1. Exposure-Response Modeling of the National Cancer
Institute Cohort 5-78
5.2.2.2. Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer
Mortality 5-81
5.2.2.3. Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer
Incidence 5-83
5.2.2.4. Sources of Uncertainty 5-85
5.2.3. Lymphohematopoietic Cancer 5-90
5.2.3.1. Exposure-Response Modeling of the National Cancer
Institute Cohort 5-90
5.2.3.2. Prediction of Lifetime Extra Risks for Hodgkin Lymphoma
and Leukemia Mortality 5-93
5.2.3.3. Prediction of Lifetime Extra Risks for Hodgkin Lymphoma
and Leukemia Incidence 5-96
5.2.3.4. Sources of Uncertainty 5-98
This document is a draft for review purposes only and does not constitute Agency policy.
I-ix DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
5.2.4. Conclusions on Cancer Unit Risk Estimates Based on Human Data 5-102
5.3. DOSE-RESPONSE MODELING OF RISK OF SQUAMOUS CELL
CARCINOMA IN THE RESPIRATORY TRACT USING ANIMAL DATA 5-104
5.3.1. Long-Term Bioassays in Laboratory Animals 5-108
5.3.1.1. Nasal Tumor Incidence Data 5-108
5.3.1.2. Mechanistic Data 5-110
5.3.2. The CUT Biologically Based Dose-Response Modeling 5-111
5.3.2.1. Major Results of the CUT Modeling Effort 5-115
5.3.3. This Assessment's Conclusions from Evaluation of Dose-Response
Models of DPX, Cell-Replication and Genomics Data, and of BBDR
Models for Risk Estimation 5-115
5.3.4. Benchmark Dose Approaches to Rat Nasal Tumor Data 5-122
5.3.4.1. Benchmark Dose Derived from BBDR Rat Model andFlux
as Dosimeter 5-122
5.3.4.2. Comparison with Other Benchmark Dose Modeling Efforts 5-129
5.3.4.3. Kaplan-Meier Adjustment 5-132
5.3.4.4. EPA Time-to-Tumor Statistical Modeling 5-132
5.4. CONCLUSIONS FROM THE QUANTITATIVE ASSESSMENT OF
CANCER RISK FROM FORMALDEHYDE EXPOSURE BY INHALATION ..5-138
5.4.1. Inhalation Unit Risk Estimates Based on Human Data 5-138
5.4.2. Inhalation Unit Risk Estimates Based on Rodent Data 5-138
5.4.3. Summary of Inhalation Unit Risk Estimates 5-139
5.4.4. Application of Age-Dependent Adjustment Factors (ADAFs) 5-140
5.4.5. Conclusions: Cancer Inhalation Unit Risk Estimates 5-142
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND
DOSE-RESPONSE 6-1
6.1. SUMMARY OF HUMAN HAZARD POTENTIAL 6-1
6.1.1. Exposure 6-1
6.1.2. Absorption, Distribution, Metabolism, and Excretion 6-1
6.1.3. Noncancer Health Effects in Humans and Laboratory Animals 6-4
6.1.3.1. Sensory Irritation 6-4
6.1.3.2. Respiratory Tract Pathology 6-6
6.1.3.3. Effects on Pulmonary Function 6-8
6.1.3.4. Asthmatic Responses and Increased Atopic Symptoms 6-9
6.1.3.5. Effects on the Immune System 6-11
6.1.3.6. Neurological Effects 6-11
6.1.3.7. Reproductive and Developmental Effects 6-12
6.1.3.8. Effects on General Systemic Toxicity 6-14
6.1.3.9. Summary 6-14
6.1.4. Carcinogenicity in Humans and Laboratory Animals 6-14
6.1.4.1. Carcinogenicity in Humans 6-15
This document is a draft for review purposes only and does not constitute Agency policy.
I-x DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
6.1.4.2. Carcinogenicity in Laboratory Animals 6-21
6.1.4.3. Carcinogenic Mode(s) of Action 6-22
6.1.5. Cancer Hazard Characterization 6-25
6.2. DOSE-RESPONSE CHARACTERIZATION 6-25
6.2.1. Noncancer Toxicity: Reference Concentration (RfC) 6-25
6.2.1.1. Assessment Approach Employed 6-26
6.2.1.2. Derivation of Candidate Reference Concentrations 6-26
6.2.1.3. Adequacy of Overall DataBase for RfC Derivation 6-27
6.2.1.4. Uncertainties in the Reference Concentration (RfC) 6-29
6.2.1.5. Conclusions 6-32
6.2.2. Cancer Risk Estimates 6-33
6.2.2.1. Choice of Data 6-33
6.2.2.2. Analysis of Epidemiologic Data 6-34
6.2.2.3. Analysis of Laboratory Animal Data 6-36
6.2.2.4. Extrapolation Approaches 6-37
6.2.2.5. Inhalation Unit Risk Estimates for Cancer 6-41
6.2.2.6. Early-Life Susceptibility 6-41
6.2.2.7'. Uncertainties in the Quantitative Risk Estimates 6-42
6.2.2.8. Conclusions 6-44
6.3. SUMMARY AND CONCLUSIONS 6-45
REFERENCES R-l
VOLUME IV
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITIONS A-2
APPENDIX B: SIMULATIONS OF INTERINDIVIDUAL AND ADULT-TO-CHILD
VARIABILITY IN REACTIVE GAS UPTAKE IN A SMALL
SAMPLE OF PEOPLE (Garcia et al., 2009) B-2
APPENDIX C: LIFETABLE ANALYSIS C-2
APPENDIX D: MODEL STRUCTURE & CALIBRATION IN CONOLLY ET AL.
(2003,2004) D-2
D.I. DPX AND MUTATIONAL ACTION D-4
D.2. CALIBRATION OF MODEL D-4
D.3. FLUX BINS D-5
D.4. USE OF LABELING DATA D-5
D.5. UPWARD EXTRAPOLATION OF NORMAL CELL DIVISION RATE D-5
This document is a draft for review purposes only and does not constitute Agency policy.
I-xi DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
D.6. INITIATED CELL DIVISION AND DEATH RATES D-7
D.7. STRUCTURE OF THE CUT HUMAN MODEL D-7
APPENDIX E: EVALUATION OF BBDR MODELING OF NASAL CANCER IN
THE F344 RAT: CONOLLY ET AL. (2003) AND ALTERNATIVE
IMPLEMENTATIONS E-2
E.I. TABULATION OF ALL ISSUES EVALUATED IN THE RAT MODELS E-2
E.2. STATISTICAL METHODS USED IN EVALUATION E-7
E.3. PRIMARY UNCERTAINTIES IN BBDR MODELING OF THE F344 RAT
DATA E-8
E.3.1. Sensitivity to Use of Historical Controls E-9
E.3.1.1. Use of Historical Controls E-9
E.3.1.2. Influence of Historical Controls on Model Calibration and
on Human Model E-10
E.3.1.3. Influence of Historical Controls on Dose-Response Curve E-13
E.3.1.4. Problem Including 1976 Study for Inhalation Historical
Control E-13
E.3.1.5. Effect of Control Data on MOA Inferences E-14
E.3.2. Characterization of Uncertainty-Variability in Cell Replication Rates E-14
E.3.2.1. Dose-Response for DN as Used in the CUT Clonal Growth
Modeling E-14
E.3.2.2. Time Variability in Labeling Data E-17
E.3.2.3. Site and Time Variability in Derived Cell Replication Rate E-18
E.3.2.4. Alternate Dose-Response Curves for Cell Replication E-22
E.3.3. Uncertainty in Model Specification of Initiated Cell Replication and
Death E-30
E.3.3.1. Biological Implications of Assumptions in Conolly et al.
(2003) E-30
E.3.3.2. Plausible Alternative Assumptions for ai and PI E-32
E.3.4. Results of Sensitivity Analyses on (XN, ai, and PI E-33
E.3.4.1. Further Constraints E-33
E.3.4.2. Sensitivity of Risk Estimates for the F344 Rat E-34
E.3.4.3. MOA Inferences Revisited E-41
E.3.4.4. Confidence Bounds: Model Uncertainty Versus Statistical
Uncertainty E-42
APPENDIX F: SENSITIVITY ANALYSIS OF BBDR MODEL FOR
FORMALDEHYDE INDUCED RESPIRATORY CANCER IN
HUMANS F-2
F.I. MAJOR UNCERTAINTIES IN THE FORMALDEHYDE HUMAN BBDR
MODEL F-2
This document is a draft for review purposes only and does not constitute Agency policy.
I-xii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
F.2. SENSITIVITY ANALYSIS OF HUMAN BBDR MODELING F-4
F.2.1. Effect of Background Rates of Nasal Tumors in Rats on Human Risk
Estimates F-4
F.2.2. Alternative Assumptions Regarding the Rate of Replication of Initiated
Cells F-6
F.2.3. Biological Plausibility of Alternate Assumptions F-10
F.2.4. Effect of Alternate Assumptions for Initiated Cell Kinetics on Human
Risk Estimates F-13
APPENDIX G: EVALUATION OF THE CANCER DOSE-RESPONSE MODELING G-2
G.I. MAJOR CONCLUSIONS IN ANDERSEN ET AL. (2008) G-2
G.2. USE OF MULTIPLE FILTERS ON THE DATA G-3
G.3. DATA FOR LOW-DOSE CANCER RESPONSE G-4
G.4. DIFFICULTIES IN INTERPRETING THE BENCHMARK MODELING G-5
G.5. STATISTICAL SENSITIVITY OF THE DATA FOR DOSE-RESPONSE G-6
G.6. LENGTH OF THE STUDY AND STOCHASTIC EVENTS G-6
G.7. OVERALL CONCLUSION G-7
APPENDIX H: EXPERT PANEL CONSULTATION ON QUANTITATIVE
EVALUATION OF ANIMAL TOXICOLOGY DATA FOR
ANALYZING CANCER RISK DUE TO INHALED
FORMALDEHYDE H-2
This document is a draft for review purposes only and does not constitute Agency policy.
I-xiii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table 2-1. Physicochemical properties of formaldehyde 2-2
Table 2-2. Ambient air levels by land use category 2-6
Table 2-3. Studies on residential indoor air levels of formaldehyde (nonoccupational) 2-9
Table 3-1. Formaldehyde kinetics in human and rat tissue samples 3-17
Table 3-2. Allelic frequencies of ADH3 in human populations 3-18
Table 3-3. Endogenous formaldehyde levels in animal and human tissues and body fluids. 3-25
Table 3-4. Percent distribution of airborne [14C]-formaldehyde in F344 rats 3-27
Table 3-5. Measurements of exhaled formaldehyde concentrations in the mouth and
nose, and in the oral cavity after breath holding in three healthy male
laboratory workers 3-31
Table 3-6. Formaldehyde, methanol and ethanol levels reported in the exhaled breath
of 34 subjects 3-32
Table 3-7. Apparent formaldehyde levels (ppb) in exhaled breath of individuals
attending a health fair, adjusted for methanol and ethanol levels which
contribute to the detection of the protonated species with a mass to charge
ratio of 31 reported as formaldehyde (m/z = 31) 3-33
Table 3-8. Extrapolation of parameters for enzymatic metabolism to the human 3-55
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function 4-14
Table 4-2. Major cohort studies of formaldehyde exposure and nasopharyngeal cancer 4-99
Table 4-3. Case-control studies of formaldehyde exposure and nasopharyngeal cancer 4-102
Table 4-4. Case-control studies of formaldehyde exposure and nasal and paranasal
(sinonasal) cancers 4-122
Table 4-5. Epidemiologic studies of formaldehyde and pharyngeal cancer 4-136
Table 4-6. Comparison of SMRs in the exposed workers to the SMRs in the unexposed
workers from NCI cohort reported by Beane-Freeman et al. (2009) 4-146
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers 4-152
LIST OF TABLES (continued)
This document is a draft for review purposes only and does not constitute Agency policy.
I-xiv DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-8. Respiratory effects of formaldehyde-induced reflex bradypnea in various
strains of mice 4-193
Table 4-9. Respiratory effects of formaldehyde-induced reflex bradypnea in various
strains of rats 4-194
Table 4-10. Inhaled dose of formaldehyde to nasal mucosa of F344 rats and B6C3F1
mice exposed to 15 ppm 4-197
Table 4-11. Exposure regimen for cross-tolerance study 4-198
Table 4-12. Summary of formaldehyde effects on mucociliary function in the upper
respiratory tract 4-208
Table 4-13. Cell proliferation in nasal mucosa, trachea, and free lung cells isolated from
male Wistar rats after inhalation exposures to formaldehyde 4-214
Table 4-14. The effect of repeated formaldehyde inhalation exposures for 3 months on
cell count, basal membrane length, proliferation cells, and two measures of
cell proliferation, LI and ULLI, in male F344 rats 4-217
Table 4-15. Formaldehyde-induced changes in cell proliferation (ULLI) in the nasal
passages of male F344 rats exposed 6 hours/day 4-219
Table 4-16. Cell population and surface area estimates in untreated male F344 rats and
regional site location of squamous cell carcinomas in formaldehyde-
exposed rats for correlation to cell proliferation rates 4-220
Table 4-17. Summary of formaldehyde effects on cell proliferation in the upper
respiratory tract 4-222
Table 4-18. Concentration regimens for ultrastructural evaluation of male CDF rat
nasoturbinates 4-227
Table 4-19. Enzymatic activities in nasal respiratory epithelium of male Wistar rats
exposed to formaldehyde, ozone, or both 4-229
Table 4-20. Lipid analysis of lung tissue and lung lavage from male F344 rats exposed
to 0, 15, or 145.6 ppm formaldehyde for 6 hours 4-236
Table 4-21. Formaldehyde effects on biochemical parameters in nasal mucosa and
lung tissue homogenates from male F344 rats exposed to 0, 15, or
145.6 ppm formaldehyde for 6 hours 4-237
LIST OF TABLES (continued)
This document is a draft for review purposes only and does not constitute Agency policy.
I-xv DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-22. Mast cell degranulation and neutrophil infiltration in the lung of rats
exposed to formaldehyde via inhalation 4-238
Table 4-23. Summary of respiratory tract pathology from inhalation exposures to
formaldehyde—short-term studies 4-241
Table 4-24. Location and incidence of respiratory tract lesions in B6C3F1 mice
exposed to formaldehyde 4-244
Table 4-25. Formaldehyde effects (incidence and severity) on histopathologic changes
in the noses and larynxes of male and female albino SPF Wistar rats
exposed to formaldehyde 6 hours/day for 13 weeks 4-246
Table 4-26. Formaldehyde-induced nonneoplastic histopathologic changes in male
albino SPF Wistar rats exposed to 0, 10, or 20 ppm formaldehyde and
examined at the end of 130 weeks inclusive of exposure 4-247
Table 4-27. Formaldehyde-induced nasal tumors in male albino SPF Wistar rats
exposed to formaldehyde (6 hours/day, 5 days/week for 13 weeks) and
examined at the end of 130 weeks inclusive of exposure 4-248
Table 4-28. Formaldehyde effects on nasal epithelium for various concentration-by-
time products in male albino Wistar rats 4-251
Table 4-29. Rhinitis observed in formaldehyde-treated animals; data pooled for male
and female animals 4-253
Table 4-30. Epithelial lesions found in the middle region of nasoturbinates of
formaldehyde-treated and control animals; data pooled for males and
females 4-253
Table 4-31. Cellular and molecular changes in nasal tissues of F344 rats exposed to
formaldehyde 4-255
Table 4-32. Percent body weight gain and concentrations of iron, zinc, and copper in
cerebral cortex of male Wistar rats exposed to formaldehyde via inhalation
for 4 and 13 weeks 4-256
Table 4-33. Zinc, copper, and iron content of lung tissue from formaldehyde-treated
male Wistar rats 4-257
Table 4-34. Total lung cytochrome P450 measurements of control and formaldehyde-
treated male Sprague-Dawley rats 4-258
This document is a draft for review purposes only and does not constitute Agency policy.
I-xvi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-35. Cytochrome P450 levels in formaldehyde-treated rats 4-258
Table 4-36. Summary of respiratory tract pathology from inhalation exposures to
formaldehyde, subchronic studies 4-261
Table 4-37. Histopathologic findings and severity scores in the naso- and
maxilloturbinates of female Sprague-Dawley rats exposed to inhaled
formaldehyde and wood dust for 104 weeks 4-265
Table 4-38. Histopathologic changes (including tumors) in nasal cavities of male
Sprague-Dawley rats exposed to inhaled formaldehyde or HC1 alone and
in combination for a lifetime 4-269
Table 4-39. Summary of neoplastic lesions in the nasal cavity of F344 rats exposed to
inhaled formaldehyde for 2 years 4-273
Table 4-40. Apparent sites of origin for the SCCs in the nasal cavity of F344 rats
exposed to 14.3 ppm of formaldehyde gas in the Kerns et al. (1983) bioassay... 4-273
Table 4-41. Incidence and location of nasal squamous cell carcinoma in male F344
rats exposed to inhaled formaldehyde for 2 years 4-275
Table 4-42. Summary of respiratory tract pathology from chronic inhalation
exposures to formaldehyde 4-282
Table 4-43. Summary of lesions observed in the gastrointestinal tracts of Wistar rats
after drinking-water exposure to formaldehyde for 4 weeks 4-293
Table 4-44. Incidence of lesions observed in the gastrointestinal tracts of Wistar rats
after drinking-water exposure to formaldehyde for 2 years 4-296
Table 4-45. Effect of formaldehyde on gastroduodenal carcinogenesis initiated by
MINING and NaCl in male Wistar rats exposed to formaldehyde in
drinking water for 8 weeks 4-299
Table 4-46. Battery of immune parameters and functional tests assessed in female
B6C3F1 mice after a 3 week, 15-ppm formaldehyde exposure 4-302
Table 4-47. Summary of the effects of formaldehyde inhalation on the mononuclear
phagocyte system (MPS) in female B6C3F1 mice after a 3 week, 15 ppm
formaldehyde exposure (6 hours/day, 5 days/week) 4-303
Table 4-48. Formaldehyde exposure regimens for determining the effects of
formaldehyde exposure on pulmonary S. aureus infection 4-305
This document is a draft for review purposes only and does not constitute Agency policy.
I-xvii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-49. Summary of immune function changes due to inhaled formaldehyde
exposure in experimental animals 4-310
Table 4-50. Study design for guinea pigs exposed to formaldehyde through different
routes of exposure: inhalation, dermal, and injection 4-316
Table 4-51. Sensitization response of guinea pigs exposed to formaldehyde through
inhalation, topical application, or footpad injection 4-317
Table 4-52. Cytokine and chemokine levels in lung tissue homogenate supernatants in
formaldehyde-exposed male ICR mice with and without Der f sensitization 4-324
Table 4-53. Correlation coefficients among ear swelling responses and skin mRNA
levels in contact hypersensitivity to formaldehyde in mice 4-333
Table 4-54. Summary of sensitization and atopy studies by inhalation or dermal
sensitization due to formaldehyde in experimental animals 4-335
Table 4-55. Fluctuation of behavioral responses when male AB mice inhaled
formaldehyde in a single 2-hour exposure: effects after 2 hours 4-343
Table 4-56. Fluctuation of behavioral responses when male AB mice inhaled
formaldehyde in a single 2-hour exposure: effects after 24 hours 4-343
Table 4-57. Effects of formaldehyde exposure on completion of the labyrinth test
by male and female LEW. IK rats 4-347
Table 4-58. Summary of neurological and neurobehavioral studies of inhaled
formaldehyde in experimental animals 4-363
Table 4-59. Effects of formaldehyde on body and organ weights in rat pups from
dams exposed via inhalation from mating through gestation 4-373
Table 4-60. Formaldehyde effects on Leydig cell quantity and nuclear damage in
adult male Wistar rats 4-382
Table 4-61. Formaldehyde effects on adult male albino Wistar rats 4-383
Table 4-62. Formaldehyde effects on testosterone levels and seminiferous tubule
diameters in Wistar rats following 91 days of exposure 4-384
Table 4-63. Effects of formaldehyde exposure on seminiferous tubule diameter and
epithelial height in Wistar rats following 18 weeks of exposure 4-3 86
This document is a draft for review purposes only and does not constitute Agency policy.
I-xviii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-64. Incidence of sperm abnormalities and dominant lethal effects in
formaldehyde-treated mice 4-386
Table 4-65. Body weights of pups born to beagles exposed to formaldehyde during
gestation 4-388
Table 4-66. Testicular weights, sperm head counts, and percentage incidence of
abnormal sperm after oral administration of formaldehyde to male Wistar
rats 4-389
Table 4-67. Effect of formaldehyde on spermatogenic parameters in male Wistar
rats exposed intraperitoneally 4-391
Table 4-68. Incidence of sperm head abnormalities in formaldehyde-treated rats 4-392
Table 4-69. Dominant lethal mutations after exposure of male rats to formaldehyde 4-393
Table 4-70. Summary of reported developmental effects in formaldehyde inhalation
exposure studies 4-395
Table 4-71. Summary of reported developmental effects in formaldehyde oral
exposure studies 4-401
Table 4-72. Summary of reported developmental effects in formaldehyde dermal
exposure studies 4-402
Table 4-73. Summary of reported reproductive effects in formaldehyde inhalation
studies 4-403
Table 4-74. Summary of reported reproductive effects in formaldehyde oral studies 4-407
Table 4-75. Summary of reported reproductive effects in formaldehyde
intraperitoneal studies 4-408
Table 4-76. Formaldehyde-DNA reactions (DPX formation) 4-410
Table 4-77. Formaldehyde-DNA reactions (DNA adduct formation) 4-416
Table 4-78. Formaldehyde-DNA interactions (single strand breaks) 4-419
Table 4-79. Other genetic effects of formaldehyde in mammalian cells 4-420
Table 4-80. In vitro clastogenicity of formaldehyde 4-422
LIST OF TABLES (continued)
This document is a draft for review purposes only and does not constitute Agency policy.
I-xix DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-81. Summary of mutagenicity of formaldehyde in bacterial systems 4-426
Table 4-82. Mutagenicity in mammalian cell systems 4-430
Table 4-83. Genotoxicity in laboratory animals 4-436
Table 4-84. MN frequencies in buccal mucosa cells of volunteers exposed to
formaldehyde 4-440
Table 4-85. MN and SCE formation in mortuary science students exposed to
formaldehyde for 85 days 4-441
Table 4-86. Incidence of MN formation in mortuary students exposed to
formaldehyde for 90 days 4-442
Table 4-87. Multivariate repression models linking genomic instability/occupational
exposures to selected endpoints from the MN assay 4-447
Table 4-88. Genotoxicity measures in pathological anatomy laboratory workers
and controls 4-447
Table 4-89. Summary of human cytogenetic studies 4-450
Table 4-90. Summary of cohort and case-control studies which evaluated the
incidence of all LHP cancers in formaldehyde-exposed populations and
all leukemias 4-491
Table 4-91. Secondary analysis of published mortality statistics to explore alternative
disease groupings within the broad category of all lymphohematopoietic
malignancies 4-493
Table 4-92. Summary of studies which provide mortality statistics for myeloid
leukemia subtypes 4-498
Table 4-93. Summary of mortality statistics for Hodgkin's lymphoma, lymphoma and
multiple myeloma from cohort analyses of formaldehyde exposed workers 4-501
Table 4-94. Leukemia incidence in Fischer 344 rats surviving at least 21 months 4-506
Table 4-95. Available evidence for susceptibility factors of concern for formaldehyde
exposure 4-549
Table 5-1. Candidate points of departure (PODs) including duration adjustments for
nervous system toxicity in key human and animal studies 5-7
LIST OF TABLES (continued)
This document is a draft for review purposes only and does not constitute Agency policy.
I-xx DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-2. Effects of formaldehyde exposure on completion of the labyrinth test by
male and female LEW. IK rats 5-23
Table 5-3. Candidate PODs including duration adjustments for developmental and
reproductive toxicity in key animal studies 5-28
Table 5-4. Summary of candidate studies for formaldehyde RfC development by
health endpoint category 5-37
Table 5-5. Adjustment for nonoccupational exposures to formaldehyde 5-63
Table 5-6. Summary of reference concentration (RfC) derivation from critical study
and supporting studies 5-67
Table 5-6. Summary of reference concentration (RfC) derivation from critical study
and supporting studies (continued) 5-68
Table 5-7. Relative risk estimates for mortality from nasopharyngeal malignancies
(ICD 8 code 147) by level of formaldehyde exposure for different exposure
metrics 5-80
Table 5-8. Regression coefficients from NCI log-linear trend test models for NPC
mortality from cumulative exposure to formaldehydea 5-81
Table 5-9. Extra risk estimates for NPC mortality from various levels of continuous
exposure to formaldehyde 5-82
Table 5-10. EC0005, LEC0005, and inhalation unit risk estimates for NPC mortality
from formaldehyde exposure based on the Hauptmann et al. (2004) log-
linear trend analyses for cumulative exposure 5-83
Table 5-11. EC0005, LEC0005, and inhalation unit risk estimates for NPC incidence
from formaldehyde exposure based on the Hauptmann et al. (2004) trend
analyses for cumulative exposure 5-84
Table 5-12. Relative risk estimates for mortality from Hodgkin lymphoma (ICD 8 code
201) and leukemia (ICD 8 codes 204-207) by level of formaldehyde
exposure for different exposure metrics 5-92
Table 5-13. Regression coefficients for Hodgkin lymphoma and leukemia mortality
from NCI trend test models 5-93
Table 5-14. Extra risk estimates for Hodgkin lymphoma mortality from various levels
of continuous exposure to formaldehyde 5-94
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 5-15. Extra risk estimates for leukemia mortality from various levels of
continuous exposure to formaldehyde 5-94
Table 5-16. EC0005, LEC0005, and inhalation unit risk estimates for Hodgkin
lymphoma mortality from formaldehyde exposure based on Beane
Freeman et al. (2009) log-linear trend analyses for cumulative exposure 5-95
Table 5-17. EC005, LEC005, and inhalation unit risk estimates for leukemia mortality
from formaldehyde exposure based on Beane Freeman et al. (2009) log-
linear trend analyses for cumulative exposure 5-96
Table 5-18. EC0005, LEC0005, and inhalation unit risk estimates for Hodgkin
lymphoma incidence from formaldehyde exposure, based on Beane
Freeman et al. (2009) log-linear trend analyses for cumulative exposure 5-97
Table 5-19. EC005, LEC005, and inhalation unit risk estimates for leukemia incidence
from formaldehyde exposure based on Beane Freeman et al. (2009) log-
linear trend analyses for cumulative exposure 5-97
Table 5-20. Calculation of combined cancer mortality unit risk estimate at 0.1 ppm 5-103
Table 5-21. Calculation of combined cancer incidence unit risk estimate at 0.1 ppm 5-103
Table 5-22. Summary of tumor incidence in long-term bioassays on F344 rats 5-109
Table 5-23. BMD modeling of unit risk of SCC in the human respiratory tract 5-129
Table 5-24. Human benchmark extrapolations of nasal tumors in rats by using
formaldehyde flux and DPX 5-132
Table 5-25. Formaldehyde-induced rat tumor incidences 5-133
Table 5-26. Summary of inhalation unit risk estimates 5-139
Table 5-27. Extra risk estimates per ppm based on ECs 5-140
Table 5-28. Total cancer risk from exposure to a constant formaldehyde exposure
level of 1 ug/m3 from ages 0-70 years 5-141
Table 6-1. Summary of candidate reference concentrations (RfC) for cocritical studies 6-27
Table 6-2. Effective concentrations (lifetime continuous exposure levels) predicted for
specified extra cancer risk levels for selected formaldehyde-related cancers 6-35
LIST OF TABLES (continued)
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxii DRAFT—DO NOT CITE OR QUOTE
-------
Table 6-3. Inhalation unit risk estimates based on epidemiological and experimental
animal data 6-41
Table B-1. Variations in overall nasal uptake, whole nose flux, and key parameters B-4
Table C-l. Extra risk calculation for environmental exposure to 0.0461 ppm
formaldehyde (the LEC0005 for NPC incidence) using a log-linear exposure-
response model based on the cumulative exposure trend results of
Hauptmann et al. (2004), as described in Section 5.2.2 C-3
Table E-l. Evaluation of assumptions and uncertainties in the CUT model for nasal
tumors in theF344 rat E-3
Table E-2. Influence of control data in modeling formaldehyde-induced cancer in the
F344rat E-ll
Table E-3. Variation in number of cells across nasal sites in the F344 rat E-20
Table E-4. Parameter specifications and estimates for clonal growth models of nasal
SCC in the F344 rat using alternative characterization of cell replication
and death rates E-39
Table E-5. Parameter specifications and estimates for clonal growth models of nasal
SCC in the F344 rat using cell replication and death rates as characterized
in Conolly et al. (2003) E-40
Table E-6. Comparison of statistical confidence bounds on added risk for two models E-42
Table F-1. Summary of evaluation of maj or assumptions and results in CUT human
BBDR model F-3
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxiii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
Figure 2-1. Chemical structure of formaldehyde 2-1
Figure 2-2. Locations of hazardous air pollutant monitors 2-6
Figure 2-3. Modeled ambient air concentrations based on 1999 emissions 2-7
Figure 2-4. Range of formaldehyde air concentrations (ppb) in different environments 2-10
Figure 3-1. Formaldehyde-mediated protein modifications 3-3
Figure 3-2. 3H/14C ratios in macromolecular extracts from rat tissues following exposure
to 14C-and 3H-labeled formaldehyde 3-14
Figure 3-3. Formaldehyde clearance by ALDH2 (GSH-independent) and ADH3
(GSH-dependent) 3-16
Figure 3-4. Metabolism of formate 3-18
Figure 3-5. Detection of the characteristic analytical product ion for formaldehyde by
proton transfer reaction mass spectrometry (PTR-MS) in gas samples
spiked with only methanol and ethanol 3-29
Figure 3-6. Reconstructed nasal passages of F344 rat, rhesus monkey, and human 3-37
Figure 3-7. Illustration of interspecies differences in airflow and verification of CFD
simulations with water-dye studies 3-38
Figure 3-8. Lateral view of nasal wall mass flux of inhaled formaldehyde simulated
in theF344 rat, rhesus monkey, and human 3-39
Figure 3-9. CFD simulations of formaldehyde flux to human nasal lining at different
inspiratory flow rates 3-40
Figure 3-10. Single-path model simulations of surface flux per ppm of formaldehyde
exposure concentration in an adult male human 3-44
Figure 3-11. Pressure drop vs. volumetric airflow rate predicted by the CUT CFD model
compared with pressure drop measurements made in two hollow molds
of the rat nasal passage or in rats in vivo 3-46
Figure 3-12. Formaldehyde-DPX dosimetry in the F344 rat 3-48
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxiv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 4-1. Delayed asthmatic reaction following the inhalation of formaldehyde
after "painting" 100% formalin for 20 minutes 4-51
Figure 4-2. McGwin Forest plot of relative risk estimates and 95% CIs from studies
included in a meta-analysis of formaldehyde exposure and asthma in
children based on the random effects models 4-60
Figure 4-3. Association between peak formaldehyde exposure and the risk of
lymphohematopoietic malignancy 4-148
Figure 4-4. Association between average intensity of formaldehyde exposure and
the risk of lymphohematopoietic malignancy 4-149
Figure 4-5. Formaldehyde effects on minute volume in naive and formaldehyde-
pretreatedmaleB6C3Fl mice and F344 rats 4-196
Figure 4-6. Sagittal view of the rat nose (nares oriented to the left) 4-202
Figure 4-7. Main components of the nasal respiratory epithelium 4-203
Figure 4-8. Decreased mucus clearance and ciliary beat in isolated frog palates
exposed to formaldehyde after 3 days in culture 4-207
Figure 4-9. Effect of formaldehyde exposure on cell proliferation of the respiratory
mucosaof rats and mice 4-211
Figure 4-10. Diagram of nasal passages, showing section levels chosen for
morphometry and autoradiography in male rhesus monkeys exposed
to formaldehyde 4-233
Figure 4-11. Formaldehyde-induced cell proliferation in male rhesus monkeys
exposed to formaldehyde 4-234
Figure 4-12. Formaldehyde-induced lesions in male rhesus monkeys exposed to
formaldehyde 4-235
Figure 4-13. Frequency and location by cross-section level of squamous metaplasia
in the nasal cavity of F344 rats exposed to formaldehyde via inhalation 4-272
Figure 4-14. Mortality corrected cumulative incidences of nasal carcinomas in the
indicated exposure groups 4-290
Figure 4-15. Alveolar MP Fc-mediated phagocytosis from mice exposed to 5 ppm
formaldehyde, 10 mg/m3 carbon black, or both 4-307
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 4-16. Compressed air in milliliters as parameter for airway obstruction
following formaldehyde exposure in guinea pigs after OVA
sensitization and OVA challenge 4-319
Figure 4-17. OVA-specific IgGl (IB) in formaldehyde-treated sensitized guinea pigs
prior to OVA challenge 4-319
Figure 4-18. Anti-OVA liters in female Balb/C mice exposed to 6.63 ppm
formaldehyde for 10 consecutive days or once a week for 7 weeks 4-320
Figure 4-19. Vascular permeability in the trachea and bronchi of male Wistar rats
after 10 minutes of formaldehyde inhalation 4-322
Figure 4-20. Effect of select receptor antagonists on formaldehyde-induced vascular
permeability in the trachea and bronchi male of Wistar rats 4-323
Figure 4-21. The effects of formaldehyde inhalation exposures on eosinophil
infiltration and goblet cell proliferation after Der f challenge in the
nasal mucosa of male ICRmice after sensitization and challenge 4-325
Figure 4-22. NGF in BAL fluid from formaldehyde-exposed female C3Ft/He mice
with and without OVA sensitization 4-328
Figure 4-23. Plasma substance P levels in formaldehyde-exposed female C3H/He
mice with and without OVA sensitization 4-329
Figure 4-24. Motor activity in male and female rats 2 hours after exposure to
formaldehyde expressed as mean number of crossed quadrants ± SEM 4-340
Figure 4-25. Habituation of motor activity was observed in control rats during the
second observation period 4-341
Figure 4-26. Motor activity was reduced in male and female LEW. IK rats 2 hours
after termination of 10-minute formaldehyde exposure 4-342
Figure 4-27. The effects of the acute formaldehyde exposures on the ambulatory and
vertical components of SLMA 4-344
Figure 4-28. Effects of formaldehyde exposure on the error rate of female LEW. IK
rats performing the water labyrinth learning test 4-348
Figure 4-29. Basal and stress-induced trunk blood corticosterone levels in male
LEW. IK rats after formaldehyde inhalation exposures 4-353
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxvi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 4-30. NGF production in the brains of formaldehyde-exposed mice 4-358
Figure 4-31. DNA-protein cross-links (DPX) and thymidine kinase (tk) mutants in
TK6 human lymphoblasts exposed to formaldehyde for 2 hours 4-433
Figure 4-32. Developmental origins for cancers of the lymphohematopoietic system 4-489
Figure 4-33. Unscheduled deaths in female F344 rats exposed to formaldehyde for
24 months 4-506
Figure 4-34. Cumulative incidence of tumor bearing animals for lymphoma in female
B6C3F1 mice exposed to formaldehyde for 24 months (p < 0.05) 4-507
Figure 4-35. Effect of various doses of formaldehyde on cell number in (A) HT-29
human colon carcinoma cells and in (B) human umbilical vein epithelial
cells (HUVEC) 4-515
Figure 4-36. Integrated MOA scheme for respiratory tract tumors 4-527
Figure 4-37. Location of intraepithelial lymphocytes along side epithelial cells in the
human adenoid 4-530
Figure 5-1. Change in number of additions made in 10 minutes following
formaldehyde exposure at 0.04, 0.21, 0.48, or 1.1 mg/m3 5-22
Figure 5-2. Effects of formaldehyde exposure on the error rate of female LEW. IK rats
performing the water labyrinth learning test 5-25
Figure 5-3. Fecundity density ratio among women exposed to formaldehyde in the high
exposure index category with 8 hour time-weighted average formaldehyde
exposure concentration of 219 ppb 5-31
Figure 5-4. Estimated reduction in peak expiratory flow rate in children in relation to
indoor residential formaldehyde concentrations 5-42
Figure 5-5. Odds ratios for physician-diagnosed asthma in children associated with
in-home formaldehyde levels in air 5-46
Figure 5-6. Prevalence of asthma and respiratory symptom scores in children associated
with in-home formaldehyde levels 5-50
Figure 5-7. Prevalence and severity of allergic sensitization in children associated
with in-home formaldehyde levels 5-50
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxvii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 5-8. Positive exposure-response relationships reported for in-home formaldehyde
exposures and sensory irritation (eye irritation) 5-53
Figure 5-9. Positive exposure-response relationships reported for in-home formaldehyde
exposures and sensory irritation (burning eyes) 5-54
Figure 5-10. Age-specific mortality and incidence rates for myeloid, lymphoid, and
all leukemia 5-101
Figure 5-11. Schematic of integration of pharmacokinetic and pharmacodynamic
components in the CUT model 5-106
Figure 5-12. Fit to the rat tumor incidence data using the model and assumptions in
Conolly et al. (2003) 5-110
Figure 5-13. Spatial distribution of formaldehyde over the nasal lining, as characterized
by partitioning the nasal surface by formaldehyde flux to the tissue per
ppm of exposure concentration, resulting in 20 flux bins 5-124
Figure 5-14. Distribution of cells at risk across flux bins in the F344 rat nasal lining 5-124
Figure 5-15. MLE and upper bound added risk of SCC in the human nose for two
BBDR models 5-127
Figure 5-16. Replot of log-probit fit of the combined Kerns et al. and Monticello et al.
data on tumor incidence showing BMC10 and BMCL10 5-130
Figure 5-17. EPA Multistage Weibull modeling: nasal tumor dose response 5-135
Figure 5-18. Multistage Weibull model fit 5-136
Figure 5-19. Multistage Weibull model fit of tumor incidence data compared with KM
estimates of spontaneous tumor incidence 5-136
Figure B-l. Gas flux across the nasal lining for the case of a "maximum uptake" gas in
Garcia etal. (2009) as a function of axial distance from the nostril B-5
Figure D-l. Dose response of normal (
-------
LIST OF FIGURES (continued)
Figure E-2. Logarithm of normal cell replication rate (XN versus formaldehyde flux
or theF344 rat nasal epithelium E-19
Figure E-3A. Logarithm of normal cell replication rate versus formaldehyde
flux with simultaneous confidence limits for the ALM E-20
Figure E-3B. Logarithm of normal cell replication rate versus formaldehyde flux with
simultaneous confidence limits for thePLM E-21
Figure E-4, Nl. Various dose-response modeling of normal cell replication rate E-25
Figure E-4, N2. Various dose-response modeling of normal cell replication rate E-25
Figure E-4, N3. Various dose-response modeling of normal cell replication rate E-26
Figure E-4, N4. Various dose-response modeling of normal cell replication rate E-26
Figure E-4, N5. Various dose-response modeling of normal cell replication rate E-27
Figure E-4, N6. Various dose-response modeling of normal cell replication rate E-28
Figure E-5A. BBDR models for the rat—models with positive added risk E-35
Figure E-5B. BBDR rat models resulting in negative added risk E-36
Figure E-6A. Models resulting in positive added rat risk: Dose-response for normal
and initiated cell replication E-37
Figure E-6B. Models resulting in negative added rat risk: Dose-response for normal
and initiated cell replication E-38
Figure F-l. Effect of choice of NTP bioassays for historical controls on human risk F-5
Figure F-2. Conolly et al. (2003) hockey-stick model for division rates of initiated
cells in rats and two modified models F-7
Figure F-3. Conolly et al. (2003) J shape model for division rates of initiated cells in
rats and two modified models F-8
Figure F-4. Very similar model estimates of probability of fatal tumor in rats for
three models in Figure F-2 F-9
Figure F-5. Cell proliferation data from Meng et al. (2010) F-12
LIST OF FIGURES (continued)
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxix DRAFT—DO NOT CITE OR QUOTE
-------
Figure F-6. Graphs of the additional human risks estimated by applying these
modified models for ai, using all NTP controls, compared to those
obtained using the original Conolly etal. (2004) model F-13
Figure G-l. Graphs of epithelial hyperplasia versus formaldehyde concentration
with 95% confidence intervals G-7
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxx DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS
ACGIH American Conference of
Governmental Industrial BC
Hygienists BCME
ADAF age-dependent adjustment BDNF
factors
ADH alcohol dehydrogenase BEIR
ADS anterior dorsal septum
AGT O6-alkylguanyl-DNA BfR
alkyltransferase
AIC Akaike Information Criterion BHR
AIE average intensity of exposure BMC
AIHA American Industrial Hygiene BMCL
Association
ALB albumin BMCR
ALDH aldehyde dehydrogenase
ALL acute lymphocytic leukemia BMD
ALM anterior lateral meatus BMDL
ALP alkaline phosphatase
ALS amyotrophic lateral sclerosis BMR
ALT alanine aminotransferase BN
AML acute myelogenous leukemia BrdU
AMM anterior medial BUN
maxilloturbinate BW
AMPase adenosine monophosphatase CA
AMS anterior medial septum CalEPA
ANAE alpha-naphthylacetate esterase
ANOVA analysis of variance CAP
APA American Psychiatric
Association CASRN
ARB Air Resources Board
AST aspartate aminotransferase CAT
ATCM airborne toxic control measure CBMA
ATP adenosine triphosphate
ATPase adenosine triphosphatase CBMN
ATS American Thoracic Society
ATSDR Agency for Toxic Substances CDC
and Disease Registry
AUC area under the curve CDHS
BAL bronchoalveolar lavage
BALT bronchus associated lymphoid CFD
tissue CGM
BBDR biologically based dose CHO
response
bronchial construction
bis(chloromethyl)ether
brain-derived neurotrophic
factor
biologic effects of ionizing
radiation
German Federal Institute for
Risk Assessment
bronchial hyperresponsiveness
benchmark concentration
95% lower bound on the
benchmark concentration
binucleated micronucleated cell
rate fluoresce
benchmark dose
95% lower bound on the
benchmark dose
benchmark response
Brown-Norway
bromodeoxyuridine
blood urea nitrogen
body weight
chromosomal aberrations
California Environmental
Protection Agency
College of American
Pathologists
Chemical Abstracts Service
Registry Number
catalase
cytokinesis-blocked
micronucleus assay
cytokinesis-blocked
micronucleus
U.S. Centers for Disease
Control and Prevention
California Department of
Health Services
computational fluid dynamics
clonal growth model
Chinese hamster ovary
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
CI confidence interval FEMA
CUT Chemical Industry Institute of FEV1
Toxicology
CLL chronic lymphocytic leukemia FISH
CML chronic myelogenous leukemia
CNS central nervous system FSH
CC>2 carbon dioxide FVC
COEHHA California Office of GALT
Environmental Health Hazard GC-MS
Assessment
CREB cyclic AMP responsive GD
element binding proteins GI
CS conditioned stimulus GO
C x t concentration times time G6PDH
DA Daltons
DAF dosimetric adjustment factor GPX
DDC/DDX DNA-DNA cross-links GR
DEI daily exposure index GM-CSF
DEN diethylnitrosamine
Der f common dust mite allergen GSH
DMG dimethylglycine GSNO
DMGDH dimethylglycine GST
dehydrogenase HAP
DNA deoxyribonucleic acid Hb
OOP AC 3,4-dihydroxyphenylacetic acid HC1
DPC/DPX DNA-protein cross-links HCT
EBV Epstein-Barr virus HEC
EC effective concentration
ED effective dose 5-HIAA
EHC Environmental Health hm
Committee HMGSH
ELISA enzyme-linked immunosorbent HPA
assay HPG
EPA U.S. Environmental Protection HPLC
Agency
ERPG emergency response planning HPRT
guideline
ET ethmoid turbinates HR
FALDH formaldehyde dehydrogenase HSA
FDA U.S. Food and Drug HSDB
Administration
FDR fecundability density ratio Hsp
FEF forced expiratory flow HUVEC
Federal Emergency
Management Agency
forced expiratory volume in 1
second
fluorescent in situ
hybridization
follicle-stimulating hormone
forced vital capacity
gut-associated lymphoid tissue
gas chromatography-mass
spectrometry
gestation day
gastrointestinal
gene ontology
glucose-6-phosphate
dehydrogenase
glutathione peroxidase
glutathione reductase
granulocyte macrophage-
colony-stimulating factor
reduced glutathione
S-nitrosoglutathione
glutathione S-transferase
hazardous air pollutant
hemoglobin
hydrochloric acid
hematocrit
human equivalent
concentration
5-hydroxyindoleacetic acid
hydroxymethyl
S-hydroxymethylglutathione
hypothalamic-pituitary adrenal
hypothalamo-pituitary-gonadal
high-performance liquid
chromatography
hypoxanthine-guanine
phosphoribosyltransferase
high responders
human serum albumin
Hazardous Substances Data
Bank
heat shock protein
human umbilical vein
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
endothelial cell
HWE healthy worker effect
I cell initiated cell
I ARC International Agency for
Research on Cancer
ICD International Classification of
Diseases
IF interfacial
IFN interferon
Ig immunoglobulin
IL interleukin
IP. intraperitoneal
IPCS International Programme on
Chemical Safety
IRIS Integrated Risk Information
System
Km Michaels-Menton constant
KM Kaplan-Meier
LD50 median lethal dose
LDH lactate dehydrogenase
LEG 95% lower bound on the
effective concentration
LED 95% lower bound on the
effective dose
LHP lymphohematopoietic
LI labeling index
LM Listeria monocytogenes
LMS linearized multistage
LLNA local lymph node assay
LOAEL lowest-observed-adverse-effect
level
LPS lipopolysaccharide
LR low responders
LRT lower respiratory tract
MA methylamine
MALT mucus-associated lymph
tissues
MCH mean corpuscular hemoglobin
MCHC mean corpuscular hemoglobin
concentration
MCS multiple chemical sensitivity
MCV mean corpuscular volume
MDA malondialdehyde
MEF maximal expiratory flow
ML myeloid leukemia
MLE maximum likelihood estimate
MMS methyl methane sulfonate
MMT medial maxilloturbinate
MN micronucleus, micronuclei
MNNG N-methyl-N'-nitro-N-
nitrosoguanidine
MO A mode of action
MoDC monocyte-derived dendritic
cell
MP macrophage
MPD multistage polynomial degree
MPS mononuclear phagocyte system
MRL minimum risk level
mRNA messenger ribonucleic acid
MVE-2 Murray Valley encephalitis
virus
MVK Moolgavkar, Venzon, and
Knudson
N cell normal cell
NaCl sodium chloride
NAD+ nicotinamide adenine
dinucleotide
NADH reduced nicotinamide adenine
dinucleotide
NALT nasally associated lymphoid
tissue
NATA National-Scale Air Toxics
Assessment
NCEA National Center for
Environmental Assessment
NCHS National Center for Health
Statistics
NCI National Cancer Institute
NEG Nordic Expert Group
NER nucleotide excision repair
NGF nerve growth factor
NHL non-Hodgkin's lymphoma
NHMRC/ARMCANZ National
Health and Medical Research
Council/Agriculture and
Resource Management Council
of Australia and New Zealand
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxiii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
NNK 4-(methylnitrosamino)-
l-(3-pyridyl)-butanone PEF
N6-hmdA N6-hydroxymethyldeoxy- PEFR
adenosine PEL
N4-hmdC N4-hydroxymethyldeoxy- PFC
cytidine PG
N2-hmdG N2-hydroxymethyldeoxy- PHA
guanosine PLA2
NICNAS National Industrial Chemicals PI
Notification and Assessment PLM
Scheme PMA
NIOSH National Institute for
Occupational Safety and PMR
Health PMS
NLM National Library of Medicine PND
NMDA N-methyl-D-aspartate POD
NMU N-methyl-N-nitrosourea POE
NNK nitrosamine, 4- PTZ
(methylnitrosamino)- l-(3- PUFA
pyridyl)-l-butanone PWULLI
NO nitric oxide
NOAEL no-observed-adverse-effect RA
level RANTES
NPC nasopharyngeal cancer
NRBA neutrophil respiratory burst
activity RB
NRC National Research Council RBC
NTP National Toxicology Program RD50
OR odds ratio
OSHA Occupational Safety and
Health Administration REL
OTS Office of Toxic Substances RfC
OVA ovalbumin RfD
PBPK physiologically based ROD
pharmacokinetic RGDR
PC Philadelphia chromosome RR
PCA passive cutaneous anaphylaxis RT
PCMR proportionate cancer mortality SAB
ratio SCC
PCNA proliferating cell nuclear SCE
antigen SCG
PCR polymerase chain reaction SD
PCV packed cell volume SDH
PECAM platelet endothelial cell
adhesion molecule
peak expiratory flow
peak expiratory flow rates
permissible exposure limit
plaque-forming cell
periglomerular
phytohemagglutinin
phospholipase A2
phagocytic index
posterior lateral meatus
phorbol 12-myristate 13-
acetate
proportionate mortality ratio
posterior medial septum
postnatal day
point of departure
portal of entry
pentilenetetrazole
polyunsaturated fatty acids
population weighted unit
length labeling index
reflex apnea
regulated upon activation,
normal T-cell expressed and
secreted
reflex bradypnea
red blood cells
exposure concentration that
results in a 50% reduction in
respiratory rate
recommended exposure limit
reference concentration
reference dose
regional gas dose
regional gas dose ratio
relative risk
reverse transcriptase
Science Advisory Board
squamous cell carcinoma
sister chromatid exchange
sodium cromoglycate
standard deviation
succinate dehydrogenase;
sarcosine dehydrogenase
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxiv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
SEER Surveillance, Epidemiology, TL
and End Results TLV
SEM standard error of the mean TNF
SEN sensitizer TP
SH sulfhydryl TRI
SHE Syrian hamster embryo TRPV
SI sensory irritation
SLMA spontaneous locomotor activity TWA
SMR standardized mortality ratio TZCA
SNP single nucleotide UCL
polymorphism UDS
SOD superoxide dismutase UF
SOMedA N6-sulfomethyldeoxyadenosine UFFI
SOMedG N2-sulfomethyldeoxyguanosine
Spl specificity protein ULLI
SPIR standardized proportionate URT
incidence ratio USDA
SSAO semicarbazide-sensitive amine VC
oxidase VOC
SSB single strand breaks WBC
STEL short-term exposure limit WDS
TEA tumor bearing animal WHO
TH T-lymphocyte helper WHOROE
THF tetrahydrofolate
TK toxicokinetics
tail length
threshold limit value
tumor necrosis factor
total protein
Toxic Release Inventory
transient receptor potential
vanilloid
time-weighted average
thiazolidine-4-carboxylate
upper confidence limit
unscheduled DNA synthesis
uncertainty factor
urea formaldehyde foam
insulation
unit length labeling index
upper respiratory tract
U.S. Department of Agriculture
vital capacity
volatile organic compound
white blood cell
wet dog shake
World Health Organization
World Health Organization
Regional Office for Europe
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxv 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 inhalation exposure to
formaldehyde. It is not intended to be a comprehensive treatise on the chemical or toxicological
nature of formaldehyde.
In Chapter 6, Major Conclusions in the Characterization of Hazard and Dose Response,
EPA has characterized its overall confidence in the qualitative and quantitative aspects of hazard
and dose response by addressing knowledge gaps, uncertainties, quality of data, and scientific
controversies. The discussion is intended to convey the limitations of the assessment and to aid
and guide the risk assessor in the ensuing steps of the risk assessment process.
For other general information about this assessment or other questions relating to IRIS,
the reader is referred to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (email address).
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxvi DRAFT—DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
JohnE. Whalan1
EPA-ORD-NCEA
CHEMICAL MANAGERS
Danielle DeVoney2
EPA-ORD-NCEA
AUTHORS
Thomas Bateson
EPA-ORD-NCEA
Sue Makris
EPA-ORD-NCEA
Ravi Subramaniam
EPA-ORD-NCEA
Susan Euling
EPA-ORD-NCEA
Kathleen Raffaele
EPA-ORD-NCEA
Suryanarayana Vulimiri
EPA-ORD-NCEA
Jennifer Jinot
EPA-ORD-NCEA
John Schaum
EPA-ORD-NCEA
Gillian Backus3
EPA-ORD-NCEA
CONTRIBUTORS
John Fox
EPA-ORD-NCEA
Larry Valcovic3
EPA-ORD-NCEA
Stanley Bar one
EPA-ORD-NCEA
David Bayliss3
EPA-ORD-NCEA
Barbara Glenn
EPA-ORD-NCEA
Rosemarie Hakim3
EPA-ORD-NCEA
John J. Vandenberg
EPA-ORD-NCEA
Lisa Vinikoor
EPA-ORD-NCEA
Ted Berner
EPA-ORD-NCEA
Karen Hogan
EPA-ORD-NCEA
Paul White
EPA-ORD-NCEA
David Bussard
EPA-ORD-NCEA
Babasaheb Sonawane
EPA-ORD-NCEA
David Farrar Chad Thompson3
EPA-ORD-NCEA EPA-ORD-NCEA
AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
1 Chemical Manager since July 2003.
2 Chemical Manager since June 2009.
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxvii DRAFT—DO NOT CITE OR QUOTE
-------
CONTRACTOR SUPPORT
The literature search and preliminary drafts of this document as well as support for
editing and formatting were provided by Oak Ridge Institute for Science and Education
(ORISE), Oak Ridge Associated Universities (ORAU), Department of Energy, under
Interagency Agreement (IAG) Project No. 03-18. The ORISE individuals who contributed to
this effort include Sheri Hester, George Holdsworth, Bobette D. Nourse, Wanda Olson, and Lutz
W. Weber.
Assistance with the biologically based dose response model evaluation was provided by
ENVIRON International Corporation of Monroe, Louisiana (subcontractors to ORISE; Project
No. 03-18). The primary scientists involved in the work were Kenny S. Crump and Cynthia Van
Landingham.
REVIEWERS
This document has been provided for review to EPA scientists and interagency reviewers from
other federal agencies and White House Offices.
3 Separated from the Agency prior to final revisions to document.
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxviii DRAFT—DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
INTERNAL EPA REVIEWERS
Daniel Axelrad, PhD Elizabeth Margosches, PhD
Office of Policy, Economics, and Innovation Office of Pollution Prevention and Toxics
Iris Camacho, PhD
Office of Pollution Prevention and Toxics
Timothy McMahon, PhD
Office of Pesticide Programs
Christina Cinalli, PhD
Office of Pollution Prevention and Toxics
Julie Migrin-Sturza, PhD
Office of Policy, Economics, and Innovation
Rebecca Edelstein, PhD
Office of Pollution Prevention and Toxics
Greg Miller, PhD
Office of Children's Health Protection and
Environmental Education
Ernest Falke, PhD
Office of Pollution Prevention and Toxics
Deirdre Murphy, PhD
Office of Air and Radiation
Stiven Foster, PhD
Office of Solid Waste and Emergency
Response
Marion Olson, PhD
EPA Region 2
Greg Fritz
Office of Pollution Prevention and Toxics
Andrea Pfahles-Hutchens, PhD
Office of Pollution Prevention and Toxics
Susan Griffin, PhD
EPA Region 8
Jennifer Seed, PhD
Office of Pollution Prevention and Toxics
Timothy Leighton, PhD
Office of Pesticide Programs
This document is a draft for review purposes only and does not constitute Agency policy.
I-xxxix DRAFT—DO NOT CITE OR QUOTE
-------
This page intentionally left blank
This document is a draft for review purposes only and does not constitute Agency policy.
I-xl DRAFT—DO NOT CITE OR QUOTE
-------
1. INTRODUCTION
1
2
3 This document presents background information and justification for the Integrated Risk
4 Information System (IRIS) Summary of the hazard and dose-response assessment of
5 formaldehyde. IRIS Summaries may include oral reference dose (RfD) and inhalation reference
6 concentration (RfC) values for chronic and other exposure durations, and a carcinogenicity
7 assessment.
8 The RfD and RfC, if derived, provide quantitative information for use in risk assessments
9 for health effects known or assumed to be produced through a nonlinear (presumed threshold)
10 mode of action. The RfD (expressed in units of mg/kg-day) is defined as an estimate (with
11 uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
12 population (including sensitive subgroups) that is likely to be without an appreciable risk of
13 deleterious effects during a lifetime. The inhalation RfC (expressed in units of mg/m3) is
14 analogous to the oral RfD, but provides a continuous inhalation exposure estimate. The
15 inhalation RfC considers toxic effects for both the respiratory system (portal of entry [POE]) and
16 for effects peripheral to the respiratory system (extrarespiratory or systemic effects). Reference
17 values are generally derived for chronic exposures (up to a lifetime), but may also be derived for
18 acute (<24 hours), short-term (>24 hours up to 30 days), and subchronic (>30 days up to 10% of
19 lifetime) exposure durations, all of which are derived based on an assumption of continuous
20 exposure throughout the duration specified. Unless specified otherwise, the RfD and RfC are
21 derived for chronic exposure duration.
22 The carcinogenicity assessment provides information on the carcinogenic hazard
23 potential of the substance in question and quantitative estimates of risk from oral and inhalation
24 exposure may be derived. The information includes a weight-of-evidence judgment of the
25 likelihood that the agent is a human carcinogen and the conditions under which the carcinogenic
26 effects may be expressed. Quantitative risk estimates may be derived from the application of a
27 low-dose extrapolation procedure. If derived, the oral slope factor is a plausible upper bound on
28 the estimate of risk per mg/kg-day of oral exposure. Similarly, an inhalation unit risk is a
29 plausible upper bound on the estimate of risk per ug/m3 air breathed.
30 Development of these hazard identification and dose-response assessments for
31 formaldehyde has followed the general guidelines for risk assessment as set forth by the National
32 Research Council (NRC) (1983). EPA Guidelines and Risk Assessment Forum Technical Panel
33 Reports that may have been used in the development of this assessment include the following:
34 Guidelines for the Health Risk Assessment of Chemical Mixtures (U.S. EPA, 1986a), Guidelines
3 5 for Mutagenicity Risk Assessment (U. S. EPA, 1986b), Recommendations for and Documentation
This document is a draft for review purposes only and does not constitute Agency policy.
1 -1 DRAFT—DO NOT CITE OR QUOTE
-------
1 of Biological Values for Use in Risk Assessment (U.S. EPA, 1988), Guidelines for
2 Developmental Toxicity Risk Assessment (U.S. EPA, 1991), Interim Policy for Particle Size and
3 Limit Concentration Issues in Inhalation Toxicity (U.S. EPA, 1994a), Methods for Derivation of
4 Inhalation Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA,
5 1994b), Use of the Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995),
6 Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines for
1 Neurotoxicity Risk Assessment (U.S. EPA, 1998), Science Policy Council Handbook: Risk
8 Characterization (U.S. EPA, 2000a), Benchmark Dose Technical Guidance Document (U.S.
9 EPA, 2000b), Supplementary Guidance for Conducting Health Risk Assessment of Chemical
10 Mixtures (U.S. EPA, 2000c), A Review of the Reference Dose and Reference Concentration
1 1 Processes (U.S. EPA, 2002a), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
12 Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens
13 (U.S. EPA, 2005b), Science Policy Council Handbook: Peer Review (U. S . EPA, 2006a), and A
1 4 Framework for Assessing Health Risks of Environmental Exposures to Children (U. S . EPA,
15 2006b).
16 The literature search strategy employed for this compound was based on the Chemical
17 Abstracts Service Registry Number (CASRN) and at least one common name. Any pertinent
18 scientific information submitted by the public to the IRIS Submission Desk was also considered
19 in the development of this document. This assessment includes a comprehensive review of
20 literature through April 2009. As periodic literature searches are conducted by EPA for the
21 formaldehyde assessment, additional literature identified through December 2009 is included
22 where that literature was determined to be critical to the assessment. This included a few articles
®
23 which were identified through PubMed searches and publically available as "e-publications" in
24 2009, but have final publication dates of 2010.
This document is a draft for review purposes only and does not constitute Agency policy.
1 -2 DRAFT—DO NOT CITE OR QUOTE
-------
2. BACKGROUND
1
2
3 This chapter provides an overview of the physical and chemical characteristics of
4 formaldehyde. Also provided in this chapter are a description of the production, uses, and
5 sources of formaldehyde and information regarding environmental levels and human exposure.
6 A description of the toxicokinetics and toxicodynamic processes involved in formaldehyde
7 toxicity for the inhalation, oral, and dermal routes can be found in Chapter 3 (Toxicokinetics).
8
9 2.1. PHYSICOCHEMICAL PROPERTIES OF FORMALDEHYDE
10 Formaldehyde (CASRN 50-00-0) is the first of the series of aliphatic aldehydes and is a
11 gas at room temperature. Its molecular structure is depicted in Figure 2-1. It is noted for its
12 reactivity and versatility as a chemical intermediate. It readily undergoes polymerization, is
13 highly flammable, and can form explosive mixtures with air. It decomposes at temperatures
14 above 150°C.
H
H
15 Figure 2-1. Chemical structure of formaldehyde.
16 At room temperature, pure formaldehyde is a colorless gas with a strong, pungent,
17 suffocating, and highly irritating odor. Formaldehyde is readily soluble in water, alcohols, ether,
18 and other polar solvents. A synopsis of its physicochemical properties is given in Table 2-1.
19
20 2.2. PRODUCTION, USES, AND SOURCES OF FORMALDEHYDE
21 Formaldehyde has been produced commercially since the early 1900s and, in recent
22 years, has been ranked in the top 25 highest volume chemicals produced in the U.S. (National
23 Toxicology Program [NTP], 2002). In 2003, 4.33 million metric tons of formaldehyde were
24 produced in the U.S. (Global Insight, 2006). In 2000, worldwide formaldehyde production was
25 estimated to be 21.5 million metric tons (International Agency for Research on Cancer [IARC],
26 2006).
This document is a draft for review purposes only and does not constitute Agency policy.
2-1 DRAFT—DO NOT CITE OR QUOTE
-------
Table 2-1. Physicochemical properties of formaldehyde
Name
International Union for Pure and
Applied Chemistry name
Synonyms
Chemical Abstracts Service Index
name
Chemical Abstracts Service
Registry Number
Formula
Molecular weight
Density
Vapor pressure
Log Kow
Henry's law constant
Conversion factors (25°C, 760 mm
Hg)
Boiling point
Melting point
Flash point
Explosive limits
Autoignition temperature
Solubility
Reactivity
Formaldehyde
Formaldehyde
Formic aldehyde
Methanal
Methyl aldehyde
Methyl ene oxide
Oxomethane
Oxymethylene
Formaldehyde
50-00-0
HCHO
30.03
Gas: 1. 067 (air =1)
Liquid: 0.815 g/mL at -20°C
3,883 mm Hg at 25°C
-0.75 to 0.35
3.4 x 10~7 atm-nrVmol at 25°C
2.2 x 10~2 Pa-m3/mol at 25°C
1 ppm =1.23 mg/m3 (v/v)
1 mg/m3 = 0.81 ppm (v/v)
-19.5°Cat760mmHg
-92°C
60°C; 83 °C, closed cup for 37%,
methanol-free aqueous solution; 50°C
closed cup for 37% aqueous solution with
15% methanol
73% upper; 7% lower by volume in air
300°C
Very soluble in water; soluble in alcohols,
ether, acetone, benzene
Reacts with alkalis, acids and oxidizers
Sources: American Conference of Governmental Industrial Hygienists (ACGIH) (2002);
International Programme on Chemical Safety (IPCS) (2002); Agency for Toxic Substances and
Disease Registry (ATSDR) (1999); Gerberich and Seaman (1994); Walker (1975).
This document is a draft for review purposes only and does not constitute Agency policy.
2-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 Formaldehyde is a chemical intermediate used in the production of some plywood
2 adhesives, abrasive materials, insulation, foundry binders, brake linings made from phenolic
3 resins, surface coatings, molding compounds, laminates, wood adhesives made from melamine
4 resins, phenolic thermosetting, resin curing agents, explosives made from
5 hexamethylenetetramine, urethanes, lubricants, alkyd resins, acrylates made from
6 trimethylolpropane, plumbing components from polyacetal resins, and controlled-release
7 fertilizers made from urea formaldehyde concentrates (IPCS, 1989). Formaldehyde is used in
8 smaller quantities for the preservation and embalming of biological specimens. It is also used as
9 a germicide, an insecticide, and a fungicide in some products. It is found (as an ingredient or
10 impurity) in some cosmetics/personal hygiene products, such as some soaps, shampoos, hair
11 preparations, deodorants, sunscreens, dry skin lotions, and mouthwashes, mascara and other eye
12 makeup, cuticle softeners, nail creams, vaginal deodorants, and shaving cream (IPCS, 2002;
13 ATSDR, 1999).
14 Formaldehyde is commonly produced as an aqueous solution called formalin, which
15 usually contains about 37% formaldehyde and 12-15% methanol. Methanol is added to
16 formalin to slow polymerization that leads eventually to precipitation as paraformaldehyde.
17 Paraformaldehyde has the formula (CH2O)n where n is 8 to 100. It is essentially a solid form of
18 formaldehyde and therefore has some of the same uses as formaldehyde (Kiernan, 2000). When
19 heated, paraformaldehyde sublimes as formaldehyde gas. This characteristic makes it useful as a
20 fumigant, disinfectant, and fungicide, such as for the decontamination of laboratories,
21 agricultural premises, and barbering equipment. Long-chain polymers (e.g., Delrin plastic) are
22 less inclined to release formaldehyde, but they have a formaldehyde odor and require additives
23 to prevent decomposition (U. S. EPA, 2008).
24 The major sources of anthropogenic emissions of formaldehyde are motor vehicle
25 exhaust, power plants, manufacturing plants that produce or use formaldehyde or substances that
26 contain formaldehyde (i.e., adhesives), petroleum refineries, coking operations, incineration,
27 wood burning, and tobacco smoke. Among these anthropogenic sources, the greatest volume
28 source of formaldehyde is automotive exhaust from engines not fitted with catalytic converters
29 (NEG, 2003). The Toxic Release Inventory (TRI) data for 2007 show total releases of
30 21.9 million pounds with about half to the air and half to underground injection (EPA TRI
31 Explorer, http://www.epa.gov/triexplorer/) (U.S. EPA, 2009a).
32 Formaldehyde is formed in the lower atmosphere by photochemical oxidation of
33 hydrocarbons or other formaldehyde precursors that are released from combustion processes
34 (ATSDR, 1999). Formaldehyde can also be formed by a variety of other natural processes such
This document is a draft for review purposes only and does not constitute Agency policy.
2-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 as decomposition of plant residues in the soil, photochemical processes in sea water and forest
2 fires (National Library of Medicine, 2001).
3 During smog episodes, indirect production of formaldehyde may be greater than direct
4 emissions (Fishbein, 1992). Grosjean et al. (1983) estimated the relative contributions of direct
5 emissions and atmospheric photochemistry to levels of formaldehyde and other carbonyls in Los
6 Angeles. They found that photochemical production predominates over direct emissions in
7 controlling formaldehyde levels in Los Angeles air. Using two models, their data were
8 translated into formaldehyde photochemical production rates of 12-161 tons per day.
9 Oxidation of methane is the dominant source of formaldehyde in regions remote from
10 hydrocarbon emissions (Staffelbach et al., 1991). Based on atmospheric measurements at a rural
11 site in Ontario, Canada and principal component analysis, Li et al. (1994) estimated that
12 formaldehyde production by atmospheric photochemical oxidation of hydrocarbons is
13 approximately 16 times that from primary emissions.
14 The input of formaldehyde into the environment is counterbalanced by its removal by
15 several pathways. Formaldehyde is removed from the air by direct photolysis and oxidation by
16 photochemically produced hydroxyl and nitrate radicals. Measured or estimated half-lives for
17 formaldehyde in the atmosphere range from 1.6 to 19 hours, depending upon estimates of radiant
18 energy, the presence and concentrations of other pollutants, and other factors (ATSDR, 1999).
19 Given the generally short daytime residence times for formaldehyde, there is limited potential for
20 long-range transport (IPCS, 2002). In cases where organic precursors are transported long
21 distances, however, secondary formation of formaldehyde may occur far from the anthropogenic
22 sources of the precursors.
23 Formaldehyde is released to water from the discharges of both treated and untreated
24 industrial wastewater from its production and from its use in the manufacture of formaldehyde-
25 containing resins (ATSDR, 1999). Formaldehyde is also a possible drinking-water disinfection
26 by-product from the use of ozone and/or hydrogen peroxide. In water, formaldehyde is rapidly
27 hydrated to form a glycol, and the equilibrium favors the glycol.
28
29 2.3. ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE
30 General population exposure to formaldehyde can occur via inhalation, ingestion and
31 dermal contact. Each of these pathways and associated media levels are discussed below.
32 Formaldehyde exposure can also occur occupationally via three main scenarios:
33
This document is a draft for review purposes only and does not constitute Agency policy.
2-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 • The production of aqueous solutions of formaldehyde (formalin) and their use in the
2 chemical industry (e.g., for the synthesis of various resins, as a preservative in medical
3 laboratories and embalming fluids, and as a disinfectant).
4 • Release from formaldehyde-based resins in which it is present as a residue and/or
5 through their hydrolysis and decomposition by heat (e.g., during the manufacture of
6 wood products, textiles, synthetic vitreous insulation products, and plastics). In general,
7 the use of phenol-formaldehyde resins results in much lower emissions of formaldehyde
8 than those of urea-based resins.
9 • The pyrolysis or combustion of organic matter (e.g., in engine exhaust gases or during
10 firefighting) (IARC, 2006).
11
12 Industries with the greatest potential for exposure include health services, business
13 services, printing and publishing, manufacture of chemicals and allied products, manufacture of
14 apparel and allied products, manufacture of paper and allied products, personal services,
15 machinery (except clerical), transport equipment, and furniture and fixtures (IARC, 1995).
16
17 2.3.1. Inhalation
18 The most current ambient air monitoring data for formaldehyde come from EPA's air
19 quality system database (EPA's AirData Web site: http://www.epa.gov/air/data/index.html)
20 (U.S. EPA, 2009b). These data have been collected from a wide variety of sources, including
21 state and local environmental agencies, but have not been collected from a statistically based
22 survey. The most recent data, for the year 2007, come from 188 monitors located in 33 states as
23 shown in Figure 2-2 (U.S. EPA, 2008). The annual means for these monitors range from
24 0.7-45.03 ug/m3 (0.56-36.31 ppb) and have an overall average of 3.44 ug/m3 (2.77 ppb). The
25 annual means are derived by EPA by averaging all available daily data from each monitor.
26 Table 2-2 shows a breakout of the data by land use category based on the annual means from
27 each monitor for 2005, 2006, and 2007. The land use is established on the basis of the most
28 prevalent land use within 0.25 miles of the monitor. The mobile category (land near major
29 highways or interstates such that it is primarily impacted by mobile sources) has the highest
30 mean levels, and agricultural lands have the lowest.
31 Under the National-Scale Air Toxics Assessment (NATA) program, EPA has conducted
32 an emissions inventory for a variety of hazardous air pollutants (HAPs), including formaldehyde
33 (U.S. EPA, 2006c). The NATA uses the emissions inventory data to model nationwide air
34 concentrations/exposures (U.S. EPA, 2006c). The results of the 1999 ambient air concentration
35 modeling for formaldehyde suggest that county median air levels range from 0 to 6.94 ug/m3
36 (0-5.59 ppb) with a national median of 0.56 ug/m3 (0.45 ppb) (see Figure 2-3). Similar results
This document is a draft for review purposes only and does not constitute Agency policy.
2-5 DRAFT—DO NOT CITE OR QUOTE
-------
Monitor Locator Map — Hazardous Air Pollutants
United States
AirData
Figure 2-2. Locations of hazardous air pollutant monitors.
Dasgupta et al. (2005) measured formaldehyde levels in 5 U.S. cities during
1999-2002. Samples were collected over approximately a one month period in
the spring or summer. Mean levels were 5.05 ppb in Nashville, TN; 7.96 ppb in
Atlanta, GA; 4.49 ppb in Houston, TX; 3.12 ppb in Philadelphia, PA; and 2.63 in
Sydney, FL.
Table 2-2. Ambient air levels by land use category
Number of data points
Mean ± standard
deviation
Minimum
Maximum
Formaldehyde exposure by category"
Agriculture
17
2.08 ±0.98
0.34
4.34
Commercial
166
3.26 ±2.76
0.20
20.61
Forest
19
2.79 ±2.17
0.40
7.33
Industrial
61
6.28 ± 14.45
0.14
74.72
Mobile"
16
6.84 ±7.28
2.02
23.39
Residential
282
2.75 ±1.71
0.17
12.35
aValues are ug/m3.
b"Mobile" is ambient air in locations primarily impacted by mobile sources.
Source: AirData for 2005, 2006, and 2007 (U.S. EPA, 2009b).
This document is a draft for review purposes only and does not constitute Agency policy.
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
1999 Estimated County Median Ambient Concentrations
Formaldehyde — United States Counties
Distribution of U.S. Ambient Concentrations
Highest I nU.S. _ _ 6.94-
95 I I 1.51
1-22 County Median Ambient Pollutant Concentration
( mlcrograms / cubic meter)
Source: US, EPA / QAQPS
-(999 f^TiA. No tonal-Scale Air Toxics Assessment
Percentlle
Figure 2-3. Modeled ambient air concentrations based on 1999 emissions.
were found for the year 2002: county concentrations ranged from 0.12 to 9.17 |ig/m3
(0.097-7.38 ppb) with median of 0.78 |ig/m3 (0.63 ppb). NATA has not provided updated
concentration maps for 2002. The 1999 map shows the highest levels in the far west and
northeastern regions of the U.S. While these modeling results can be useful, it is important to
consider their limitations. Some of the geographical differences result from differences in
methods used by states supplying the data. For example, the high levels indicated for Idaho
result from the large amount of wood burned during forest fires and the relatively high emission
factor that Idaho uses (compared with other states) to estimate formaldehyde emissions from
forest fires. A comparison of modeling results from NATA to measured values at the same
locations is presented in EPA (2006c). For 1999, it was found that formaldehyde levels were
underestimated at 76% of the sites (n = 68). One possible reason why the NATA results appear
low compared to measurements is that the modeling has not accounted for secondary formation
of formaldehyde in the atmosphere.
In general, ambient levels of formaldehyde in outdoor air are significantly lower than
those measured in the indoor air of workplaces or residences (ATSDR, 1999; IARC, 1995).
This document is a draft for review purposes only and does not constitute Agency policy.
2-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 Indoor sources of formaldehyde in air include volatilization from pressed wood products,
2 carpets, fabrics, insulation, permanent press clothing, latex paint, and paper bags, along with
3 emissions from gas burners, kerosene heaters, and cigarettes (NLM, 2001). In general, the major
4 indoor air sources of formaldehyde can be described in two ways: (1) those sources that have the
5 highest emissions when the product is new with decreasing emission over time, as with the first
6 set in the examples above; and (2) those sources that are reoccurring or frequent such as the
7 second set of examples above. Gilbert et al. (2006) studied 96 homes in Quebec City, Canada
8 and found elevated levels in homes with new wood or melamine furniture purchased within the
9 previous 12 months. A summary of indoor data is provided in Table 2-3. Results vary
10 depending on housing characteristics and date of study.
11 Salthammer et al. (2010) present a thorough review of formaldehyde sources and levels
12 found in the indoor environment. Based on an examination of international studies carried out in
13 2005 or later they conclude that the average exposure of the population to formaldehyde is 20 to
14 40 ug/m3 under normal living conditions. They used the diagram shown in Figure 2-4 to
15 summarize data they found on the range of formaldehyde air concentrations (in ppb) in different
16 environments.
17 Data on formaldehyde levels in outdoor and indoor air were collected under Canada's
18 National Air Pollution Surveillance program (IPCS, 2002; Health Canada and Environment
19 Canada, 2001). The effort included four suburban and four urban sites sampled in the period
20 1990-1998. A Monte Carlo analysis applied to the pooled data (n = 151) was used to estimate
21 the distribution of time-weighted 24-hour air exposures. This study suggested that mean levels
22 in outdoor air were 3.3 |ig/m3 (2.7 ppb) and mean levels in indoor air were 35.9 |ig/m3
23 (29.2 ppb) (Health Canada and Environment Canada, 2001). The simulation analysis also
24 suggested that general population exposures averaged 33-36 |ig/m3 (27-30 ppb).
25 Since the early to mid 1980s, manufacturing processes and construction practices have
26 been changed to reduce levels of indoor formaldehyde emissions (ATSDR, 1999). A 2008 law
27 enacted by the California Air Resource Board (CARB. 2008, Final Regulation Order: Airborne
28 Toxic Control Measure to Reduce Formaldehyde Emissions from Composite Wood Products;
29 http://www.arb.ca.gov/regact/2007/compwood07/fro-final.pdf) has limited the amount of
30 formaldehyde that can be released by specific composite wood products (i.e., hardwood
31 plywood, particle board, and medium density fiberboard) sold, supplied, or manufactured for use
32 in California. For this reason the mean indoor air levels presented by Health Canada and
33 Environment Canada (2001) (based on samples collected from 1989-1995) may overestimate
This document is a draft for review purposes only and does not constitute Agency policy.
2-8 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 2-3. Studies on residential indoor air levels of formaldehyde
(nonoccupational)
Citation
Goldetal., 1993
Hareetal., 1996
Hareetal., 1996
Gammage and
Hawthorne, 1985
Hawthorne et al.,
1986a, b
U.S. EPA, 1987
Health Canada
and Environment
Canada, 2001
Zhang et al.,
1994a, b
Gilbert et al.,
2006
Shah and Singh,
1988
Stock, 1987
Krzyzanowski et
al., 1990
No. of
samples
>1,200
131
>500
260
18
11
11
40
560
151
6
96
315
43
202
Target population/house type
Complaint homes1
Older conventional homes
Newly built homes
30 days after installing pressed wood
Homes with UFFI
Homes without UFFI
Complaint mobile homes
Newer mobile homes
Older mobile homes
Conventional homes 0-5 yr
Conventional homes 5-15 yr
Conventional homes >15 yr
Conventional homes overall
Noncomplaint, conventional, randomly
selected
Noncomplaint, mobile homes, randomly
selected
Residential (Canadian) noncomplaint
homes
Residential, carpeted, nonsmoking
homes
Residential (Canadian)
Residential and commercial
Conventional homes
Conventional homes
Mean
(ug/m3)
<60
91
42-540
60-144
30-84
120-1080
1032
300
96
48
36
72
32-109
109-744
35
66
29.5
59
84
31
Range
(ug/m3)
24-960
12-4080
12-204
0-5040
24-480
6-576
12-3480
7-148
42-89
9.6-90.0
23-89
96-216
1 The "complaint" homes are ones where the occupants have complained about formaldehyde irritant symptoms.
Note: lppb= 1.2 ug/m3.
This document is a draft for review purposes only and does not constitute Agency policy.
2-9 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
0
Remote Air
Rurj
al Air
Urban /
Jr
Normal Indoor ALT
Polk
b
ited Indoor Air
* *
Extreme Conditions
A
1 1 10 100 1000
Figure 2-4. Range of formaldehyde air concentrations (ppb) in different
environments.
Source: Salthammer et al. (2010).
5 current levels. In addition, the Canadian indoor air data may overestimate formaldehyde levels
6 in U.S. homes, because many residential homes in Canada use wood burning stoves more
7 frequently and have tighter construction (due to colder winters), leading to less dilution of indoor
8 emissions. The outdoor air levels, however, appear to have remained fairly constant over recent
9 years, and the median outdoor level from the Canadian study (2.8 |ig/m3) (2.3 ppb) is very
10 similar to the median of the U.S. monitoring data (2.83 |ig/m3) (2.3 ppb) in 1999.
11 Even though formaldehyde levels in construction materials have declined, indoor
12 inhalation concerns still persist. For example, recent studies have measured formaldehyde levels
13 in mobile homes/trailers (these terms are used interchangeably here to refer to homes with
14 wheels that are designed to be moved). ATSDR (2007) reported on air sampling in 96
15 unoccupied trailers provided by the Federal Emergency Management Agency (FEMA) used as
16 temporary housing for people displaced by Hurricane Katrina. Formaldehyde levels in closed
17 trailers averaged 1,250 ± 828 |ig/m3 (mean ± standard deviation [SD]) (1.04 ± 0.69 ppm), with a
18 range of 12-4,390 |ig/m3 (0.01-3.66 ppm). The levels decreased to an average of 468 ±
This document is a draft for review purposes only and does not constitute Agency policy.
2-10 DRAFT—DO NOT CITE OR QUOTE
-------
1 324 |ig/m3 (0.39 ± 0.27 ppm), with a range of 0.00-1,960 |ig/m3 (0.00-1.63 ppm) when the air
2 conditioning was turned on. Levels also decreased to an average of 108 ± 96 |ig/m3 (0.09 ±
3 0.08 ppm), with a range of 12-588 |ig/m3 (0.01-0.49 ppm) when the windows were opened.
4 ATSDR (2007) found an association between temperature and formaldehyde levels; higher
5 temperatures were associated with higher formaldehyde levels in trailers with the windows
6 closed. They also noted that different commercial brands of trailers yielded different
7 formaldehyde levels.
8 In December 2007 and January 2008, the Centers for Disease Control and Prevention
9 (CDC) measured formaldehyde levels in a stratified random sample of 519 FEMA-supplied
10 occupied travel trailers, park models, and mobile homes ("trailers") (CDC, 2008). At the time of
11 the study, sampled trailers were in use as temporary shelters for Louisiana and Mississippi
12 residents displaced by hurricanes Katrina and Rita. The geometric mean level of formaldehyde
13 in sampled trailers was 95 |ig/m3 (77 ppb), and the range was 3.7-730 |ig/m3 (3-590 ppb).
14
15 2.3.2. Ingestion
16 Limited U.S. data indicate that concentrations in drinking water may range up to
17 approximately 10 |ig/L in the absence of specific contributions from the formation of
18 formaldehyde by ozonation during water treatment or from leaching of formaldehyde from
19 polyacetyl plumbing fixtures (IPCS, 2002). In the absence of other data, one-half this
20 concentration (5 |ig/L) was judged to be a reasonable estimate of the average formaldehyde in
21 Canadian drinking water. Concentrations approaching 100 |ig/L were observed in a U.S. study
22 assessing the leaching of formaldehyde from domestic polyacetal plumbing fixtures, and this
23 concentration was assumed to be representative of a reasonable worst case (IPCS, 2002).
24 Formaldehyde is a natural component of a variety of foodstuffs (IARC, 1995; IPCS,
25 1989). However, foods may be contaminated with formaldehyde as a result of fumigation (e.g.,
26 grain fumigation), cooking (as a combustion product), and release from formaldehyde resin-
27 based tableware (IARC, 1995). Also, the compound has been used as a bacteriostatic agent in
28 some foods, such as cheese (IARC, 1995). There have been no systematic investigations of
29 levels of formaldehyde in a range of foodstuffs that could serve as a basis for estimation of
30 population exposure (Health Canada and Environment Canada, 2001). According to the limited
31 available data, concentrations of formaldehyde in food are highly variable. In the few studies of
32 the formaldehyde content of foods in Canada, the concentrations were within a range of
33 <0.03-14 mg/kg (Health Canada and Environment Canada, 2001). Data on formaldehyde levels
34 in food have been presented by Feron et al. (1991) and IPCS (1989) from a variety of studies,
35 yielding the following ranges of measured values:
This document is a draft for review purposes only and does not constitute Agency policy.
2-11 DRAFT—DO NOT CITE OR QUOTE
-------
1 • Fruits and vegetables: 3-60 mg/kg
2 • Meat and fish: 6-20 mg/kg
3 . Shellfish: 1-100 mg/kg
4 • Milk and milk products: 1-3.3 mg/kg
5
6 Daily intake of formaldehyde was estimated by IPCS (1989) to be in the range of
7 1.5-14 mg for an average adult. Similarly, Fishbein (1992) estimated that the intake of
8 formaldehyde from food is 1-10 mg/day but discounted this on the belief that it is not available
9 in free form. Although the bioavailability of formaldehyde from the ingestion of food is not
10 known, it is not expected to be significant (ATSDR, 1999). Using U.S. Department of
11 Agriculture (USDA) (1979) consumption rate data for various food groups, Owen et al. (1990)
12 calculated that annual consumption of dietary formaldehyde results in an intake of about
13 4,000 mg or approximately 11 mg/day.
14
15 2.3.3. Dermal Contact
16 The general population may have dermal contact with formaldehyde-containing
17 materials, such as some building products and cosmetics (see Section 2.2 for the details on these
18 products). Generally, though, dermal contact is more of a concern in occupations that involve
19 handling concentrated forms of formaldehyde, such as those occurring in embalming and
20 chemical production.
21
This document is a draft for review purposes only and does not constitute Agency policy.
2-12 DRAFT—DO NOT CITE OR QUOTE
-------
3. TOXICOKINETICS
1
2
3 This chapter presents chemical specific information on the toxicokinetics of formaldehyde
4 which helps to inform the potential for health effects from formaldehyde exposure. As a water
5 soluble and reactive gas (see Chapter 2), the chemical reactions of formaldehyde at the site of
6 first contact in biological systems is important to understanding its toxic potential. Therefore,
7 before a discussion of the absorption, distribution, and metabolism of formaldehyde (which
8 normally comprises the heart of the toxicokinetic discussion of an agent) a section is provided
9 which discusses some key issues regarding formaldehyde's reactivity. Section 3.1 provides
10 information regarding the hydration of formaldehyde in biological aqueous systems and the
11 equilibrium which exists between free formaldehyde and methylene glycol. Additional
12 information is provided on what is known of the nature of chemical reactions of free
13 formaldehyde with proteins. These discussions are provided to give context to the following
14 Sections of Chapter 3.
15 Sections 3.2 and 3.3 present the available studies which describe the absorption and
16 distribution of formaldehyde, including animal studies of radiolabeled formaldehyde. The
17 influence of formaldehyde's reactivity at the site of first contact and effects on the mucociliary
18 apparatus are presented here as well, as these effects may modify the uptake of formaldehyde.
19 Metabolism of formaldehyde is presented in Section 3.4, but the endogenous production of
20 formaldehyde from normal metabolic processed, as well as metabolism of other xenobiotics.
21 The last section of Chapter 3 present the available models which apply to the toxicokinetics of
22 formaldehyde—in this case primarily modeling of the flux of formaldehyde through tissues at
23 the sight of first contact using computational fluid dynamics models.
24
25 3.1. CHEMICAL PROPERTIES AND REACTIVITY
26 Formaldehyde (HCHO) is the smallest aldehyde (30 g/mol) and is a gas at room
27 temperature. It is highly water soluble and reactive. In water, less than 0.1% of formaldehyde
28 exists unhydrated, with the majority reported to be in the hydrated form, methylene glycol
29 (CH2(OH)2) (Priha et al., 1996). Formaldehyde reacts readily with high and low molecular
30 weight biological constituents.
31
32 3.1.1. Hydration of Formaldehyde
33 In aqueous solution formaldehyde exists in equilibrium with its hydrated form
34 methanediol (CH2OH2) (Kd = 5.5 x 10~4). The equilibrium favors methanediol at physiological
This document is a draft for review purposes only and does not constitute Agency policy.
3-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 temperature and pH (>99.9%) and is readily reversible. In biological systems, as free
2 formaldehyde is removed from aqueous solution through binding with serum proteins and
3 cellular components, the equilibrium is reestablished by dehydration of methanediol to free
4 formaldehyde. The reversible nature of this hydration reaction describes how a pool of free
5 formaldehyde may be sustained in biological systems.
6
7 3.1.2. Binding of Formaldehyde to Proteins
8 Formaldehyde is a reactive molecule that is likely to react with both low molecular
9 weight cellular components (e.g., reduced glutathione[GSH]) as well as high molecular weight
10 components. Unlike deoxyribonucleic acid (DNA), which has some additional barriers to
11 exposure (i.e., nucleus), extracellular and intracellular proteins are obvious targets for interacting
12 with formaldehyde. Formaldehyde is a well-known cross-linking agent that is used in the
13 fixation of tissues, preparation of vaccines, and study of protein-protein interactions (Metz et al.,
14 2006). However, the exact nature of the protein modifications used for these purposes is not yet
15 fully characterized (Metz et al., 2006, 2004). Figure 3-1 provides a general reaction scheme for
16 formaldehyde-mediated modifications of amino acids. Instep 1, formaldehyde reacts with
17 primary N-terminal amines to form a labile methylol adduct. This adduct can undergo
18 dehydration (step 2) to form an imine, or Schiff base (-N=CH2). Metz et al. (2004) examined
19 the types of formaldehyde-protein reactions that are likely to occur in vivo by synthesizing
20 several identical polypeptides with one varying amino acid (X) within the sequence VELXVLL
21 (V = valine, E = glutamate, L = Leucine, X = varying amino acid). Several peptides with
22 reactive amino acids did not exhibit modifications, suggesting that the peptide sequence/structure
23 affects the ability of formaldehyde to react with amino acids. Peptides that were modified
24 indicated formation of methylol adducts (see Figure 3-1, step 1) or a mixture of methylol and
25 imine adducts (see Figure 3-1, step 2).
26 Mucus is composed of water, electrolytes, polysaccharides, and about 0.5% soluble
27 proteins (Priha et al., 1996; Bogdanffy et al., 1987). Bogdanffy et al. (1987) showed that
28 although human nasal mucus can bind 70% of 100 mM formaldehyde, irreversible binding of
29 [14C]-formaldehyde to serum albumin (the major protein in mucus) was shown to be insignificant
30 after a 1-hour incubation. Irreversible binding (50% or more) did not occur until after about
31 7 hours of incubation. These data suggest that the protein content of mucus may not provide a
32 significant formaldehyde irreversible sink. Nonetheless, the solubility of formaldehyde in mucus
33 along with mucus flow and ingestion likely indicate that much of the inhaled dose is removed—
34 perhaps as much as 42% in rodents (IARC, 2005; Schlosser, 1999).
35
This document is a draft for review purposes only and does not constitute Agency policy.
3-2 DRAFT—DO NOT CITE OR QUOTE
-------
[1] Protein-NH
H H
H
Protein -N
[2] Protein-N
(Am=+30)
Protein-N=CH
(Am=+l2)
[3] Protein—N=CH +
Protein
Protein'
Protein
Figure 3-1. Formaldehyde-mediated protein modifications.
Note: Formaldehyde reacts with primary TV-terminal amines to form a methylol
adduct [1], which increases the molecular weight by 30 Da (Am). This labile
adduct can rearrange to form an amine, or Schiff base [2], that results in an
increase in MW of 12 Da. Schiff bases can react with certain amino acids to form
intra- or intermolecular methylene bridges [3]. The two amino acids depicted in
step 3 may be within the same protein or possibly from two different proteins.
Source: Metz et al. (2004).
2 In general, formaldehyde interacts with proteins. Studies carried out in cell culture media
3 containing serum and formaldehyde have shown that such mixtures are quite labile. For
4 example, during a 60-minute incubation of formaldehyde with complete cell media (i.e., with
5 fetal calf serum) at 38°C, gas chromatography-mass spectrometry (GC-MS) exhibited very
6 different peak profiles at different points during the incubation (Proctor et al., 1986). In contrast,
7 GC-MS chromatograms of cell media containing formaldehyde but no serum proteins appeared
8 relatively unchanged throughout the incubation. Compared to cell culture medium alone,
9 complete media were considered to provide a more suitable model for the hypothetical
10 interactions that formaldehyde could undergo in vivo (including perhaps blood).
11
This document is a draft for review purposes only and does not constitute Agency policy.
3-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.2. ABSORPTION
2 3.2.1. Oral
3 Oral absorption of [14C]-formaldehyde (7 mg/kg) in rats resulted in 40% elimination as
4 14C-carbon dioxide (14CO2), with 10% excretion in urine, 1% excretion in feces, and much of the
5 remaining 49% retained within the carcass, presumably due to metabolic incorporation (IARC,
6 1995; Buss etal., 1964).
7 3.2.2. Dermal
8 Jeffcoat et al. (1983) reported on the disposition of various doses of [14C]-formaldehyde
9 dermally administered to rats, guinea pigs, and monkeys. Very little (<1% of the applied dose)
10 of the radiolabel was found in the major organs excised during necropsy. As noted by the
11 authors, the disposition of formaldehyde when administered via the dermal route was markedly
12 different to that observed when the compound was administered intravenously or
13 intraperitoneally. In the latter cases, there was much evidence of metabolic activity, and
14 substantial portions of the load were expired as CO2. The difference appeared to be the result of
15 a reaction of dermally applied formaldehyde with macromolecules at or near the skin surface or
16 of its evaporation. In general, portions of the load that succeed in entering the circulation
17 probably do so bound to macromolecules or by incorporation of the radiolabel via the one-
18 carbon pool. Likewise, Bartnik et al. (1985) who applied [14C]-formaldehyde to the shaved
19 backs of rats concluded that the overwhelming majority of the formaldehyde load remained
20 sequestered in the outer layers of skin at or near the site of application. At the end of the various
21 measurements, approximately 70% of the dose was found in the treated skin, with a marked
22 localization of the remaining radioactivity in the uppermost layers. This fraction of the load was
23 considered to be permanently sequestered, most likely as a result of irreversible binding to
24 macromolecular components.
25
26 3.2.3. Inhalation
27 Studies indicate that the majority of inhaled formaldehyde is absorbed in the upper
28 respiratory tract (URT) but that the extent of the scrubbing in this region varies significantly
29 across species. In dogs, nearly 100% of nasally inhaled formaldehyde is absorbed (Egle, 1972).
30 Lower respiratory tract (LRT) studies designed to collect formaldehyde via a tube inserted into
31 the lower trachea revealed that nearly 95% of formaldehyde was absorbed during the first pass
32 through the upper respiratory tract (Egle, 1972), an effect observed with multiple ventilation
33 rates. The rat nasal passages also scrub nearly all of the inhaled formaldehyde (on average
34 -97%) (Morgan et al., 1986). In computational dosimetry modeling based on anatomically
This document is a draft for review purposes only and does not constitute Agency policy.
3-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 realistic representation of the human nasal airways from a single individual, approximately 90%
2 of inhaled formaldehyde was predicted to be absorbed in the nose at resting inspiration. As the
3 inspiratory rate increased, this fraction decreased to about 70% at light exercise and to 58% at
4 heavy exercise conditions (see Figure 1 in Kimbell et al. [2001b]). The normal human breathing
5 mode during heavy exercise is oronasal (with -54% of airflow being oral) (ICRP 66, 1994).
6 Consequently, it is estimated that during heavy exercise breathing (50 L/minute) the flux of
7 formaldehyde into tissue (or rate of mass transported per mm2 of tissue surface area) in the first
8 six to eight generations of the tracheobronchial airways is comparable to that in the nasal region
9 (Overton et al., 2001).
10 It is important to note that the computer simulations mentioned above are based on
11 anatomical representations of a single individual. Significant anatomical variations occur in
12 human nasal airways. For example, the nasal volumes of 10 adult nonsmoking subjects between
13 18 and 50 years of age in a study in the U.S. varied between 15 and 60 mL (Santiago et al.,
14 2001), and disease states can result in considerable further variation (Singh et al., 1998).
15 Species differences in kinetic factors have been argued to be the key determinants of
16 species-specific lesion distributions for formaldehyde and other reactive inhaled gases. Airway
17 geometry is an important determinant of inhaled-formaldehyde dosimetry in the respiratory tract
18 and its differences across species. These issues will be discussed in a later section on dosimetry
19 modeling.
20
21 3.2.3.1. Formaldehyde Uptake Can Be Affected by Effects at the Portal of Entry
22 Certain formaldehyde-related effects have the potential to modulate its uptake and
23 clearance. The mucociliary apparatus of the upper respiratory tract is the first line of defense
24 against airborne toxins. Comprising a thick mucus layer (epiphase), hydrophase, and a ciliated
25 epithelium, the mucociliary apparatus may entrain, neutralize, and remove particulates and
26 airborne chemicals from inspired air. As reviewed by Wolfe (1986), airborne pollutants and
27 reactive gases have been shown to decrease mucus flow rates in several animal models (Mannix
28 et al., 1983; Iravani, 1974; Carson et al., 1966; Dalhamn, 1956; Cralley, 1942). Degradation in
29 the continuity or function of this mucociliary apparatus could result in a lower clearance of
30 inhaled pollutants at the portal of entry.
31 Morgan et al. (1983) first reported defects in mucociliary function in F344 rats exposed
32 to 15 ppm formaldehyde 6 hours/day for 1-9 days. Mucostasis occurred in several regions in all
33 rats after a single 15 ppm exposure. Ciliastasis occurred with greater frequency and across more
34 regions of the nasoturbinate in subsequent days of exposure. The authors observed that
35 mucostasis preceded ciliastasis in most cases, and vigorous ciliary activity was noted in areas
This document is a draft for review purposes only and does not constitute Agency policy.
3-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 without mucus flow. Morgan et al. (1984a) also studied formaldehyde effects on the mucociliary
2 apparatus of isolated frog palates in vitro. Mucostasis was evident as mucus became stiff and
3 eventually rigid with increasing formaldehyde concentration and time of exposure. Ciliary beat
4 continued even after mucostasis, but ciliastasis ultimately occurred when exposure reached 4 and
5 9 ppm.
6 When a rodent is exposed to an irritant, its inhaled dose and pattern of deposition can be
7 profoundly affected by reflex bradypnea, a protective reflex seen in rodents but not in humans.
8 Reflex bradypnea can occur when the trigeminal nerve is exposed to a sufficient concentration of
9 an irritant, such as formaldehyde. It is manifest as markedly decreased activity or prostration,
10 reduced metabolism, hypothermia (as much as 5°C), significantly reduced respiratory rate and
11 minute volume, and altered blood and brain chemistry. Because of their small size, rodents are
12 able to rapidly lower their metabolism and body temperature and therefore their oxygen demand.
13 The consequence is that their inhaled dose of an irritating chemical is dramatically lowered.
14 Reflex bradypnea is quantified as the RD50, which is the concentration of a chemical that results
15 in a 50% decrease in respiratory rate. It can take as much as two hours for rodents to fully
16 recover from the effects of reflex bradypnea. The clinical manifestations of reflex bradypnea can
17 easily be misconstrued as toxicity. None of the studies described in this assessment took into
18 account the fact that reflex bradypnea may have confounded the results. Reflex bradypnea is
19 discussed in depth in Section 4.2.1.1.
20 Sensory irritation studies suggest that formaldehyde activates the trigeminal nerve by
21 activating nociceptors through the modification of receptor amino acids, possibly including thiol
22 groups. Cassee et al. (1996) measured sensory irritation to formaldehyde, acetaldehyde, and
23 acrolein in male Wistar rats, following a 30-minute nose-only exposure. Formaldehyde and
24 acrolein elicited similar responses, whereas acetaldehyde was far less irritating. The authors
25 suggested that the differences in sensitivity to the aldehydes might be explained by differences in
26 physicochemical properties and by regional differences in activities of detoxifying enzymes for
27 each chemical. In addition, it has been suggested that acetaldehyde might interact with sensory
28 nerves via an amino group (Steinhagen and Barrow, 1984), whereas the receptor-binding site for
29 formaldehyde and acrolein is believed to be a thiol group. Differential binding sites for sensory
30 irritants in the trigeminal nerve have been reported (Nielsen, 1991).
31 Sensory irritation effects are discussed in depth in Chapter 4 but are noted here because
32 stimulation of the trigeminal nerve by formaldehyde can result in significantly lower pulmonary
33 ventilation, and formaldehyde exposure in rodents at concentrations that approach the RDso.
34 Barrow et al. (1983) have estimated the "inhaled dose" equivalent to an exposure concentration
35 of 15 ppm in mice and rats used in the chronic formaldehyde bioassays by Kerns et al. (1983)
This document is a draft for review purposes only and does not constitute Agency policy.
3-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 and Monticello and Morgan (1994). Their results indicate that, because mice are observed to
2 decrease their minute volume by approximately 75% as compared to 45% in rats, a twofold
3 greater inhaled dose would be expected in rats versus mice. This difference may be relevant to
4 the increased incidence of squamous cell carcinoma of the nasal cavity in F344 rats as compared
5 to B6C3F1 mice. Chang et al. (1983) estimated a reduction of 25% in the minute volume of
6 F344 rats. Yokley et al. (2008) have recently published a model that accounts for physiological
7 changes in ventilation rate induced by sensory irritation in rats. Thus, the "standard" minute
8 volumes used for rats and mice need to be adjusted downward when calculating dosimetric
9 adjustment factors for extrapolation of adverse effects to humans (Thompson et al., 2008). This
10 question is further discussed in the section on modeling the dosimetry.
11 Another effect that modulates dosimetry is the dynamic tissue remodeling of nasal
12 airways that occurs as a consequence of exposure to reactive gases. For example, formaldehyde
13 dosimetry is influenced by the occurrence of squamous metaplasia, an adaptive tissue conversion
14 to squamous that occurs in nasal epithelium exposed to toxic levels of formaldehyde. The
15 metaplasia has been observed to occur in rats at exposure concentrations of 3 ppm and higher
16 (Kimbell et al., 1997b). Squamous epithelium is known to absorb considerably less
17 formaldehyde than other epithelial types (Kimbell et al., 1997b). Overall, the highest flux levels
18 of formaldehyde in the simulations of the rat nose in Kimbell et al. (200 la) are estimated in the
19 region just posterior to the nasal vestibule. A consequence of squamous metaplasia would be to
20 "push" the higher levels of formaldehyde flux toward the more distal regions of the nose
21 (Kimbell et al., 1997b). Subramaniam et al. (2008) discussed this issue further in the context of
22 uncertainties in the modeling of formaldehyde dosimetry.
23
24 3.3. DISTRIBUTION
25 3.3.1. Transport of Methylene Glycol
26 In biological systems, formaldehyde is known to exist in equilibrium with its hydrated
27 form, as methanediol (CFkOFk) (Kd = 5.5 x 10~4) at physiological temperatures and pH
28 (>99.9%) in the body and is readily reversible. When free formaldehyde is removed from
29 aqueous solution through binding with serum proteins and cellular components, the equilibrium
30 is reestablished by dehydration of methanediol to free formaldehyde. Thus, a pool of free
31 formaldehyde may be sustained in biological systems due to the reversible nature of this
32 hydration reaction.
33 There is strong and consistent evidence in biological testing systems in vitro that treating
34 cells with formaldehyde in an aqueous media results in significant cytotoxicity, cell proliferation,
35 clastogenic effects and clear evidence of mutational events (see Section 4.3). Similarly, animal
This document is a draft for review purposes only and does not constitute Agency policy.
3-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 bioassays where formaldehyde is administered in drinking water report portal of entry toxicity
2 including hyperplasia, increased cell proliferation, focal lesions and tumors (see Section 4.2.1).
3 It should be noted that URT tissues are covered by an aqueous mucous layer, through which
4 formaldehyde must pass to react the cellular components of the URT. It has been postulated that
5 formaldehyde transports through this mucous layer and the underlying tissues as methanediol
6 (Georgieva et al., 2003).
7 The dynamic equilibrium between the hydrated and unhydrated forms of formaldehyde in
8 biological systems is well understood. Since the hydration reaction favors methanediol, it is
9 expected that exogenous formaldehyde which reaches the blood will primarily exist as
10 methanediol and is subject to physiological elimination. As free, unhydrated formaldehyde
11 continues to react with serum proteins and cellular components, the blood levels of methanediol
12 are expected to reduce as it is dehydrated to maintain equilibrium. Although some attempts to
13 measure significant changes in free formaldehyde levels in blood after inhalation exposure have
14 not been successful, the half-life in blood has been measured after i.v. injection at approximately
15 2 minutes (McMartin et al., 1979). Additionally, the detection of antibodies to formaldehyde-
16 hemoglobin adducts and formaldehyde-albumin adducts in exposures workers, smokers and
17 laboratory animals exposed via inhalation provides direct evidence that formaldehyde is able to
18 react with serum albumin and hemoglobin in biological systems (Li et al., 2007; Varro et al.,
19 1997; Grammer et al., 1993; Dykewicz et al., 1991; Thrasher et al., 1990, Grammer et al., 1990).
20 These data support the hypothesis that exogenous formaldehyde may reach and transport through
21 the blood. If so, formaldehyde (or methanediol) may reach sites distal to the portal of entry.
22
23 3.3.2. Formaldehyde-GSH Conjugate as a Method of Systemic Distribution
24 Formaldehyde is primarily metabolized by alcohol dehydrogenase (ADH3) which uses
25 the formaldehyde-glutathione hemiacetal adduct as the substrate. Sanghani et al. (2000) have
26 shown that due to high circulating concentrations (50-fold) of glutathione in human blood, the
27 S-(hydroxymethyl)glutathione (HMGSH) adduct, the nonenzymatic product of formaldehyde
28 with glutathione is the major form of formaldehyde seen in vivo (Sanghani et al., 2000). It is
29 likely that the reversibly bound HMGSH may be transported to different tissues through
30 circulation, but, specific experimental evidence is lacking.
31
32 3.3.3. Levels in Blood
33 Inhalation studies in several species indicate that exposure to formaldehyde does not
34 result in elevated levels in blood. These studies were carried out over a wide range of exposure
35 concentrations and durations. Rats exposed to 14 ppm formaldehyde for 2 hours exhibited no
This document is a draft for review purposes only and does not constitute Agency policy.
3-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 increase in blood formaldehyde levels [2.25 ± 0.07 |ig/(g blood) in treated animals compared
2 with 2.24 ± 0.07 |ig/(g blood) in control animals] when measured by GC-MS using a stable
3 isotope dilution technique (Heck et al., 1985, 1982). Similarly, mean formaldehyde blood levels
4 in humans (n = 6) exposed to 1.9 ppm formaldehyde for 40 minutes in a walk-in chamber
5 (2.77 ± 0.28 |ig/g blood) were not statistically different from measurements in the same
6 population before exposure (mean of 2.61 ± 0.14 |ig/g) (Heck and Casanova-Schmitz, 1984).
7 The variability in the levels was large. At the individual level, the data showed both increase
8 and decrease in blood levels relative to pre-exposure levels, which was attributed by the authors
9 as plausibly due to temporal variations in baseline levels in humans, particularly since the
10 experiment did not control food intake prior to exposure. Studies in rhesus monkeys have
11 revealed endogenous formaldehyde levels (2.4 jig/g blood) comparable to humans and that levels
12 were also unaltered following exposure to 6 ppm formaldehyde via inhalation 6 hours/day for
13 4 weeks, measurements being taken at both 7 minutes and 45 hours post final exposure
14 (Casanova etal., 1988).
15 It is important to keep in mind that the GC-MS method is not capable of detecting
16 irreversibly bound formaldehyde; for example, formaldehyde levels detected by this method,
17 even in the anterior nasal mucosa of rats exposed to 6 ppm of formaldehyde, were not elevated
18 over control levels. Furthermore, the GC-MS method does not differentiate between free and
19 reversibly bound adducts of formaldehyde (Heck et al., 1982). Thus, measured levels represent
20 total formaldehyde concentration that includes free formaldehyde as well as reversibly bound
21 adducts. Based on the known Michaelis-Menten constant, Km, for formaldehyde dehydrogenase
22 with respect to the GSH adduct formation, Heck et al. (1982) estimated under certain
23 assumptions that free formaldehyde comprised only about 1-2% of the total formaldehyde
24 measured by their method. Furthermore, as shown by Metz et al. (2006, 2004), formaldehyde
25 reactions with primary amino and thiol groups can, in a second step, react with many other
26 amino acids to form stable methylene bridges. Presumably, such reactions would not be
27 detectable by using the methods employed by Heck et al. (1982).4 Thus, the limited
28 interpretation of GC-MS measurements of blood levels suggests that formaldehyde does not
29 appreciably reach the blood,
30
4 Additionally, note that, although Heck et al. (1982) demonstrated that formaldehyde concentration can be
accurately measured from glutathione and tetrahydrofolate adducts, similar experiments were not performed by using
protein samples or cellular extracts (i.e., in the presence of various amino acids). In addition, standard curves for
predicting formaldehyde concentration in tissues were generated in aqueous solutions rather than biological samples.
This document is a draft for review purposes only and does not constitute Agency policy.
3-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 is rapidly metabolized or interacts with macromolecules when it escapes metabolism, or is
2 otherwise undetected.
3 Results from an earlier experiment using radiolabeled formaldehyde in rats are consistent
4 with the conclusion based on the GC-MS measurements of no appreciable increase in blood
5 levels of formaldehyde. Following a 6-hour exposure of F344 rats to 15 ppm of
6 [14C]-formaldehyde (Heck et al., 1983), the concentrations of 14C in the nasal mucosa were
7 28-fold higher than those in the blood. The observed half-life of the terminal phase of the
8 radioactivity was long (55 hours); on the other hand, it is known that the half-life of free
9 formaldehyde in the rat blood is very short. Therefore, the authors concluded that the
10 radioactivity was likely due to modification of macromolecules or metabolic incorporation rather
11 than slow metabolic clearance of formaldehyde. The terminal decline of the radioactivity in the
12 packed cell fraction of the blood was much slower and observed to be consistent with
13 incorporation into erythrocytes.
14 In the same paper, Heck et al. (1983) report on the similarity in the pharmacokinetics of
15 radiolabeled formaldehyde and radiolabeled formate in the rat blood, supporting their hypothesis
16 that oxidation of formaldehyde to formate and subsequent incorporation of this compound
17 through one-carbon metabolism were major factors in the disposition of formaldehyde. Studies
18 by Gottschling et al. (1984) have also established that the main product of metabolic clearance of
19 formaldehyde is formate, which is either further metabolized to CO2 and water, incorporated into
20 the one-carbon pool, and/or eliminated in the urine as a sodium salt at about 13 mg/L urine.
21
22 3.3.4. Levels in Various Tissues
23 The radiolabeling studies indicated high levels of 14C in the rat nasal mucosa (equivalent
24 concentrations of 14C-formaldehyde in the nasal mucosa of rats naively exposed to 15 ppm
25 14C-formaldehyde were 2,148 ± 255 nmol/g compared with 76 ± 11 nmol/g in plasma). In
26 contrast, the GC-MS studies did not detect elevated formaldehyde in this region. This is not to
27 be interpreted as a discrepancy, because the radiolabeling study did not distinguish among
28 radiolabeled species and thus the measured radioactivity could potentially be free or bound
29 formaldehyde, formate, or any [14C] metabolically incorporated into macromolecules.
30 In concurrent studies, Casanova-Schmitz et al. (1984) resolved the question as to whether
31 the higher [14C] levels in the nasal mucosa were a consequence of GSH depletion and a
32 subsequent reduction in GSH-dependent clearance of formaldehyde. An important result in
33 these studies was that there was no significant difference in labeling in either the nasal mucosa or
34 in plasma between naive F344 rats and those pre-exposed to unlabeled 15 ppm formaldehyde
35 6 hours/day for the 9 previous days. These findings indicated little or no apparent effect on the
This document is a draft for review purposes only and does not constitute Agency policy.
3-10 DRAFT—DO NOT CITE OR QUOTE
-------
1 disposition of formaldehyde following short-term exposure to relatively high levels of
2 formaldehyde. In contrast, Farooqui et al. (1986) reported decreases in GSH in several tissues
3 3 hours after a sublethal IP. injection of formaldehyde but not after 6 and 9 hours. Taken
4 together, these data suggest that formaldehyde exposure does not result in long-term alterations
5 in cellular GSH levels and that repeated inhalation exposure does not alter the dosimetry to the
6 bloodstream or formaldehyde body burden.
7 Heck et al. (1983) determined the 14C concentrations in different tissues in the F344 rat
8 body by exposing rats in a head-only chamber to various concentrations (5-24 ppm) of
9 radiolabeled formaldehyde for 6 hours. (Concentrations of 14C in internal organs and tissues
10 relative to that in plasma did not appear to vary much as exposure concentrations increased;
11 therefore only averages over the concentration range were reported.) Except for the esophagus,
12 levels in the heart, spleen, lung, intestines, liver, and kidney were 1-3 times higher relative to
13 that in plasma. Labeling in the esophagus was high (fivefold relative to plasma). The authors
14 attributed this relatively higher dose to mucociliary action in the nose and trachea. The data also
15 indicate that the brain, testes, and erythrocytes appear to have about threefold lower 14C levels
16 than plasma. Pre-exposure to formaldehyde (for 9 days) did not alter the measured radioactivity
17 in the nasal mucosa or plasma. Thus, it was concluded that the single exposure findings may
18 also be qualitatively extended to chronic exposures.
19 The total radiolabel measured in the bone marrow (femur) of F344-rats exposed for
20 6 hours to 0.3-15 ppm of radiolabeled formaldehyde in the Casanova et al. (1984) experiment
21 was high (generally within a factor of 0.5 of the total labeling in the nasal respiratory mucosa).
22 Nearly half of the 14C was contained in the DNA in this tissue presumably on account of the high
23 rate of cell turnover in the bone marrow, indicating that the carbon derived from
24 14C-formaldehyde was utilized for DNA synthesis (Casanova-Schmitz et al., 1984).
25 Chang et al. (1983) described visceral labeling (via autoradiography) in rats, following
26 exposure to 15 ppm [14C]-formaldehyde 6 hours/day for 4 days. The authors attributed this
27 labeling to mucociliary clearance and grooming-related ingestion of formaldehyde.
28 In summary, following exposure to radiolabeled formaldehyde, the radioactivity was very
29 high in the nasal mucosa but was also extensively distributed to various tissues. In particular,
30 levels in the bone marrow were high. On the other hand, formaldehyde levels in the blood
31 measured by GC-MS were not significantly elevated. Thus, the authors considered it unlikely
32 that the elevated 14C in various tissues was due to free formaldehyde. Instead, these levels were
33 thought to arise from either rapid metabolic incorporation or formation of covalent adducts or
34 incorporation via carboxylation reactions of the 14CC>2 formed during metabolism.
This document is a draft for review purposes only and does not constitute Agency policy.
3-11 DRAFT—DO NOT CITE OR QUOTE
-------
1 The data presented thus far in this section illustrate that measuring the distribution of the
2 absorbed formaldehyde based on 14C-radiolabeling and GC-MS studies alone is problematic
3 because it is difficult to resolve (through these studies) whether it is free, reversibly bound,
4 irreversibly bound, formate, one-carbon pool, etc. This is of significance with regard to
5 understanding the availability of the absorbed formaldehyde. More indirect methods had to be
6 developed to further examine the disposition of formaldehyde; however, as discussed below, the
7 interpretation of these approaches may also not be straightforward.
8
9 3.3.4.1. Disposition of Formaldehyde: Differentiating Covalent Binding and Metabolic
10 Incorporation
11 The motivation in presenting this section is twofold, as follows:
12
13 1. As concluded above, subsequent studies were necessary to ascertain whether measured
14 radiolabeling in different experiments was due to formaldehyde adducts or incorporation
15 of [14C] one-carbon units of formaldehyde into macromolecules via the one-carbon pool.
16 2. DNA protein cross-links (DPXs) formed by formaldehyde (covalently bound in this case)
17 have been regarded as a surrogate dose metric for the intracellular concentration of
18 formaldehyde (Hernandez et al., 1994; Casanova et al., 1991, 1989). This is particularly
19 relevant because of the nonlinear dose response for DPX formation due to saturation of
20 enzymatic defenses at high concentrations (Casanova et al., 1991, 1989). Thus, the
21 ability to measure DPX is an important development.
22
23 An important question is whether the formaldehyde disposed in the form of DPX is
24 detected in remote tissues. A set of elegant but complex experiments involving dual isotope
25 labeling (14C and 3H) was carried out to this end by the Heck and Casanova-Schmitz and their
26 coworkers. Casanova-Schmitz et al. (1984) and Casanova-Schmitz and Heck (1983) used dual
27 isotope labeling of formaldehyde as a way to partially distinguish between formaldehyde adducts
28 formation and metabolic incorporation. In separate experiments, F344 rats were exposed to 3H-
29 and 14C-formaldehyde at different exposure concentrations (0.3-15.0 ppm), and the 3H/14C ratios
30 of different phases of DNA were measured. Only the highlights of the results and significant
31 issues are presented here. The overall conclusions from these experiments were as follows:
32
33 • Labeling in the nasal mucosa was due to both covalent binding and metabolic
34 incorporation.
35 • DPX was formed at 2 ppm and greater concentrations in the respiratory mucosa.
36 • In the bone marrow, formaldehyde did not bind covalently to bone marrow
37 macromolecules at any exposure concentration. The labeling of bone marrow
This document is a draft for review purposes only and does not constitute Agency policy.
3-12 DRAFT—DO NOT CITE OR QUOTE
-------
1 macromolecules was found to be entirely due to metabolic incorporation and not due to
2 covalent binding.
3
4 Macromolecules such as DNA and protein can be isolated from tissue homogenates by
5 extraction into three phases: an organic phase consisting of proteins, an aqueous phase consisting
6 of only double-stranded DNA, and an interfacial phase consisting of both DNA and protein.
7 Single-stranded (but not double-stranded) DNA was particularly likely to form adducts. DNA
8 from this interfacial phase can be further purified and has been shown to consist of DPXs
9 (Casanova-Schmitz and Heck, 1983). Because both [14C]-formaldehyde and [3H]-formaldehyde
10 can become incorporated into DNA and protein metabolically as well as by cross-linking, the
11 3H/14C ratio in such cross-linked material should be higher than in material that primarily
12 contains metabolically incorporated formaldehyde. Figure 3-2 shows the labeling of tissue from
13 the nasal respiratory mucosa and bone marrow (distal femur) in rats exposed to
14 [14C]-formaldehyde and [3H]-formaldehyde vapor.
15 In the nasal mucosa the interfacial phase has a significantly higher 3H/14C ratio than the
16 material in the aqueous phase. This suggests that interfacial DNA has significantly more 3H, a
17 phenomenon likely explained by additional [3H]-formaldehyde molecules present as DPXs prior
18 to extraction. The amount of interfacial DNA was found to have a clear dose response. These
19 cross-links were also judged to be due to exogenous formaldehyde. Likewise, the organic phase
20 of the nasal mucosa showed a similar increase in 3H/14C ratio at higher concentrations, a result
21 that could be attributed to various inter- and intraprotein adducts (Metz et al., 2004; Trezl et al.,
22 2003; Skrzydlewska, 1996).
23 In contrast, analysis of macromolecules at the distal femur location presents a different
24 pattern (see Figure 3-2, part B). First, the interfacial phase was not detected during extraction,
25 suggesting that there were few or no DPXs to be detected. Second, there was no increase in
26 3H/14C ratio in the organic (i.e., protein) phase as a function of dose. Therefore, it was concluded
27 that either radiolabeled formaldehyde or formate reached the distal site and was subsequently
28 incorporated into macromolecules. According to the mechanistic interpretation of these studies,
29 the quantity plotted on the ordinate in Figure 3-2 (the ratio of 3H/14C between the tissue and the
30 exposure gas) should approach unity as metabolism becomes saturated and more adduct
31 formation occurs, particularly for protein. Indeed, this is what is observed (see Figure 3-2,
32 Part A). In contrast, there is no dose effect in the femur, suggesting that the labeling at all doses
33 in that tissue may be due to metabolic incorporation and not due to the parent formaldehyde.
This document is a draft for review purposes only and does not constitute Agency policy.
3-13 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
A
ro 1 Or
- el
(7 «•£••<
I 0.1 L
C^ 1
tasal Respiratory >
, I
'"", 1
r^^~^-Jl n_
r-^TTL 1----
*-• f
lucosa
In
T
* T '*•*
B
cc
^
I
en
s
1_1
I
n
>o
Of
Oft
Of
0*
05
04
01
Of
O 1
o
Bone .Marrow
•
•
li 1^~ •
.
•*
10 >1 14
[CH20], ppm
[CH2O], ppm
Figure 3-2. 3H/14C ratios in macromolecular extracts from rat tissues
following exposure to 14C- and 3H-labeled formaldehyde (0.3, 2, 6,10,
15 ppm).
Note that the small yield of interfacial (IF) phase from bone marrow tissue
precluded further analysis; this isprima facie evidence for the lack of significant
DPXs in this tissue.
Source: Casanova-Schmitz et al. (1984a).
4
5
6
1
8
9
10
11
12
13
14
15
16
11
18
(Note: These data were originally shown in the absence of an analysis of isotope effects
on covalent binding and metabolism. Subsequent studies determined that [3H]-formaldehyde is
oxidized less rapidly than [14C]-formaldehyde and unlabeled formaldehyde. This suggests that
the 3H/14C ratio, and therefore the amount of formaldehyde covalently bound to tissue, is likely
overestimated because more [3H]-formaldehyde remains unmetabolized, i.e., free to bind [Heck
and Casanova, 1987]. The authors hypothesized that this overestimate was relatively greater at
the lower concentrations.)
Similar results were obtained in GSH-depleted rats (Casanova and Heck, 1987). Again,
these authors observed a dose-dependent increase in the 3H/14C ratio in the interfacial DNA and
organic fractions of disrupted cells of the respiratory and olfactory mucosa and no such increases
in bone marrow. Interestingly, at 10 ppm exposure (only), GSH-depleted rats exhibited a higher
3H/14C ratio in the organic phase than did normal rats. Casanova and Heck (1987) posited that
much of the covalent binding at 6 ppm and lower was due to binding to extracellular proteins,
whereas the higher 3H/14C ratio in GSH-depleted rats at 10 ppm was due to more intracellular
binding.
This document is a draft for review purposes only and does not constitute Agency policy.
3-14 DRAFT—DO NOT CITE OR QUOTE
-------
1 In their first experiment to measure DPX concentrations, Casanova-Schmidt et al. (1984)
2 and Casanova and Heck (1987) used the dual isotope method (3H/14C) mentioned above. In this
3 experiment, DPX was observed only at formaldehyde concentrations >2 ppm. Subsequently,
4 Casanova et al. (1989) developed a more sensitive method using high-performance liquid
5 chromatography (HPLC) for measuring DPX. In this method, tissue homogenates were digested
6 with a proteolytic enzyme and extracted with a phenolic solvent. DPX was detected in the nasal
7 mucosa of rats at formaldehyde concentrations as low as 0.3 ppm. This method was also used to
8 measure DPX in the nasal region, the larynx, trachea and carina, and major intrapulmonary
9 airways (airway diameters >2 mm) of rhesus monkeys exposed for 6 hours to 0.7, 2.0, and
10 6.0 ppm of formaldehyde. DPX was detected in the nose (including the nasopharynx) at all
11 concentrations and at 2.0 and 6.0 ppm in the larynx, trachea, carina, and other lower airways.
12 However, DPX was not detectable in the bone marrow of these monkeys at any concentration.
13 Overall, Heck and Casanova-Schmitz and their coworkers interpreted the results of these
14 various experiments to mean that inhaled formaldehyde could not reach distant sites in the body.
15 It may be noted in this context that Shaham et al. [1996] reported elevated DPX levels in the
16 white blood cells of laboratory workers exposed to formaldehyde. These data are further
17 reported in Chapter 4.)
18
19 3.4. METABOLISM
20 Formaldehyde is primarily metabolized by glutathione-dependent formaldehyde
21 dehydrogenase (FALDH) and aldehyde dehydrogenases (ALDHs). Numerous studies now
22 recognize FALDH as a member of the alcohol dehydrogenase (ADH) family, specifically ADH3
23 (Thompson et al., 2009; Liu et al., 2004, 2001; Hedberg et al., 2003; H00g et al., 2003; and the
24 references in each of these). The remainder of this report will refer to FALDH as ADH3.
25
26 3.4.1. In Vitro and In Vivo Characterization of Formaldehyde Metabolism
27 Formaldehyde is oxidized to formate by two metabolic pathways (see Figure 3-3). The
28 first pathway involves conversion of free formaldehyde to formate by the so-called low-Km
29 (Km = 400 jiM) mitochondrial aldehyde dehydrogenase-2 (ALDH2). The second pathway
30 involves a two-enzyme system that converts glutathione-conjugated formaldehyde
31 (S-hydroxymethylglutathione [HMGSH]) to the intermediate S-formylglutathione, which is
32 subsequently metabolized to formate and GSH by S-formylglutathione hydrolase.
This document is a draft for review purposes only and does not constitute Agency policy.
3-15 DRAFT—DO NOT CITE OR QUOTE
-------
ADH1 || GSH
Methand ^—-^^- ^ ^ *~ HMGSH
NAD+ NADH
NAD+
NADH „,, NADH
ALDH2
Formate ** ^^—S-fbrmyl GSH
S-formyl GSH hydrolase
1 Figure 3-3. Formaldehyde clearance by ALDH2 (GSH-independent) and
2 ADH3 (GSH-dependent).
The Km value for ALDH2 and free formaldehyde is about 400 jiM (Teng et al.,
2001), whereas the Km value for HMGSH and ADH3 is 6.5 |iM (Uotila and
Koivusalo, 1974a, b). The ADH-mediated reactions are reversible in the presence
of excess reduced nicotinamide adenine dinucleotide (NADH).
Source: Adapted from Teng et al. (2001).
3 Though ADH3 is rate limiting in this second pathway, the affinity of HMGSH for ADH3
4 (Km = 6.5 jiM) is about 100-fold higher than that of free formaldehyde for ALDH2. In addition
5 to the kinetic properties, this member of the ADH gene family (H00g et al., 2003, 2001; Liu et
6 al., 2001; Jornvall et al., 2000; Estonius et al., 1996) appears to be ubiquitously expressed in
7 organ tissues (Molotkov et al., 2002; Ang et al., 1996a, b), exhibits cytoplasmic and nuclear
8 localization (Fernandez et al., 2003), and is the most abundant ADH family member in the liver
9 and brain (Gaiter et al., 2003).
10 In vitro studies have examined the clearance of formaldehyde in several human and rat
11 tissues (see Table 3-1). Examination of formaldehyde metabolism in the rat nasal and olfactory
12 mucosa indicates nearly identical pharmacokinetics in the rat liver on a per mg of cell lysate
13 basis (Casanova-Schmitz et al., 1984b). Similar results have been obtained in the absence of
14 GSH, where other ALDH family members oxidize formaldehyde, albeit with significantly lower
15 affinity (i.e., higher Km). Hedberg et al. (2000) demonstrated that human buccal tissue lysate
16 kinetics are in close agreement with those reported for purified human liver ADH3 (Uotila and
17 Koivusalo, 1974a). Additionally, micro-array analysis indicates that these cells express far more
18 ADH3 and S-formylglutathione hydrolase than ALDH1 or ALDH2 (Hedberg et al., 2001a). The
This document is a draft for review purposes only and does not constitute Agency policy.
3-16 DRAFT—DO NOT CITE OR QUOTE
-------
1 results of Ovrebo et al. (2002) are not easily compared with the other studies in Table 3-1
2 because these studies were in intact cell cultures. However, it is apparent that the
3 pharmacokinetic values in these human cells are comparable to intact rat liver cells.
4 Table 3-1. Formaldehyde kinetics in human and rat tissue samples
Source
Purified human liver ADH3
Rat olfactory mucosa (+ GSH)
Rat olfactory mucosa (- GSH)
Rat respiratory mucosa (+ GSH)
Rat respiratory mucosa (- GSH)
Rat liver (+ GSH)
Human bronchial explantsa
Human bronchial epithelial21
Rat hepatocytesa
Human buccal tissue (+ GSH)
Human buccal tissue (- GSH)
Human keratinocytes
Human fibroblasts
Km(uM)
6.5
2.6 + 0.5
647 + 43
2.6 + 2.6
481 + 88
5.0 + 1.9
5,100
1,400
1,250
11+2
360 + 90
n.d.b
n.d.
Vmax (nmol/mg
protein x min)
2.77 + 0.12
1.77 + 0.12
4.39 + 0.14
0.90 + 0.24
4.07 + 0.35
2.0 + 0.3
3.3
6.1
4.2
2.9 + 0.6
1.2 + 0.7
14.5 + 1.8
17.9 + 1.4
Reference
Uotila and Koivusalo (1974a, b)
Casanova-Schmitz et al. (1984b)
Casanova-Schmitz et al. (1984b)
Casanova-Schmitz et al. (1984b)
Casanova-Schmitz et al. (1984b)
Casanova-Schmitz et al. (1984b)
Ovrebo et al. (2002)
Ovrebo et al. (2002)
Ovrebo et al. (2002)
Hedberg et al. (2000)
Hedberg et al. (2000)
Hedberg et al. (2000)
Hedberg et al. (2000)
5
6
7
9
10
11
12
13
14
aThese studies were carried out in intact cells by measuring the formation of formate. This likely explains the nearly
1,000-fold increase in apparent Km, since much of the formaldehyde was likely to be bound extracellularly. The
remaining studies used either purified enzyme or cell lysates (as indicated) and measured the formation of NADH.
bn.d. = not determined.
The data in Table 3-2 along with data indicating the ubiquity of ADH3, indicate that
many human tissues and cells, particularly in the respiratory tract, appear to exhibit significant
capacity to metabolize formaldehyde. Molecular biology techniques have demonstrated the
importance of ADH3 in formaldehyde clearance. For example, ADH-knockout studies have
shown that the median lethal dose (LDso) values for formaldehyde in wild type, ADH1 7 ,
ADH3 7 , and ADH4 7 mice strains were 0.200, 0.175, 0.135, and 0.190 g/kg, respectively
(Deltour et al., 1999). Although the statistical significance was not reported, the data indicate
that deletion of ADH3 increases the sensitivity of mice to formaldehyde.
This document is a draft for review purposes only and does not constitute Agency policy.
3-17 DRAFT—DO NOT CITE OR QUOTE
-------
Table 3-2. Allelic frequencies of ADH3 in human populations
Population, n
Chinese, 83
Spanish, 95
Swedish, 96
Allele frequencies (%)
AA_197/_19fi
22
41
47
GG-197/-196
78
59
53
A_79
100
62
67
G_79
-
38
33
T+9
-
-
1.5
^9
100
100
98.5
Source: Adapted from Hedberg et al. (200Ib).
2
3
4
5
6
7
8
9
10
11
12
The pharmacokinetics of formate are complex. Formate can undergo adenosine
triphosphate (ATP)-dependent addition to tetrahydrofolate (THF), which can carry either one or
two one-carbon groups. Formate can conjugate with THF to form N10-formyl-THF and its
isomer N5-formyl-THF, both of which can be converted to N5,N10-methenyl-THF and
subsequently to other derivatives that are ultimately incorporated into DNA and proteins via
biosynthetic pathways (see Figure 3-4).
FOMATC
^
In
C02
1 I jf
\
5-10-CH-H(,FOL*Tt
t
filT 1 OtJBP OTHP
HCv HET
Figure 3-4. Metabolism of formate.
Note: 1, formyl-THF synthetase; 2, formyl-THF dehydrogenase.
Source: Adapted from Black et al. (1985).
This document is a draft for review purposes only and does not constitute Agency policy.
3-18 DRAFT—DO NOT CITE OR QUOTE
-------
1 Elevated levels of formate in urine have been detected following inhalation of methanol
2 or formate under certain conditions (Liesivuori and Savolainen, 1987), although the
3 interpretation of this finding is unclear. There is also evidence that formate generates CO2
4 radicals and can be metabolized to CO2 via catalase and via the oxidation of N10-formyl-THF
5 (Dikalova et al., 2001, and references therein). The significance of formate in formaldehyde
6 toxicity is unclear. Black et al. (1985) reported that hepatic tetrahydrofolate levels in monkeys
7 are 60% of those in rats and that primates are far less efficient in clearing formate than are rats
8 and dogs. Studies in rats involving [14C]-formate suggest that about 80% is exhaled as 14CO2,
9 2-7% is excreted in the urine, and about 10% undergoes metabolic incorporation (Hanzlik et al.,
10 2005, and references therein). Mice deficient in formyl-THF dehydrogenase exhibit no change
11 in LD50 (via IP. dose) for methanol or in oxidation of high doses of formate (Cook et al., 2001).
12 It has been suggested that rodents efficiently clear formate via folate-dependent pathways,
13 peroxidation by catalase, or an unknown third pathway. Conversely, primates do not appear to
14 exhibit such capacity and are more sensitive to metabolic acidosis following methanol poisoning
15 (Cook etal., 2001).
16
17 3.4.2. Formaldehyde Exposure and Perturbation of Metabolic Pathways
18 The enzyme ADH3 has received renewed attention in recent years because of new
19 functions that have been attributed to it. ADH3 is central to the metabolism of formaldehyde;
20 however, exposure to formaldehyde in turn alters the activity of ADH3 (in multiple dose-
21 dependent ways), thereby leading to perturbation of critical metabolic pathways. These are
22 briefly mentioned below (refer to cited papers for details).
23
24 1. Exposure to formaldehyde increases cell replication. These proliferating epithelial and
25 inflammatory cells are rich in both the messenger ribonucleic acid (mRNA) and protein
26 of ADH3 (Nilsson et al., 2004; Hedberg et al., 2000). Studies in the rodent lung suggest
27 that increases in ADH3 in such cells dramatically alter the biology of other important
28 ADH3 substrates that are involved in protein modification and cell signaling (Que et al.,
29 2005).
30 2. ADH3 also participates in the oxidation of retinol and long-chain primary alcohols, as
31 well as the reduction of S-nitrosoglutathione (GSNO) (Staab et al., 2009; Thompson et
32 al., 2009; Hedberg et al., 2003; H00g et al., 2003; Molotkov et al., 2002; Liu et al., 2001;
33 Jornvall et al., 2000; Jensen et al., 1998). The activity of ADH3 toward some of these
34 substrates has been shown to be significantly increased in the presence of formaldehyde.
35 Staab et al. (2009) showed that (in cultured cells) GSNO can accelerate ADH3-mediated
36 formaldehyde oxidation and, likewise, that formaldehyde increases ADH3-mediated
37 GSNO reduction nearly 25-fold. The following effects may be noted with regard to the
38 relevance of such perturbations.
This document is a draft for review purposes only and does not constitute Agency policy.
3-19 DRAFT—DO NOT CITE OR QUOTE
-------
1 a. GSNO is an endogenous bronchodilator and reservoir of nitric oxide (NO)
2 activity (Jensen et al., 1998). Details on the ADH3-mediated reduction of GSNO
3 are shown in Thompson and Grafstrom (2008).
4 b. ADH3 is implicated in playing a central role in regulating bronchiole tone and
5 allergen-induced hyperresponsiveness (Gerard, 2005; Que et al., 2005).
6 c. As concluded by California Environmental Protection Agency (CalEPA) (2008),
7 "the dysregulation of NO by formaldehyde [in this manner] helps to explain the
8 variety and variability in the toxic manifestations following formaldehyde
9 inhalation."
10
11 3.4.3. Evidence for Susceptibility in Formaldehyde Metabolism
12 Teng et al. (2001) provided evidence that inhibition of ADH1, ALDH2, and ADH3 has
13 significant impact on formaldehyde toxicity. The authors speculated that deficiencies in any of
14 these enzymes would confer an increased susceptibility to formaldehyde toxicity (Teng et al.,
15 2001). Polymorphism in ALDH2 has been shown to have implications in human risk
16 assessment, specifically with regard to acetaldehyde metabolism (Ginsberg et al., 2002). It is
17 worth noting, however, that Teng et al. (2001) only demonstrated the importance of ALDH2 in
18 rat hepatocytes with formaldehyde concentrations of 2.5 mM and greater. Since this
19 concentration is fivefold greater than the 0.5 mM Km for free formaldehyde, ALDH2
20 involvement is not unexpected at such high concentrations. Teng et al. (2001) also demonstrated
21 the importance of ADH1 in driving the reverse reaction (i.e., formaldehyde to methanol) by
22 coadministration of NADH-generators. This would have the effect of prolonging the life of
23 formaldehyde by continuous recycling. This is not surprising, given that many ADH reactions
24 are reversible. However, levels of nicotinamide adenine dinucleotide (NAD+) are normally
25 much higher than NADH.
26 To date, two studies have reported polymorphisms in ADH3, using the new
27 nomenclature.5 ADH3 transcription appears to be regulated by specificity protein (Spl), with a
28 minimal promoter located at positions -34 to +61. The reported polymorphisms in ADH3
29 involve four base-pair substitutions in the promoter region and no polymorphisms in the coding
30 region (Hedberg et al., 2001b). The three polymorphisms include -197/-196 (GG-»AA), -79
31 (G^A), and +9 (C—»T). The genotype frequencies are shown in Table 3-2. Of these alleles, the
32 +9 (C—»T) polymorphism (in the putative Spl minimal promoter region) reduced transcriptional
5 Other epidemiologic studies investigating links between ADH3 and oral cancer use the older nomenclature and
thus refer to Class I ADH (i.e., ADH1) enzymes.
This document is a draft for review purposes only and does not constitute Agency policy.
3-20 DRAFT—DO NOT CITE OR QUOTE
-------
1 activity twofold in in vitro reporter gene experiments. According to Hedberg et al. (2001b), no
2 studies have demonstrated differences in ADH3 enzyme activity in humans. More recently,
3 single nucleotide polymorphisms in ADH3 have been reported to be associated with childhood
4 risk of asthma, although the functional relevance of these polymorphisms has not been published
5 (Wu et al., 2007).
6 Alterations in THF pathways may also have an impact on formaldehyde toxicity. These
7 could result from polymorphisms in various enzymes or differences in folate intake and
8 absorption. Species differences in tetrahydrofolate levels (Black et al., 1985) are thought to play
9 a role in the differential responses to methanol across species. Cook et al. (2001) speculate that
10 rats have redundant pathways for formate clearance that may be absent or less efficient in
11 primates.
12
13 3.5. ENDOGENOUS SOURCES OF FORMALDEHYDE
14 Endogenous formaldehyde is produced through normal cellular metabolism through
15 enzymatic or nonenzymatic reactions, and also as a detoxification product of xenobiotics during
16 cellular metabolism.
17
18 3.5.1.1. Normal Cellular Metabolism (Enzymatic)
19 Formaldehyde is produced during normal metabolism of methanol, amino acids (e.g.,
20 glycine, serine, and methionine), choline, dimethylglycine, and methylamine and through the
21 folate-dependent endogenous one-carbon pool, etc.
22
23 a) One of the endogenous sources for formaldehyde production is methanol, formed during
24 normal cellular metabolism. However, this fraction may also be derived through
25 consumption of fruits, vegetables and alcohol (Shelby et al., 2004; IPCS, 1997). In
26 studies conducted with healthy humans whose diet was devoid of methanol-containing or
27 methanol-generating foods (such as cereals containing aspartame, a precursor of
28 methanol) and who abstained from alcohol consumption, the background blood levels of
29 methanol range from 0.25-4.7 mg/L (reviewed in Shelby et al., 2004 [CERHR]).
30 Methanol is metabolized to formaldehyde predominantly by hepatic alcohol
31 dehydrogenase-1 (ADH1) in primates and by ADH1 and catalase (CAT) in rodents,
32 ADH1 requiring nicotinamide adenine dinucleotide (NAD+) as a cofactor.
33 b) Dimethylglycine (DMG), one of the byproducts of choline metabolism endogenously
34 present in the body, is an indirect source of endogenous formaldehyde. Two specific
35 dehydrogenases, (a) dimethylglycine dehydrogenase (DMGDH) which converts DMG to
36 sarcosine (methylglycine) and (b) sarcosine dehydrogenase (SDH) which converts
37 sarcosine to glycine, have been shown to noncovalently bind to the folate enzyme,
This document is a draft for review purposes only and does not constitute Agency policy.
3-21 DRAFT—DO NOT CITE OR QUOTE
-------
1 tetrahydrofolate (THF). Further, these dehydrogenases form "active formaldehyde" by
2 removing the 1-carbon groups from THF (Binzak et al., 2000).
3 c) Another source of endogenous formaldehyde is methylamine (MA), an intermediary
4 component of the metabolism of adrenaline, sarcosine, creatine, lecithin, and other
5 dietary sources (Yu and Zuo, 1996). The enzyme semicarbozole-sensitive amine oxidase
6 (SSAO), predominantly present in the plasma membrane of endothelial smooth muscle
7 cells and in circulating blood, converts methylamine to formaldehyde, hydrogen peroxide
8 and ammonia. The formaldehyde thus released has been shown to cause endothelial
9 injury eventually leading to atherosclerosis (Kalasz, 2003). Yu et al. (1997) have shown
10 that adrenaline, released in the body as a response to stress, is known to be deaminated by
11 the enzyme monoamine oxidase, with further conversion of methylamine to
12 formaldehyde by SSAO (Yu et al., 1997). Creatine is another precursor for methylamine
13 which is metabolized by SSAO to form formaldehyde. It has been shown that short-term,
14 high-dose dietary supplementation of creatine in healthy humans causes a significant
15 increase in urinary methylamine and formaldehyde levels (Poortmans et al., 2005).
16 d) Endogenous formaldehyde is also a constituent of the one-carbon pool, a network of
17 interrelated biochemical reactions that involve the transfer of one-carbon groups from
18 one compound to another (usually the transfer of the hydroxymethyl group of serine to
19 tetrahydrofolic acid).
20
21 Tyihak et al. (1998) have demonstrated that formaldehyde, but not the methyl radical or
22 methyl cation, is involved in the enzymatic transmethylation and demethylation reactions, and
23 suggested the presence of a formaldehyde cycle in cells for the production and removal of
24 formaldehyde utilizing the transfer through methionine -^ S-adenosylmethionine -^
25 S-adenosyl-homocysteine -^ homocysteine (Tyihak et al., 1998). However, these studies did not
26 clearly show whether the formaldehyde released in this cycle is in free or bound form.
27 Formaldehyde has been shown to be produced in normal and leukemic leukocytes from
28 N5-methyl-THF by enzymatic degradation (Thorndike and Beck, 1977). This is a two-step
29 reaction involving (1) enzymatic conversion of the methyl-THF to formaldehyde followed by (2)
30 nonenzymatic reaction of formaldehyde with an amine. Thorndike and Beck (1977) showed that
31 leukocyte (granulocyte and lymphocyte) cell extracts from normal individuals and patients with
32 chronic lymphocytic leukemia (CLL) or chronic myelocytic leukemia (CML) incubated with
33 14C-methyl-THF and saturating amounts of tryptamine produced free formaldehyde which is
34 detected as its corresponding carboline derivative formed with tryptamine. These results
35 demonstrate the activity of the enzyme N5, N10-methylene THF reductase which oxidizes
36 N5-methyltetrahydrofolate to N5, N10 methylene THF. The authors noted that the enzyme levels
37 were in the order of normal granulocytes < normal lymphocytes < granulocytes from a CML
38 individual < lymphocytes from a CLL individual (Thorndike and Beck, 1977), suggesting
This document is a draft for review purposes only and does not constitute Agency policy.
3-22 DRAFT—DO NOT CITE OR QUOTE
-------
1 increased activity of formaldehyde producing enzyme in leukemic cells compared to normal
2 leukocytes. Overall, formaldehyde might be a byproduct as well as an intermediary product in
3 several of these reactions.
4
5 3.5.1.2. Normal Metabolism (Nonenzymatic)
6 i) Formaldehyde can also be formed nonenzymatically by the spontaneous reaction of
7 methanol with hydroxyl radicals, wherein cellular hydrogen peroxide is the precursor for
8 hydroxyl radicals generated through Fenton reaction (Cederbaum and Qureshi, 1982).
9 ii) Another mechanism of nonenzymatic production of formaldehyde is through lipid
10 peroxidation of polyunsaturated fatty acids (PUFA) (Shibamoto, 2006; Slater, 1984). In
11 this mechanism, reactive oxygen species (ROS) generated during oxidative stress abstract
12 a hydrogen atom from a methylene group of polyunsaturated fatty acids (PUFA) in cell
13 membranes causing autooxidation of lipids with the eventual production of free radicals
14 (e.g., peroxy radical). It is known that a certain level of oxidative stress and lipid
15 peroxidation does occur in normal individuals, and these cellular metabolic processes are
16 likely to contribute to endogenous formaldehyde production.
17
18 3.5.1.3. Exogenous Sources of Formaldehyde Production
19 Microsomal cytochrome P450 enzymes catalyze oxidative demethylation of N-, O- and
20 S-methyl groups of xenobiotic compounds whereby formaldehyde is produced as a primary
21 product, which is subsequently incorporated into the one-carbon pool by reacting with
22 tetrahydrofolic acid or is oxidized to formate (Dahl and Hadley, 1983; Heck et al., 1982). Also,
23 some special peroxidases, such as peroxide-dependent horseradish peroxidase enzymatically
24 catalyze xenobiotics to generate formaldehyde in the body. In particular, an ethyl peroxide-
25 dependent horseradish peroxidase has been shown to act on 7V,jV-dimethylaniline and produce
26 equimolar amounts of 7V-methylaniline and formaldehyde (Kedderis and Hollenberg, 1983).
27 The tobacco-specific nitrosamine, 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone
28 (NNK), is another source of formaldehyde. It has been shown that formaldehyde is also
29 produced during the methyl hydroxylation of NNK by rat liver microsomes (Castonguay et al.,
30 1991). Also recent studies have demonstrated the formation of formaldehyde-DNA adducts in
31 NNK-treated rats using a highly sensitive liquid chromatography-electrospray ionization-tandem
32 mass spectrometry with selected reaction monitoring (Wang et al., 2007), suggesting formation
33 of formaldehyde from nitrosamines. Cigarette smoke is also a source of exogenously produced
34 methylamine which is converted to formaldehyde by SSAO (Yu, 1998).
35
This document is a draft for review purposes only and does not constitute Agency policy.
3-23 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.5.1.4. Metabolic Products of Formaldehyde Metabolism (e.g., Formic Acid)
2 Formate is converted to carbon dioxide (CO2) in rodents predominantly by a folate-
3 dependent enzyme pathway (Dikalova et al., 2001). Formate is also oxidized to CC>2 and water
4 by a minor pathway involving catalase located in rat liver peroxisomes (Waydhas et al., 1978;
5 Oshinoetal., 1973). In the folate-dependent pathway, tetrahydrofolate (THF)-mediated
6 oxidation of formate and the transfer of one-carbon compounds between different derivatives of
7 TFIF has been described.
8 Endogenous levels of formate also will be affected by dietary intake of methanol -
9 producing or methanol-containing diets since methanol is initially converted to formaldehyde
10 and eventually metabolized to formate. It has been shown in several studies in human subjects
11 who were restricted on consuming methanol producing diets, aspartame or alcohol, that the
12 endogenous blood concentrations of formate ranged from 3.8 to 19.1 mg/L (Shelby et al., 2004
13 [CERHR]). The biological half life of formic acid is 77-90 minutes (Owen et al., 1990b). The
14 levels of formate in the urine of unexposed individuals range from 11.7 to 18 mg/L (Boeniger,
15 1987). One source of formic acid intake is through diet which ranges from 0.4 to 1.2 mg per day
16 (Boeniger, 1987). The half life for plasma formate is -30 minutes or longer (Boeniger, 1987).
17
18 3.5.1.5. Levels of Endogenous Formaldehyde in Animal and Human Tissues
19 Heck et al. (1982) estimated that endogenous levels of formaldehyde (free as well as
20 bound) in rats ranged from 0.05 to 0.5 |imole/g (1.5-15 |ig/g) of wet tissue as analyzed by the
21 stable isotope dilution with GC-MS method (Heck et al., 1982). Although the levels of free
22 formaldehyde cannot be measured due to their high reactivity and short half life, they were
23 calculated by Heck et al. (1985) using an indirect method. They added a molar excess of GSH or
24 THF to the test tube containing formaldehyde in aqueous solution enabling complete binding.
25 When estimated, they observed that the amount of formaldehyde detected was equal to the total
26 amount added to the reaction suggesting that the formaldehyde measured contained both free and
27 bound forms. Further, they calculated the free formaldehyde concentration using the
28 dissociation constant of the HMGSH adduct and cellular concentration of GSH. Human
29 formaldehyde dehydrogenase has been shown to have a dissociation constant of 1.5 mM for the
30 formaldehyde-GSH hemithioacetal adduct (Uotila and Koivusalo, 1974), while the folate
31 enzyme product N5,N10-methylene-THF has a dissociation constant of 30 mM (Kallen and
32 Jencks 1966a, b). This could be evaluated using the Michaels-Menton constant (Km) of
33 formaldehyde dehydrogenase for the GSH adduct (~4 jiM at 25°C), whereby they calculated the
34 free formaldehyde level to be around 3-7 jiM or 1-2% of the total formaldehyde as measured by
35 GC-MS in rat tissues (Heck 1982).
This document is a draft for review purposes only and does not constitute Agency policy.
3-24 DRAFT—DO NOT CITE OR QUOTE
-------
1 Cascieri and Clary (1992) estimated the total body content of formaldehyde in human
2 body based on the following assumptions. For an individual with an average body wt of 70 kg
3 and with body fluids accounting for 70% of body weight, total formaldehyde content is
4 distributed in -49 kg of body mass or 49 L of body fluids, owing to the water solubility and
5 uniform distribution of formaldehyde in body fluids. It has been shown that the average blood
6 concentration (mean ± S.E.) of formaldehyde in unexposed rats and humans was 2.24 ± 0.07 and
7 2.61 ± 0.14 |ig/g of blood, respectively (Heck et al., 1985), and in unexposed rhesus monkeys it
8 was 2.42 ± 0.09 jig/g of blood (Casanova et al., 1988), overall giving an average of
9 approximately 2.5 ppm (2.5 mg/L) formaldehyde across the species. All these studies used
10 pentafluorophenyl hydrazine derived formaldehyde using GC-MS analysis (see Table 3-3).
11 Assuming these values, the body content of total formaldehyde is 122.5 mg (49 L x 2.5 mg/L) or
12 1.75 mg/kg body wt at any given time. Formaldehyde given intravenously to rhesus monkeys
13 has been shown to have a half life of-1.5 minutes in blood, wherein formaldehyde in blood was
14 measured by the dimedone method (McMartin et al., 1979). Using this information Cascieri and
15 Clary (1992) calculated that the human body generates approximately 40.83 mg/minute
16 [(122.5 mg/2 x 1.5] of formaldehyde. Biotransformation of formaldehyde to carbon dioxide in
17 the liver alone has been estimated at 22 mg/minute (Owen et al., 1990a).
18 Free formaldehyde is detected in body fluids and tissues using dimedone (Szarvas et al.,
19 1986) or 2,4-dinitrohenyhydrazine (DNPH) or pentafluorophenyl hydrazine (PFPH) derivative
20 (Heck et al., 1985) or as a fluorescent derivative (Luo et al., 2001) as trapping agent and detected
21 by analytical techniques such as thin-layer chromatography (TLC), high-performance liquid
22 chromatography (HPLC) and gas-chromatography mass spectrometry (GC-MS). Data from
23 several studies is summarized in Table 3-3. Using 14C-labeled dimedone, a chemical which
24 condenses with free formaldehyde forming a product termed "formaldemethone" enabling
25 radiometric detection, Szarvas et al. (1986) estimated the levels of endogenous formaldehyde in
26 human blood plasma to be 0.4-0.6 |ig/mL and in human urine to be 2.5-4 |ig/mL
27 (Szarvas et al., 1986).
28 Hileman (1984) reported that the endogenous levels of metabolically derived
29 formaldehyde will be in the range of 3-12 ng/g of tissue (Hileman, 1984). So for an average
30 70 kg individual, the endogenous level of metabolically derived formaldehyde would be 210 jig
31 to 840 |ig (3-12 ng/g x 0.001 |ig/ng x 1,000 g/kg x 70 kg).
32 Table 3-3. Endogenous formaldehyde levels in animal and human tissues
33 and body fluids
Tissue
Method
Detected as
Formaldehyde levels
Reference
This document is a draft for review purposes only and does not constitute Agency policy.
3-25 DRAFT—DO NOT CITE OR QUOTE
-------
Not
specified
Not
specified
Blood
Blood
Plasma
Heart
perfusate
Blood
Plasma
Urine
Not specified
GC-MS with stable
isotope dilution method
GC-MS with select ion
monitoring
GC-MS with select ion
monitoring
Reverse phase HPLC-
fluorescent detection
HPLC
GC-MS with select ion
monitoring
Radiometric method
Radiometric method
Not specified
As PFPH-
derivative
As PFPH-
derivative
As PFPH-
derivative
As product of
ampicillin
As DNPH adduct
As PFPH-
derivative
As formalde-
methone adduct
As formalde-
methone adduct
0.003-0. 012 ppm
(3-12 ng/g)
1.5-15 ppm
(0.05-0.5 (imole/g)
2.24 ± 0.07 ppm
(2.24 ± 0.07 jig/g)
2.61 ±0.14 ppm
(2.61±0.14jig/g)
1.65 ppm
(1.65 (ig/mL)
0.089-0. 126 ppm
(2.98-4.21 nmol/mL)
2.42 ± 0.09 ppm
(2.42 ± 0.09 ng/g)
0.4-0.6 ppm
(0.4-0.6 jig/mL)
2.5-4.0 ppm
(2.5-4.0 jig/mL)
Hileman 1984
Hecketal., 1982a
Hecketal., 1985
Hecketal., 1985
Luoetal, 2001
Shibamoto 2006
Casanova et al.,
1988
Szarvas et al., 1986
Szarvas et al., 1986
Values in the parenthesis, originally cited in the references, are converted to parts per million (ppm) as indicated.
PFPH, pentafluorophenyl hydrazone derivative; DNPH, dinitrophenyl hydrazine; GC-MS, gas-chromatography mass
spectrometry; HPLC, high performance liquid chromatography.
1
2 3.6. EXCRETION
3 The main product of metabolic clearance of formaldehyde is formate, which is further
4 metabolized to CC>2 and water, incorporated into the one-carbon pool, or eliminated in the urine.
5 There is also some evidence that formaldehyde is present in exhaled breath; however, it is
6 unclear whether this originates from endogenous sources, or is simply a function of ambient
7 formaldehyde dissolved in fluids lining POEs. The following sections describe first experiments
8 in laboratory species and then available data in humans. Broadly, these studies address two
9 important questions that may be of relevance for risk assessment. First, it may be of interest to
10 know what levels of formaldehyde are exhaled for comparison with inhaled levels, and whether
11 there is any relationship between external exposure and exhaled levels. Second, there are recent
12 studies that have attempted to relate genetic polymorphisms and changes in gene transcription
13 level to levels of putative urinary formaldehyde biomarkers.
This document is a draft for review purposes only and does not constitute Agency policy.
3-26 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.6.1. Formaldehyde Excretion in Rodents
2 Heck et al. (1983) determined the relative contributions of various excretion pathways in
3 F344 rats following inhalation exposure to formaldehyde. Table 3-4 indicates that the relative
4 excretion pathways were independent of exposure concentration (at least between 0.63 and
5 15 ppm). Nearly 40% of inhaled [14C]-formaldehyde appeared to be eliminated via expiration,
6 probably as CO2 (it should be recalled that nearly 100% of inhaled formaldehyde is absorbed).
7 Within 70 hours of a 6-hour exposure to formaldehyde, about 17 and 5% were eliminated in the
8 urine and feces, respectively. Nearly 40% of inhaled [14C]-formaldehyde remained in the
9 carcass, presumably due to metabolic incorporation.
10 Table 3-4. Percent distribution of airborne [14C]-formaldehyde in F344 rats
Source
Expired air
Urine
Feces
Tissues and carcass
Concentration of formaldehyde (ppm)
0.63
13.1
Distribution (%)a
39.4 ± 1.5
17.6 ± 1.2
4.2 ± 1.5
38.9 ± 1.2
41.9 ±0.8
17.3 ±0.6
5.3 ± 1.3
35.2 ±0.5
aValues are means ± standard deviations (n = 4).
Source: Heck etal. (1983).
11 Mashford and Jones (1982) examined elimination pathways of formaldehyde in rats
12 exposed by IP. injection. Urine and exhaled gases were collected from rats exposed to 4 or
13 40 mg/kg [14C]-formaldehyde. At 48 hours postinjection, 82 and 78% of the radiolabel were
14 exhaled as 14CC>2, whereas exhaled [14C]-formaldehyde was not detected. Mashford and Jones
15 (1982) also further identified the urinary metabolites. Five hours after injection of the higher
16 dose, formate was determined to comprise 80% of the urinary metabolites. The authors were
17 unable to detect cysteine derivatives observed in other studies (see below) in the urine of these
18 rats prior to or after formaldehyde exposure. The authors stated that if formaldehyde were to be
19 excreted in urine containing cysteine, then thiazolidine-4-carboxylate (TZCA) would likely be
20 produced. They speculated that species differences in urinary compounds may produce
21 formaldehyde conjugates (or artifacts).
This document is a draft for review purposes only and does not constitute Agency policy.
3-27 DRAFT—DO NOT CITE OR QUOTE
-------
1 Hemminki (1982) reacted formaldehyde and acetaldehyde with cysteine,
2 N-acetylcysteine, and GSH and found that formaldehyde reacted most rapidly with cysteine to
3 form TZCA. Similarly, acetaldehyde reacted preferentially with cysteine, albeit slower than
4 formaldehyde, to form a thiazolidine derivative. However, when each aldehyde was
5 administered IP. (10% formaldehyde, 50% acetaldehyde), thioether concentrations (nmol/mol
6 creatinine) significantly increased in the 24 and 48 hour urine of acetaldehyde-treated rats but
7 not formaldehyde-treated rats. These data suggest that formaldehyde is not appreciably excreted
8 in urine and thus cysteine conjugates are not likely to represent formaldehyde exposure.
9 Most recently, Shin et al. (2007) attempted to show that formaldehyde inhalation
10 increased urinary TZCA levels in Sprague-Dawley rats. Treated rats were exposed to 3.1 and
11 38.1 ppm formaldehyde for 6 hours/day for 2 weeks, and urine was collected for 3 days. The
12 TZCA level in four control rats was 0.07 ± 0.02 mg/L, whereas levels in the 3 and 38 ppm
13 groups were 0.18 ± 0.045 and 1.01 ± 0.36, respectively. Notably, the concentrations in the four
14 highest exposed animals (0.71, 0.70, 1.20, and 1.43 ppm) exhibited a nearly twofold range.
15 However, these comparisons are confounded if the exposures have any influence on urine
16 production and urine cysteine levels. The study does not provide any data that might allow one
17 to examine this issue.
18
19 3.6.2. Formaldehyde Excretion in Exhaled Human Breath
20 Several human and animal studies have attempted to measure the concentration of
21 formaldehyde in exhaled breath. However, study design and limitations of available analytical
22 techniques have resulted in little data which provide a basis for determining levels of
23 formaldehyde in exhaled breath either from normal metabolism (in humans), or when
24 formaldehyde is administered (animal study). The two major limitations of studies of human
25 breath include the potential for false positives for formaldehyde from the primary analytical
26 technique for breath analysis and the need for concurrent room air controls.
27 A recent study has illustrated that the use of proton transfer reaction in SIFT-MS may
28 result in false positive results for formaldehyde as the characteristic analytical product ion for
29 formaldehyde is also produced from methanol and ethanol found in exhaled breath (Spanel and
30 Smith, 2008). Proton transfer reaction mass spectrometry (PTR-MS) has been applied to
31 measure trace compounds in exhaled breath including volatile organics and specifically
32 formaldehyde. The basic method of PTR-MS is based on the transfer of protons from H?0+ to
33 gases in exhaled breath and the in-line monitoring of products where gases are tentatively
34 identified by the mass to charge ratio (m/z) where an m/z of 31 is consistent with protonated
35 formaldehyde (Hansel et al., 1995; Lindinger et al., 1998). It is important to note that reaction
This document is a draft for review purposes only and does not constitute Agency policy.
3-28 DRAFT—DO NOT CITE OR QUOTE
-------
1 products from methanol and ethanol may also produce fragments with an m/z ratio of 31 (Kusch
2 et al., 2008). Selected ion flow tube mass spectrometry (SIFT-MS) is an application of PTR-MS
3 developed for real-time analysis of trace gases in breath (Smith and Spanel, 2005; Spanel and
4 Smith, 2007). As shown in Figure 3-5 up to 1% of the mass of ethanol and methanol in exhaled
5 breath may be detected with a mass-to-charge ratio (m/z ratio) of 31—which may have been
6 reported as formaldehyde in earlier publications (Kusch et al., 2008; Spanel and Smith, 2008).
7 The authors have improved the SIFT-MS software used in exhaled breath analysis to adjust the
8 reported formaldehyde levels by accounting for the contribution of methanol and ethanol to the
9 characteristic analytical product ion for formaldehyde (m/z- = 31). No published articles were
10 available on formaldehyde in exhaled breath which adjusted for methanol and ethanol levels in
11 exhaled breath. Therefore, the available articles discussed below will be evaluated with respect
12 to the potential for ethanol or methanol to influence the reported formaldehyde levels.
13
140
14
15
16
17
18
19
20
21
22
23
5000 10000 15000 20000 Z5000 30000 35000 40000 45000 50000
ethanol + mathanol [ppb]
Figure 3-5. Detection of the characteristic analytical product ion for
formaldehyde (m/z ratio of 31) by proton transfer reaction mass
spectrometry (PTR-MS) in gas samples spiked with only methanol and
ethanol. Open circles show the reported formaldehyde without adjustment for
the methanol and ethanol present (each of which produces a small fraction of the
analytical product with an m/z ratio of 31). Closed circles represent the same
data, corrected by the SIFT-MS software to control for methanol and ethanol.
Source: Spanel and Smith (2008).
This document is a draft for review purposes only and does not constitute Agency policy.
3-29 DRAFT—DO NOT CITE OR QUOTE
-------
1 Six articles were located which reported formaldehyde levels in exhaled breath, three of
2 which provide level of methanol and ethanol in exhaled breath in the same individuals or study
3 group and are further discussed below (Wang et al., 2008; Cap et al., 2008; and Moser et al.,
4 2005). Although Wehinger et al., (2007) report a compound tentatively identified as
5 formaldehyde correlated with a diagnosis of lung caner, the PTR-MS was not controlled for any
6 contribution of ethanol and methanol, and the levels of these compounds were not provided for
7 comparison so it is not further discussed here. Turner et al. (2008) measured levels of volatile
8 compounds including formaldehyde in exhaled breath five healthy males. The subjects fasted
9 overnight, and measurements were taken before and after ingesting 75 g of glucose. The source
10 of the inhaled air was laboratory air which contained an unreported concentration of
11 formaldehyde. Formaldehyde was not detected in the exhaled breath of any subjects (5 ppb limit
12 of detection) ethanol and methanol levels were not reported.
13 In a study designed to compare volatile organics in exhaled breath of smokers and
14 nonsmokers, compounds tentatively identified as formaldehyde and methanol were not different
15 between the populations (Kushch et al., 2008). The authors acknowledge that the reported
16 formaldehyde (m/z = 31) might also represent fragments of reaction products from methanol and
17 ethanol. Reported formaldehyde levels were approximately 5% of the methanol (e.g., mean of
18 9.9 ppb versus 208 ppb respectively).
19 Wang et al. (2008) measured the concentrations volatile organics, including
20 formaldehyde, in the exhaled breath through the nose or mouth, and oral cavity during breath
21 holding of three healthy male laboratory workers. Measurements were taken in each individual
22 over a period of a month, 20 workdays. Formaldehyde levels (4-7 ppb) were lower than the
23 inspired laboratory air (9.6 ppb) (see Table 3-5). Formaldehyde in the mouth during breath
24 holding, did not differ from the exhaled air (nose or mouth). The SIFT-MS analysis did not
25 adjust for any contribution of ethanol or methanol to the tentatively identified formaldehyde
26 levels. Although only means are reported, a comparison of results in Table 3-5 does indicate that
27 1% of the reported ethanol and methanol may have contributed significantly to the reported
28 formaldehyde levels.
29 Cap et al. (2008) evaluated relationships between volatile organic compounds measured
30 in exhaled breath and exhaled breath condensate. Exhaled breath condensate consists of
31 aerosolized particles of airway lining fluid evolved from the airway wall by turbulent airflow
32 that serve as seeds for substantial water vapor condensation, which then serves to trap water
33 soluble volatile gases. This study also attempted to ascertain whether the source of each
34 compound was endogenous or exogenous. According to the published article and electronic
This document is a draft for review purposes only and does not constitute Agency policy.
3-30 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
communication with Dr. Patrik Spanel, a coauthor for this study, the limit of quantification was 3
ppb or better.
Table 3-5. Measurements of exhaled formaldehyde concentrations in the
mouth and nose, and in the oral cavity after breath holding in three healthy
male laboratory workers. The median levels are estimated as the geometric
mean with the associated standard deviation (o.)
Subject
A
B
C
Mouth
Nose
Oral cavity
Mouth
Nose
Oral cavity
Mouth
Nose
Oral cavity
Laboratory air
Methanol
(median ppb/o)
178/1.2
167/1.2
149/1.2
300/1.4
396/1.4
358/1.4
228/1.5
229/1.5
162/1.7
44 ±9
Ethanol
(median ppb/o)
236/1.6
28/1.3
412/1.4
64/1.6
27/1.4
93/1.4
153/1.5
26/1.4
163/1.4
101 ±52
Formaldehyde
(median ppb/o)
5/2.3
7/2.1
5/2.3
7/2.3
5/2.1
6/1.9
4/2.5
6/1.9
6/1.9
9.6 ± 1.5
Notes: The limit of quantification for formaldehyde was not reported.
Source: Wang et al. (2008).
8 However, the SIFT-MS protocol used in this study did not adjust for any contribution of ethanol
9 or methanol to reported formaldehyde levels. Unadjusted reported formaldehyde levels in the
10 direct exhaled breath of 34 subjects (25 to 62 years; 11 males; 2 smokers) varied from 0 to 12
11 ppb with a mean of 2 ppb and a median of 1 ppb (see Table 3-6). Measurements of
12 formaldehyde in exhaled breath condensate ranged from 0 to 12 ppb with a mean of 2 ppb and a
13 median of 0 ppb. All but one measurement was below the average ambient room air
14 concentration of 9.6 ±1.5 ppb. Although comparisons on the individual level could not be made
15 from the data as reported, the range of ethanol and methanol levels in exhaled breath indicate
16 that 1% of the reported ethanol and methanol may have contributed significantly to the reported
17 formaldehyde levels in exhaled breath (see Table 3-6). It is unclear if the reported formaldehyde
18 may represent in part inhaled formaldehyde, reduced by absorption in the upper respiratory tract,
19 or is an artifact of the reported methanol and ethanol levels.
This document is a draft for review purposes only and does not constitute Agency policy.
3-31 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 3-6 Formaldehyde, methanol and ethanol levels reported in the
exhaled breath of 34 subjects (25 to 62 years; 11 males; 2 smokers)
Chemical
Methanol
Ethanol
1% of the reported levels of both ethanol and
methanol
Formaldehyde (tentatively identified with a
m/z ratio n = 31)
Minimum
(PPb)
102
27
1.3
0
Maximum
(PPb)
2319
10262
125
12
Mean
(PPb)
297
447
7.3
2
Median
(PPb)
189
82
2.6
1
Source: Cap et al. (2008).
3 Moser et al. (2005) measured levels of 179 volatile organic compounds (VOCs) in the
4 exhaled breath of 344 individuals. This study was not designed to ascertain whether exhaled
5 formaldehyde is of endogenous origin, but rather to demonstrate that proton transfer reaction-
6 mass spectrometry can be used as a new method for rapid screening of large collectives for risk
7 factors (e.g., smoking behavior), potential disease biomarkers, and ambient air characterization.
8 The study was conducted at a health fair. The test subjects had a mean age of 61.6 years; 63%
9 were males and 14% were smokers. Samples of room air were collected and evaluated in
10 parallel with exhaled breath samples. The authors note that formaldehyde was detected in room
11 air, but did not report the levels; rather they stated that the background concentrations were
12 negligible. Of the 179 volatile organic compounds measured, data were reported for 14,
13 including formaldehyde and formic acid. The report by Moser et al. (2005) does not provide the
14 limit of detection for any of the compounds measured or details of the analytical method. Moser
15 et al. (2005) do note that significant differences in exhaled breath composition could be found
16 between smokers and nonsmokers for 32 of the 179 chemicals measured, but the 32 chemicals
17 were not named and no substantiating data were provided.
18 The formaldehyde levels in exhaled breath spanned from 1.2 to 72.7 ppb with a median
19 of 4.3 ppb and 75th percentile of 6.3 ppb (see Table 3-7) (Moser et al., 2005). The reported
20 levels of formaldehyde (m/z ratio = 31) we not adjusted for any potential contribution from
21 methanol or ethanol in exhaled breath. The levels of methanol and ethanol in exhaled breath
22 were reported by Moser et al. (2005). Although the summary statistics do not allow comparison
This document is a draft for review purposes only and does not constitute Agency policy.
3-32 DRAFT—DO NOT CITE OR QUOTE
-------
1 of individual results, it is possible that reaction fragments from methanol and ethanol may have
2 contributed to the reported formaldehyde levels (see Table 3-7).
4
5
6
Table 3-7. Apparent formaldehyde levels (ppb) in exhaled breath of
individuals attending a health fair, adjusted for methanol and ethanol levels
which contribute to the detection of the protonated species with a mass to
charge ratio of 31 reported as formaldehyde (m/z = 31)
Chemical
Methanol
Ethanol
1% of the reported levels of
both ethanol and methanol
Mass of m/z = 31 reported
as formaldehyde
Minimum
13.367
11.583
0.25
1.23
25th
percentile
106.227
23.1
1.29
3.1
Median
161.179
34.664
1.96
4.26
75th
percentile
243.185
64.24
3.07
6.33
97.5th
percentile
643.614
549.24
11.93
39.8
Maximum
1536.499
9779.768
113.16
72.7
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Source: Moser et al. (2005).
The range of reported formaldehyde is much greater in this study of the general
population (attendees at a health fair) than that observed in healthy volunteers discussed above
(Wang et al., 2008; Cap et al., 2008; Turner et al., 2008; Kushch et al., 2008). Moser et al.
(2005) do not discuss potential causes for this wide range in values, and there was no distinction
of the data by sex, age, or health. However, reported formaldehyde in exhaled breath
(unadjusted) has been correlated to lung cancer diagnosis with a median of 7.0 ppb and upper
95th CI greater than 30 ppb (Wehinger et al., 2007). Although it is unknown if these results
represent only formaldehyde, or are in part an artifact of increased ethanol and methanol in
exhaled breath, the higher levels reported by Moser et al. (2005) may reflect volatile levels in
unhealthy individuals who attended the public health fair.
Selected ion flow tube mass spectrometry (SIFT-MS), with the recent improvements by
Spanel and Smith (2008) to account for the fragments of methanol and ethanol reaction products,
have the ability to detect formaldehyde in exhaled breath. However, to date, no data has been
published which makes this adjustment for reporting formaldehyde levels. Therefore all of the
above reports of formaldehyde in exhaled breath should be carefully interpreted as the mass
reported as formaldehyde—is only tentatively identified as formaldehyde. A careful review of
the data where methanol and ethanol levels are also provided, indicate that levels of
formaldehyde (tentatively identified as m/z = 31) may reflect a significant contribution from
This document is a draft for review purposes only and does not constitute Agency policy.
3-33 DRAFT—DO NOT CITE OR QUOTE
-------
1 reaction products of methanol and ethanol. In summary, there are insufficient data at this time to
2 confidently establish a concentration of formaldehyde in exhaled breath that can be attributed to
3 endogenous sources. Additional research is needed to further clarify.
4
5 3.6.3. Formaldehyde Excretion in Human Urine
6 Gottschling et al. (1984) examined urinary formic acid in 35 veterinary students.
7 Personal monitoring badges were worn and returned after class, and urine samples were taken
8 prior to class and within 2 hours after the class. Mean exposure levels were about 100 ppb.
9 Baseline averages of urinary formic acid (as a sodium salt) were 12.47 mg/L and ranged from
10 2.43 to 28.38 mg/L among subjects. Post exposure formate levels were slightly elevated but
11 were not statistically significant. Moreover, formate levels decreased in several individuals
12 relative to pre-exposure levels. The authors concluded that variability in urinary formate may
13 mask any changes and that monitoring formate within 2 hours of exposure is not informative. It
14 is worth noting, however, that interpretation of this finding is confounded due to the fact that diet
15 was not controlled and because no markers for urinary normalization were employed (Boeniger,
16 1987).
17 Boeniger (1987) reviewed previously published data on formate in urine (some of which
18 were in German). In one occupational study, workers were exposed to an average formaldehyde
19 exposure of 1.28 mg/m3 over a 6-hour work shift. This implies an average intake of 6 mg;6
20 Boeniger reported a range of 2.5 to 13 mg. However, the original study reported that post-shift
21 formate levels were 152 mg/L, whereas the levels were only 24 mg/L 6 days later (no exposure).
22 Considering that only a small percentage of inhaled formaldehyde would be excreted in urine, it
23 is unclear how (or whether) formaldehyde exposure, with the highest total dose of 13 mg, could
24 be responsible for the observed increase.
25 In the previously described study by Shin et al. (2007), human urine samples were shown
26 to contain TZCA, although variability was not reported. A subsequent study reported that urine
27 TZCA levels were higher in individuals living in newer apartments (0.18 ± 0.121 mg/g
28 creatinine) as compared to older apartments (0.097 ± 0.040 mg/g creatinine) (Li et al., 2007).7
29 The authors cited this as evidence that TZCA is a urinary marker for formaldehyde exposure,
30 even though TZCA levels were not correlated to measured (or estimated) formaldehyde
31 exposures. The individuals also differed significantly in age (21.5 vs. 28.6,/? = 0.053) and
' 1.28 mg/m3/l,000 L/m3 x 13.8 L/minute x 60 minutes/hour x 6 hours.
This study is described in greater detail in Chapter 5.
This document is a draft for review purposes only and does not constitute Agency policy.
3-34 DRAFT—DO NOT CITE OR QUOTE
-------
1 differed in smoking percentage (10 vs. 27%). Clearly these two studies do not establish a
2 relationship between human formaldehyde exposure and urine TZCA levels.
3
4 3.7. MODELING THE TOXICOKINETICS OF FORMALDEHYDE AND DPX
5 3.7.1. Motivation
6 Airway geometry is expected to be an important determinant of inhaled formaldehyde
7 dosimetry in the respiratory tract and its differences across species. The uptake of formaldehyde
8 in the upper respiratory tract is highly nonhomogeneous and spatially localized and exhibits
9 strong species differences. Species differences in kinetic factors have been argued to be the key
10 determinants of species-specific lesion distributions for formaldehyde and other reactive inhaled
11 gases. Section 3.7.2 details the benefits to the quantitative risk assessment of modeling these
12 dosimetric differences in the upper respiratory tract. While frank effects were seen only in the
13 upper respiratory tract in rodents, mild lesions were also present in the major bronchiolar region
14 of the rhesus monkey. Therefore, with regard to extrapolation of cancer risk from animal
15 bioassays to humans, it appears that the upper and lower human respiratory tract should both be
16 considered potentially at risk of developing formaldehyde-induced squamous cell carcinoma.
17 Therefore, formaldehyde dose to the lower human respiratory tract also needs to be quantified in
18 order to develop a dose-response relationship that considers the entire respiratory tract.
19 This assessment uses internal dose metrics computed by using fluid dynamic models to
20 compute regional formaldehyde uptake in the F344 rat and human nasal passages and in the
21 human lower respiratory tract. The assessment also uses estimates of DPX levels in the nasal
22 lining predicted by physiologically-based pharmacokinetic models which use the fluid dynamic
23 model derived estimates of formaldehyde flux to the tissue as input. These computational
24 models enable the derivation of more accurate human equivalent concentrations from the animal
25 bioassays than would be obtained by averaging over the respiratory surface area. The following
26 sections provide the motivation for these calculations, and discuss the strengths and uncertainties
27 associated with the data and the models and their relevance to the hypothesized mode of action
28 are discussed in some length.
29
30 3.7.2. Species Differences in Anatomy: Consequences for Gas Transport and Risk
31 As discussed earlier, formaldehyde is highly reactive and water soluble (categorized as a
32 category 1 gas), thus its absorption in the mucus layer and tissue lining of the upper respiratory
33 tract is known to be significant. The regional inhaled dose of formaldehyde to the respiratory
34 tract of a given species depends on the amount of formaldehyde delivered by inhaled air, the
35 absorption characteristics of the nasal lining, and reactions in the tissue. The amount delivered
This document is a draft for review purposes only and does not constitute Agency policy.
3-35 DRAFT—DO NOT CITE OR QUOTE
-------
1 by inhaled air is a function of the major airflow patterns, air-phase diffusion, and absorption at
2 the airway-epithelial tissue interface. The dose of formaldehyde to the epithelial tissue, which is
3 different from the amount delivered, depends on the amount absorbed at the airway-tissue
4 interface, water solubility, mucus-to-tissue phase diffusion, and chemical reactions, such as
5 hydrolysis, protein binding, and metabolism. It has been argued strongly that species differences
6 in these kinetic factors are determinants of species-specific lesion distributions for formaldehyde
7 and other inhaled gases (Moulin et al., 2002; Bogdanffy et al., 1999; Ibanes et al., 1996;
8 Monticello et al., 1996; Monticello and Morgan, 1994; Morgan et al., 1991).
9 Because of the convoluted nature of the airways in the upper respiratory tract, the
10 absorption of such gases in the upper respiratory tract is highly nonhomogeneous. There are
11 large differences across species in the anatomy of the upper respiratory tract (see Figure 3-6) and
12 in airflow patterns (see Figure 3-7). Therefore, as shown in the simulations in Figure 3-8, it may
13 be expected that the uptake patterns, and thus risk due to inhaled formaldehyde, will also show
14 strong species dependence. Morgan et al. (1991) concluded that airflow-driven dosimetry plays
15 a critical role in determining the site specificity of various formaldehyde-induced responses,
16 including tumors, in the nose of the F344 rat. The convoluted geometry of the airway passages
17 in the upper respiratory tract, as seen from the cross sections of the nose in Figure 3-6, renders an
18 idealized representation of fluid flow and uptake profiles almost impossible. For these reasons,
19 Kimbell et al. (1998, 1993), Kepler et al. (1998), and Subramaniam et al. (1998) developed
20 anatomically realistic finite-element representations of the noses of humans, F344 rats, and
21 rhesus monkeys. These representations were subsequently used in physical and computational
22 models (see Figure 3-6). This assessment utilizes dosimetry derived from these representations.
23 An accurate calculation of species differences in formaldehyde dosimetry in the upper
24 respiratory tract is important to the extrapolation problem for another reason. The upper
25 respiratory tract in rats is an extremely efficient scrubber of reactive gases (97% uptake)
26 (Morgan et al., 1986), thereby protecting the lower respiratory tract from gaseous penetration.
27 On the other hand, there is considerably more fractional penetration of formaldehyde into the
28 lower respiratory tract of the rhesus monkey than in the rat (see Figure 3-8). Therefore, an
29 accurate determination of scrubbing in the upper respiratory tract is important to delineate
30 species differences in dosimetry in both the upper and lower respiratory tract. Thus, in the case
31 of the rhesus monkey, the model by Kepler et al. (1998) included the trachea. It is important to
32 note that the models mentioned above represent nasal passages reconstructed from a single
33 individual from each species (Kimbell et al., 2001a, b; Conolly et al., 2000; CUT, 1999;
34 Subramaniam et al., 1998). This is discussed later in the context of intraspecies variability.
This document is a draft for review purposes only and does not constitute Agency policy.
3-36 DRAFT—DO NOT CITE OR QUOTE
-------
XH
Human
2
3
Figure 3-6. Reconstructed nasal passages of F344 rat, rhesus monkey, and
human.
Note: Nostril is to the right, and the nasopharynx is to the left. Right side shows
the finite element mesh. Left-hand side shows tracings of airways obtained from
cross sections of fixed heads (F344 rat and rhesus monkey) and magnetic
resonance image sectional scans (humans). Aligned cross sections were
connected to form a three-dimensional reconstruction and finite-element
computational mesh. Source: Adapted from Kimbell et al. (200la). Additional
images provided courtesy of Dr. J.S. Kimbell, CUT Hamner Institutes.
This document is a draft for review purposes only and does not constitute Agency policy.
3-37 DRAFT—DO NOT CITE OR QUOTE
-------
(A) F344Rat
CFD-Mociel simulated inspiratory airflow
Streams observed in experimental water-dye studies
(B) Rhesus Monkey
CFD-Model simulated inspiratory airflow
Streams observed in experimental water-dye studies
(C) Human
CFD-Model simulated inspiratory airflow
Black bars = Axial velocities measured in hollow molds
White bars = CFD-simulated axial velocities
ii till Ilii
£
Urn
• I • I •
• I I I I I •
• I • • •
Normalized airway diameter
1
2
Figure 3-7. Illustration of interspecies differences in airflow and verification
of CFD simulations with water-dye studies.
Note: Panels A and B show the simulated airflow pattern versus water-dye
streams observed experimentally in casts of the nasal passages of rats and
monkeys, respectively. Panel C shows the simulated inspiration airflow pattern,
and the histogram depicts the simulated axial velocities (white bars) vs.
experimental measurements made in hollow molds of the human nasal passages.
Dye stream plots were compiled for the rat and monkey over the physiological
range of inspiration flow rates. Modeled flow rates in humans were 15 L/minute.
Source: Adapted from Kimbell et al. (2001a).
This document is a draft for review purposes only and does not constitute Agency policy.
3-38 DRAFT—DO NOT CITE OR QUOTE
-------
F344 Rat
Rhesus Monkey
Key
pmol/(mm2-hr-ppm)
2000
1500
1000
500
0
1
2
Figure 3-8. Lateral view of nasal wall mass flux of inhaled formaldehyde
simulated in the F344 rat, rhesus monkey, and human.
Note: Nostrils are to the right. Simulations were exercised in each species at
steady-state inspiration flow rates of 0.576 L/minute in the rat, 4.8 L/minute in the
monkey, and 15 L/minute in the human. Flux was contoured over the range from
0-2,000 pmol/(mm2-hour-ppm) in each species.
Source: Kimbell et al. (200 la).
3 The highly localized nature of uptake patterns shown in Figure 3-8 means that averaging
4 uptake over the entire nasal surface area would dilute the regional dose over areas where
5 response was observed and that an extrapolation based on such averaging would clearly not be
6 accurate.
7 Another factor to consider in the extrapolation is that monkeys and humans are oronasal
8 breathers while rats are obligate nose-only breathers. Thus, for humans and monkeys, oronasal
9 or oral breathing implies a significantly higher uptake in the lower respiratory tract. It is known
This document is a draft for review purposes only and does not constitute Agency policy.
3-39 DRAFT— DO NOT CITE OR QUOTE
-------
1 that a significant fraction of the human population breathes normally through the mouth.
2 Finally, activity profiles are also determinants of extraction efficiency (see Figure 3-9) and of
3 breathing route (Niinimaa et al., 1981). Given the fact that formaldehyde-induced lesions were
4 observed as far down the respiratory tract as the first bifurcation of the lungs in exposed
5 monkeys, the entire human respiratory tract should be considered when extrapolating data from
6 rats. Thus, for the human, Overton et al. (2001) attached an idealized single-path model of the
7 lower respiratory tract to a model of the upper respiratory tract.
2082 pmol/(mm2-hr-ppm)
15L/min
37 L/min
9
10
11
12
13
14
15
Figure 3-9. CFD simulations of formaldehyde flux to human nasal lining at
different inspiratory flow rates.
Note: Right lateral view. Uptake is shown for the nonsquamous portion of the
epithelium. The front portion of the nose (vestibule) is lined with keratinized
squamous epithelium and is expected to absorb relatively much less
formaldehyde.
Source: Kimbell et al. (2001b).
This document is a draft for review purposes only and does not constitute Agency policy.
3-40 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.7.3. Modeling Formaldehyde Uptake in Nasal Passages
2 Computational models for air flow and formaldehyde uptake in the F344 rat, rhesus
3 monkey, human nose, and human lung were developed by several scientists (Kimbell et al.,
4 1998, 1993; Kepler et al., 1998; Subramaniam et al., 1998; Kimbell et al., 2001a, b). The F344
5 strain of the rat was chosen since it was assumed to be anatomically representative of its species
6 and because it is widely used experimentally, most notably in bioassays sponsored by the
7 National Toxicology Program. The approximate locations of squamous, mucus-coated, and
8 nonmucus epithelial cells were mapped onto the reconstructed nasal geometry of the computer
9 models. Taken together, these regions of nonmucus and mucus-coated cells comprise the entire
10 surface area of the nasal passages (see original papers and CUT [1999] for further details on
11 reconstruction and morphometry). Types of nasal epithelium overlaid onto the geometry of the
12 models were assumed to be similar in characteristics across all three species (rat, monkey, and
13 human) except for thickness, surface area, and location. Species-specific mucosal thickness,
14 surface area, and location were estimated from the literature or by direct measurements (Conolly
15 et al., 2000; CUT, 1999). The nasal passages of all three species were assumed to have a
16 continuous mucus coating over all surfaces except specific areas in the nasal vestibule. As
17 discussed at the beginning of this chapter, formaldehyde hydrolyzes in water and reacts readily
18 with a number of components of nasal mucus. Absorption rates of inhaled formaldehyde by the
19 nasal lining were therefore assumed to depend on where the epithelial lining is coated by mucus
20 and where it is not.
21 To calculate an airflow rate that would be comparable among species, the amount of
22 inspired air (tidal volume, VT) was divided by the estimated time involved in inhalation (half the
23 time a breath takes, or (1/2)(1/[breathing frequency, f]). Thus, an inspiratory flow rate was
24 calculated to be 2Vjf, or twice the minute volume. Predicted flux values represent an average of
25 one nasal cycle. Minute volumes were allometrically scaled to 0.288 L/minute for a 315 g rat
26 from data given by Mauderly (1986). Simulations were therefore carried out at 0.576 L/minute
27 for the rat.
28 The fluid dynamics modeling in the respiratory tract comprises two steps: modeling the
29 airflow through the lumen (solution of Navier-Stokes equations) and modeling formaldehyde
30 uptake by the respiratory tract lining (solution of convective-diffusion equations for a given
31 airflow field). Details of these simulations, including boundary conditions for air flow and mass
32 transfer, are provided in Kimbell et al. (2001a, b; 1998, 1993) and Subramaniam et al. (1998).
33 Formaldehyde absorption at the airway-to-epithelial tissue interface was assumed to be
34 proportional to the air-phase formaldehyde concentration adjacent to the nasal lining layer in
This document is a draft for review purposes only and does not constitute Agency policy.
3-41 DRAFT—DO NOT CITE OR QUOTE
-------
1 monkeys and humans (see the original paper [Kimbell et al., 200la, b] for a more detailed
2 elaboration of the calculations for these coefficients).
3 Because formaldehyde is highly water soluble and reactive, Kimbell at al. (2001a)
4 assumed that absorption occurred only during inspiration. Thus, for each breath, flux into nasal
5 passage walls (rate of mass transport in the direction perpendicular to the nasal wall per mm2 of
6 the wall surface) was assumed to be zero during exhalation, with no backpressure to uptake built
7 up in the tissues. Overton et al. (2001) estimated the error due to this assumption to be small,
8 roughly an underestimate of 3% in comparison to cyclic breathing. Also, this assumption is the
9 same as that used in default methods for reference concentration determination and has been
10 used in other PBPK model applications to describe nasal uptake (Andersen and Jarabek, 2001).
11
12 3.7.3.1. Flux Bins
13 A novel contribution of the CUT biologically motivated dose-response model is that cell
14 division rates and DPX concentrations are driven by the local concentration of formaldehyde.
15 These were determined by partitioning the nasal surface by flux, resulting in 20 "flux bins."
16 Each bin was comprised of elements (not necessarily contiguous) of the nasal surface that
17 receive a particular interval of formaldehyde flux per ppm of exposure concentration (Kimbell et
18 al., 2001a, b). The spatial coordinates of elements comprising a particular flux bin were fixed
19 for all exposure concentrations, with formaldehyde flux in a bin scaling linearly with exposure
20 concentration (ppm). Thus, formaldehyde flux was expressed as pmol/(mm2-hour-ppm).
21
22 3.7.3.2. Flux Estimates
23 Formaldehyde flux was estimated for the rat, monkey, and human over the entire nasal
24 surface and over the portion of the nasal surface that was lined by nonsquamous epithelium.
25 Formaldehyde flux was also estimated for the rat and monkey over the areas where cell
26 proliferation measurements were made (Monticello et al., 1991, 1989) and over the anterior
27 portion of the human nasal passages that is lined by nonsquamous epithelium. Figure 3-8 shows
28 the mass flux of inhaled formaldehyde to the lateral wall of nasal passages in the F344 rat, rhesus
29 monkey, and human (Kimbell et al., 2001a, b).
30 Maximum flux estimates for the entire upper respiratory tract were located in the mucus-
31 coated squamous epithelium on the dorsal aspect of the dorsal medial meatus near the boundary
32 between nonmucus and mucus-coated squamous epithelium in the rat, at the anterior or rostral
33 margin of the middle turbinate in the monkey, and in the nonsquamous epithelium on the
34 proximal portion of the mid-septum near the boundary between squamous and nonsquamous
This document is a draft for review purposes only and does not constitute Agency policy.
3-42 DRAFT—DO NOT CITE OR QUOTE
-------
1 epithelium in the human (see Kimbell et al. [2001a, b] for tabulations of comparative estimates
2 of formaldehyde flux across the species).
3 The rat-to-monkey ratio of the highest site-specific fluxes in the two species was 0.98. In
4 the rat, the incidence of formaldehyde-induced squamous cell carcinomas in chronically exposed
5 animals was high in the anterior lateral meatus (Monticello et al., 1996). Flux predicted per ppm
6 in this site and flux predicted near the anterior or proximal aspect of the inferior turbinate and
7 adjacent lateral walls and septum in the human were similar, with a rat-to-human ratio of 0.84.
8
9 3.7.3.3. Mass Balance Errors
10 Overall uptake of formaldehyde was calculated as 100% x (mass entering nostril - mass
11 exiting outlet)/(mass entering nostril). Mass balance errors for air, 100% x (mass of air entering
12 nostril - mass exiting outlet)/(mass entering nostril), and inhaled formaldehyde, 100% x (mass
13 entering nostril - mass absorbed by airway walls - mass exiting outlet)/(mass entering nostril),
14 were calculated. Mass balance errors associated with simulated formaldehyde uptake from air
15 into tissue ranged from less than 14% for the rat, monkey, and human at 7.4 and 15 L/minute to
16 approximately 27% at the highest inspiratory flow rates of 31.8 and 37 L/minute (Kimbell et al.,
17 2001b). Kimbell et al. (2001b) corrected the simulation results for these errors by evenly
18 distributing the lost mass over the entire nasal surface.
19
20 3.7.4. Modeling Formaldehyde Uptake in the Lower Respiratory Tract
21 Lesions were observed in the lower respiratory tract of rhesus monkeys exposed to 6 ppm
22 formaldehyde. Therefore it is appropriate to consider the human lower respiratory tract as
23 potentially at risk for formaldehyde-induced cancer. Accordingly, fluid flow and formaldehyde
24 uptake in the lower respiratory tract were also modeled for the human in the CUT approach by
25 using dosimetry estimates for the human lower respiratory tract.
26 The single-path idealization of the human lung anatomy captures the geometrical
27 characteristics of the airways for a given lung depth, and of airflow through these airways, in an
28 average, homogeneous sense. For particulates, this has provided a reasonable representation of
29 the average deposition in a given generation of the lung airways for a normal human population.
30 The one-dimensional model by Weibel (1963) is generally considered adequate unless the fluid
31 dynamics at airway bifurcations need to be explicitly modeled, and such an idealization of the
32 lung geometry has been successfully used in various models for the dosimetry of ozone and
33 particulate and fibrous matter. Most likely, the lung geometries of the susceptible population,
34 such as those with chronic obstructive pulmonary disease, would depart significantly from the
35 geometry described in Weibel (1963). Unlike the accurate representation of the nasal anatomy
This document is a draft for review purposes only and does not constitute Agency policy.
3-43 DRAFT—DO NOT CITE OR QUOTE
-------
1 used in the CFD modeling, the lung geometry is idealized in the CUT approach as a typical path
2 Weibel geometry. The single-path model used to calculate formaldehyde uptake in the human
3 respiratory tract (Overton et al., 2001; CUT, 1999) applied a one-dimensional equation of mass
4 transport to each generation of an adult human symmetric, bifurcating Weibel-type respiratory
5 tract anatomical model, augmented by an upper respiratory tract. The detailed CFD modeling of
6 the upper respiratory tract was made consistent with the upper respiratory tract in the single-path
7 model by requiring that the one-dimensional version of the nasal passages have the same
8 inspiratory air-flow rate and uptake during inspiration as the CFD simulations for four daily
9 human activity levels. The reader is referred to Overton et al. (2001) for further details of the
10 simulations. Results most relevant to this assessment are shown in Figure 3-10.
11
IdUO
7.5 Umin (Nasal E
9.0 L'min
25.0 Unin (NasaJ E
— 50.0 L/min (Orona
-200
(J 5 10 15
MODEL GENERATION
12
13
14
15
16
17
Figure 3-10. Single-path model simulations of surface flux per ppm of
formaldehyde exposure concentration in an adult male human.
Source: Overton et al. (2001).
The primary predictions of the model, as shown in Figure 3-10, were that more than 95%
of the inhaled formaldehyde would be retained and formaldehyde flux in the lower respiratory
This document is a draft for review purposes only and does not constitute Agency policy.
3-44 DRAFT—DO NOT CITE OR QUOTE
-------
1 tract would increase for several lung airway generations from that in the posterior-most segment
2 of the nose and then decrease rapidly, resulting in almost zero flux to the alveolar sacs.
3 Overton et al. (2001) modeled uptake at higher inspiratory rates, including those at
4 50 L/minute of minute volume (well beyond levels where the oronasal switch occurs in the
5 normal nasal breathing population). At these rates Figure 3-8 indicates that formaldehyde flux in
6 the mouth cavity is comparable (but a bit less) to that occurring in the nasal passages. Overton et
7 al. (2001) did not model uptake in the oral cavity at minute volumes less than 50 L/minute. This
8 would be of interest because mouth breathers form a large segment of the population.
9 Furthermore, at concentrations of formaldehyde where either odor or sensory irritation becomes
10 a significant factor, humans are likely to switch to mouth breathing even at resting inspiration.
11 At a minute volume of 50 L/minute, Overton et al. (2001) assumed, citing Niinimaa et al. (1981),
12 that 0.55 of the inspired fraction is through the mouth. Therefore, based on the results in
13 Figure 3-8, it is not unreasonable to assume that for mouth breathing conditions at resting or
14 light exercise inspiratory rates, average flux across the human mouth lining would be
15 comparable to the average flux across the nasal lining computed in Kimbell et al. (2001a, b).
16
17 3.7.5. Uncertainties in Formaldehyde Dosimetry Modeling
18 3.7.5.1. Verification of Predicted Flow Profiles
19 The simulated streamlines of steady-state inspiration airflow predicted by the CFD model
20 agreed reasonably well with experimentally observed patterns of water-dye streams made in
21 casts of the nasal passages for the rat and monkey as shown in panels A and B in Figure 3-7.
22 The airflow velocity predicted by CFD model simulations of the human also agreed well with
23 measurements taken in hollow molds of the human nasal passages (panel C, Figure 3-8) (Kepler
24 et al., 1998; Subramaniam et al., 1998; Kimbell et al., 1997a, 1998, 1993). However, the
25 accuracy and relevance of these comparisons are limited. The profiles were verified by video
26 analysis of dye streak lines in the molds of rats and rhesus monkeys, although this method is
27 reasonable for only the major airflow streams.
28 Plots of pressure drop vs. volumetric airflow rate predicted by the CFD simulations
29 compared well with measurements made in rats in vivo (Gerde et al., 1991) and in acrylic casts
30 of the rat nasal airways (Cheng et al., 1990) as shown in Figure 3-11. This latter comparison
31 remains qualitative due to differences among the simulation and experiments as to where the
32 outlet pressure was measured and because no tubing attachments or other experimental apparatus
33 were included in the simulation geometry. The simulated pressure drop values were somewhat
34 lower, possibly due to these differences.
35
This document is a draft for review purposes only and does not constitute Agency policy.
3-45 DRAFT—DO NOT CITE OR QUOTE
-------
800
1
2
3
4
Nasal Mold C1
In Vivo
Nasal Mold C2
CIITCFD
100 200 300 400 500 600 700 800 900 1000
Airflow rate (ml/min)
Figure 3-11. Pressure drop vs. volumetric airflow rate predicted by the CUT
CFD model compared with pressure drop measurements made in two hollow
molds (Cl and C2) of the rat nasal passage (Cheng et al., 1990) or in rats
in vivo (Gerde et al., 1991).
Source: Kimbell et al. (1997a).
5 Inspiratory airflow was assumed to be constant in time (steady state). Subramaniam et al.
6 (1998) considered this to be a reasonable assumption during resting breathing conditions based
7 on a value of 0.02 obtained for the Strouhal number. Unsteady effects are insignificant when
8 this number is much less than one. However, this assumption may not be reasonable for light
9 and heavy exercise breathing scenarios.
10
11 3.7.5.2. Level of Confidence in Formaldehyde Uptake Simulations
12 Unlike the airflow simulations, it was not possible to evaluate the formaldehyde uptake
13 calculations directly. Since the mass transfer boundary conditions were set by fitting overall
14 uptake to the average experimental data for various exposure concentrations, it was not possible
15 to independently verify even the overall uptake values with empirical data. This assessment has
16 relied on several indirect qualitative and quantitative lines of evidence listed below to provide
17 general confidence in the uptake profile for the F344 rat nasal passages, as modeled in CUT
18 (1999), when gross averages are considered over certain regions of the nasal lining.
This document is a draft for review purposes only and does not constitute Agency policy.
3-46 DRAFT—DO NOT CITE OR QUOTE
-------
1 In an earlier simulation, where the nasal walls were set to be infinitely absorbing of
2 formaldehyde, uptake of inhaled formaldehyde in the upper respiratory tract was predicted to be
3 90% in the rat for simulations corresponding to the resting minute volume in the F344 rat. This
4 estimate compared reasonably well with the range of 91-98% observed by Morgan et al.
5 (1986a).
6 Morgan et al. (1991) showed general qualitative correspondence between the main routes
7 of flow and lesion distribution induced by formaldehyde in the rat nose. In their initial work
8 with a CFD model that represented a highly reactive and soluble gas, Kimbell et al. (1998, 1993)
9 described similarities in computed regional mass flux patterns and lesion distribution due to
10 formaldehyde. When the results from this work in the coronal section immediately posterior to
11 the vestibular region were considered, simulated flux levels over regions such as the medial
12 aspect of the maxilloturbinate and the adjacent septum (where lesions were seen) were an order
13 of magnitude higher than over other regions, such as the nasoturbinate (where lesions were not
14 seen).8
15 The results of a PBPK model by Cohen-Hubal et al. (1997) provide a reasonable level of
16 confidence in regional uptake simulations for the F344 rat when gross averages over nasal sites
17 are carried out. Cohen-Hubal et al. (1997) linked the CFD dosimetry model for formaldehyde to
18 a PBPK model for formaldehyde-DPX concentration in the F344 rat. This PBPK model was
19 calibrated by optimizing the model to combined DPX data from all regions of the rat nose (high-
20 tumor and low-tumor incidence regions) that were obtained in separate experiments by Casanova
21 et al. (1991, 1989). These data were obtained at 0.3, 0.7, 2.0, 6.0, and 10 ppm for both regions.
22 DPX data were also obtained at 15 ppm exposure from the high-tumor region; however these
23 were not included for the calibration. Model prediction of DPX concentrations were then
24 compared with data for the high-tumor region only and compared well with the experimental
25 data, including the 15 ppm data for which the model had not been calibrated. This is shown in
26 Figure 3-12. Such a verification, albeit indirect, is not available for the simulation of uptake
27 patterns in the human.
28 The CFD simulations do not model reflex bradypnea, a protective reflex seen in rodents
29 but not in humans. As discussed at length in Sections 3.2.3.1 and 4.2.1.1, it is reasonable to
8 However, this 1993 CFD model differed somewhat from the subsequent model by Kimbell et al. (200 la) used in
this assessment. In the 1993 model, the limiting mass-transfer resistance for the gas was assumed to be in the air
phase; that is, the concentration of formaldehyde was set to zero at the airway lining. Furthermore, this same
boundary condition was used on the nasal vestibule as well, while, in the more recent model, the vestibule was
considered to be nonabsorbing. Unfortunately, Kimbell at al. (200 la) did not report on correspondences between
flux patterns and lesion distribution.
This document is a draft for review purposes only and does not constitute Agency policy.
3-47 DRAFT—DO NOT CITE OR QUOTE
-------
HCHO Flux (^MtnmAnirO
Z19.B Z55.5
0.8
2.5
0,0
2.5 5.8 7.5 10.B li.5 15.8 17.5
HCHO Exposure Concentration (ppm)
HCHO Flux (nM-mmAnln)
34.s ea.e 103.s IBB.B i?;.5 z»7,e 241.5
2.5 S.8 7.5 !«.» 12.S 15.8 17.5
HCHO Exposure Concentration (ppm)
Figure 3-12. Formaldehyde-DPX dosimetry in the F344 rat.
Panel A: calibration of the PBPK model using data from high and low tumor
incidence sites. Panel B: model prediction compared against data from high
tumor incidence site. Dashed line in panel A shows the extrapolation outside the
range of the calibrated data.
Source: Cohen-Hubal etal. (1997).
2 expect a range of 25% (Chang et al., 1983) to 45% (Barrow et al., 1983) decrease in minute
3 volume in F344 rats at the exposure concentration of 15 ppm. Explicit omission of this effect in
4 the modeling is, however, not likely to be a source of major uncertainty in the modeled results
This document is a draft for review purposes only and does not constitute Agency policy.
3-48 DRAFT—DO NOT CITE OR QUOTE
-------
1 for uptake of formaldehyde in the rat nose for the following reason. The CFD model for the
2 F344 rat was calibrated to fit the overall experimental result for formaldehyde uptake in the F344
3 rat at 15 ppm exposure concentration. This was carried out by adjusting the mass transfer
4 coefficient used as boundary condition on the absorbing portion of the nasal lining. Thus, the
5 reflex bradypnea occurring in those experimental animals is phenomenologically factored into
6 the value used for the boundary condition. Nonetheless, some error in the localized distribution
7 of uptake patterns may be expected, even if the overall uptake is reproduced correctly.
8 Furthermore, since the same value for the mass transfer coefficient was used in human
9 simulations (as obtained from calibration of the rat model), there is additional uncertainty in the
10 modeled human flux estimates. This issue was not addressed by Kimbell et al. (2001a, b),
11 Conolly et al. (2004), or Schlosser et al. (2003), and we are unable to assess the extent of this
12 error more accurately.
13
14 3.7.6. PBPK Modeling of DNA Protein Cross-Links (DPXs) Formed by Formaldehyde
15 3.7.6.1. PBPK Models for DPXs
16 As can be seen from the previous sections, measuring the distribution of the absorbed
17 formaldehyde and identifying its form have proven difficult. Because of the high reactivity of
18 formaldehyde, rapid metabolism of formaldehyde, and complexity of formate clearance, dose
19 surrogates (or biomarkers) of exposure have been used to characterize the extent of absorption
20 and distribution of formaldehyde. As with other soluble and reactive gases, typical PBPK
21 models that predict steady-state blood concentrations are not useful for predicting formaldehyde
22 dosimetry at this time. As noted previously, inhalation exposure to formaldehyde has not been
23 shown to increase blood formaldehyde levels. Thus, most modeling efforts for formaldehyde
24 have focused on disposition at the site of contact.
25 As discussed earlier, the concentration of DPXs formed by formaldehyde has been
26 treated as a surrogate for the tissue dose of formaldehyde in earlier efforts by Casanova et al.
27 (1991) and in EPA's efforts to update its health assessment of formaldehyde (Hernandez et al.,
28 1994). These efforts used data from rats and rhesus monkeys (Casanova et al., 1991, 1989).
29 Using DPXs in this manner allowed the incorporation of both clearance and metabolism of
30 formaldehyde and the incorporation of the effect of saturation on detoxification of formaldehyde
31 at higher doses. Calculation of the average DPX concentration from these data was seen as a
32 surrogate for the area under the curve (AUC) of the reactive formaldehyde species in the
33 epithelium. Based on these data, Casanova et al. (1991) developed a PBPK model for predicting
34 DPXs in these species and for extrapolating to the human.
This document is a draft for review purposes only and does not constitute Agency policy.
3-49 DRAFT—DO NOT CITE OR QUOTE
-------
1 The Casanova et al. (1991) model consists of three anatomical compartments
2 representing different parts of the upper respiratory tract of the rhesus monkey. The results
3 indicated a 10-fold difference in DPX formation between rats and monkeys, due primarily to
4 species differences in minute volume and differing quantities of DNA in the nasal mucosa.
5 Casanova et al. (1991) then developed a monkey /rat scaling factor for these parameters by taking
6 the ratio of nasal mucosa tissue between the two species, a determinant that was proportional to
7 the total body weight differences between the two species. Using these scaling factors in their
8 model, the authors' predictions in monkey (based on the rat data) were in close agreement with
9 observed DPXs in monkey, particularly at higher formaldehyde concentrations. However, the
10 model overpredicted DPX formation in the monkey at lower formaldehyde concentrations.
11 Subsequent rat-human and monkey-human scaling results predicted much lower DPX formation
12 in man. Again, the values obtained at lower concentrations may have been overpredicted, as was
13 the case for the rat-monkey extrapolation.
14 Georgieva et al. (2003) developed a model for the uptake and disposition of
15 formaldehyde in the rat nasal lining. This model was designed to predict the distribution of
16 formaldehyde in the nasal mucosa. The model indicated that, at 6 ppm exposure, a steady-state
17 elevation of 15-20 jiM formaldehyde would be achieved within 30 seconds. Furthermore, this
18 same elevation was predicted when the exposure was 6 ppm formaldehyde for 60 minutes.
19 Given that human blood formaldehyde levels are predicted to be about 100 ±15 jiM (Heck et al.,
20 1985) and assuming that blood formaldehyde concentration is roughly equivalent to the
21 concentration predicted at the basement membrane of the epithelium, this model predicts roughly
22 a 15-20% increase in blood formaldehyde. However, it should be noted that a 40-minute
23 inhalation exposure of humans to 1.99 ppm formaldehyde did not lead to a measurable increase
24 in blood formaldehyde (Heck et al., 1985).
25 Franks (2005) published a mathematical model for predicting the disposition of
26 formaldehyde in the human nasal mucosa and blood. The calculated concentrations of
27 formaldehyde in the mucus, the epithelium, and the blood attained steady-state profiles within a
28 few seconds of exposure. The increase of the formaldehyde concentration in the blood was
29 predicted to be insignificant compared with the existing pre-exposure levels in the body: an
30 increase of 0.00044 mg/L in blood formaldehyde following exposure to 1.9 ppm formaldehyde
31 for up to 8 hours. The model described formaldehyde concentration gradients across the mucus,
32 epithelial, and submucosal compartments in the human nose. Transport of formaldehyde was
33 governed by the following processes: diffusional (in the mucus); a combination of diffusional,
34 two first order terms representing intrinsic reactivity of formaldehyde and binding to DNA, and
35 Michaelis-Menten kinetics representing enzymatic metabolism (in the epithelial layer); a first-
This document is a draft for review purposes only and does not constitute Agency policy.
3-50 DRAFT—DO NOT CITE OR QUOTE
-------
1 order term representing nonenzymatic removal governed by the blood perfusion rate (in the
2 submucosal compartment). The model used the values for the first order reaction rate constants
3 and the Michaelis-Menten parameters (Vmax and Km) estimated by Conolly et al. (2000) in their
4 model for extrapolating the rat and rhesus monkey data to the human. The modeling in Franks
5 (2005) was not calibrated or validated against experimental data, but the predictions of
6 negligible penetration of free formaldehyde to the blood are qualitatively in agreement with the
7 conclusions in Heck et al. (1985).
8 Following the efforts by Casanova and coworkers, Cohen-Hubal et al. (1997), Conolly et
9 al. (2000), and Georgieva et al. (2003) developed models that linked local formaldehyde flux
10 from CFD models to DPX predictions. The focus here will be on the Conolly et al. (2000) effort
11 for the following two reasons: it explicitly incorporates regional formaldehyde dosimetry in the
12 nasal lining by using results from CFD modeling of airflow and gas uptake and it brings data
13 across species (rat and rhesus monkey) to bear on model calibration, such a situation being
14 relatively rare in chemical health risk assessments.
15
16 3.7.6.2. A PBPK Model for DPXs in the F344 Rat and Rhesus Monkey that Uses Local
17 Tissue Dose of Formaldehyde
18 In earlier risk assessment efforts (Hernandez et al., 1994; Casanova et al., 1991; U.S.
19 EPA, 1991b), the average DPX concentration was considered a surrogate tissue dose metric for
20 the AUC of the reactive formaldehyde species. Conolly et al. (2003) assigned a more specific
21 role for DPXs, treating local DPX concentration as a dose surrogate indicative of the
22 intercellular concentration of formaldehyde, leading to formaldehyde-induced mutations. These
23 authors indicated that it was not known whether DPXs directly induced mutations (Conolly et
24 al., 2003; Merk and Speit, 1998). This is discussed in detail in the mode-of-action sections in
25 this document. The Conolly et al. (2000) model for the disposition of inhaled formaldehyde gas
26 and DPX in the rat and rhesus nasal lining is relatively simple in terms of model structure
27 because it consists of a single well-mixed compartment for the nasal lining as follows:
28
29 1. Formaldehyde flux to a given region of the nasal lining is provided as input to the
30 modeling and is obtained in turn as the result of a CFD model. This flux is defined as the
31 amount of formaldehyde delivered to the nasal lining per unit time per unit area per ppm
32 of concentration in the air in a direction transverse to the airflow. It is locally defined as
33 a function of location in the nose and the inspiratory flow rate and is linear with exposure
34 concentration.
35 2. The clearance of formaldehyde from the tissue is modeled as follows:
This document is a draft for review purposes only and does not constitute Agency policy.
3-51 DRAFT—DO NOT CITE OR QUOTE
-------
1 a. a saturable pathway representing enzymatic metabolism of formaldehyde, which
2 is primarily by formaldehyde dehydrogenase (involving Michaelis-Menten
3 parameters Vmax and Km)
4 b. a separate first-order pathway, which is assumed to represent the intrinsic
5 reactivity of formaldehyde with tissue constituents (rate constant kf)
6 c. first-order binding to DNA that leads to DPX formation (rate constant kb)
7 3. The clearance or repair of this DPX is modeled as a first order process (rate constant
8 ki0ss).
9
10 DPX data. DPX concentrations were estimated from a study by Casanova et al. (1994) in which
11 rats were exposed 6 hours/day, 5 days/week, plus 4 days for 11 weeks to filtered air (naive) or to
12 0.7, 2, 6, or 15 ppm (0.9, 2.5, 7.4, or 18 mg/m3) formaldehyde (pre-exposed). On the 5th day of
13 the 12th week, the rats were then exposed for 3 hours to 0, 0.7, 2, 6, or 15 ppm 14C-labeled
14 formaldehyde (with pre-exposed animals exposed to the same concentration as during the
15 preceding 12 weeks and 4 days). The animals were sacrificed and DPX concentrations
16 determined at two sites in the nasal mucosa. Conolly et al. (2000) used these naive rat data to
17 develop a PBPK model that predicted the time-course of DPX concentrations as a function of
18 formaldehyde flux at these sites.9
19
20 3.7.6.3. Uncertainties in Modeling the Rat and Rhesus DPX Data
21 3.6.6.3.1. Half-life of DPX repair. In the development of the PBPK model for DPXs, Conolly
22 et al. (2000) assumed a value of 6.5 x io~3 minute"1 for kioss, the first-order rate constant for the
23 clearance (repair) of DPXs, such that the DPXs predicted at the end of a 6-hour exposure to
24 15 ppm were reduced to exactly the detection limit for DPXs in 18 hours (the period between the
25 end of 1 day's 6-hour exposure and the beginning of the next). This determination of rapid
26 clearance was based on an observation by Casanova et al. (1994) that the DPX concentrations
27 observed in the pre-exposed animals were not significantly higher than those in naive animals (in
28 which there was no significant DPX accumulation). However, in vitro data (Quievryn and
29 Zhitkovich, 2000) indicate a much slower clearance, with an average kioss of
30 9.24 x 10~4 minute"1.
31 Subramaniam et al. (2007) examined the Casanova et al. (1994) data and argued that
32 there was a significantly decreased (~ 40%) level of DPXs in high tumor regions of pre-exposed
33 animals vs. naive animals at 6 and 15 ppm and that the weight of the tissues dissected from those
9 Subramaniam et al. (2007) who also used the same data verified that they were on naive rats; however, Conolly et
This document is a draft for review purposes only and does not constitute Agency policy.
3-52 DRAFT—DO NOT CITE OR QUOTE
-------
1 regions increased substantially, indicating a thickening of the tissues. After testing the outcome
2 of changing the tissue thickness in the PBPK model for DPXs, it was apparent to these authors
3 that such a change alone could not account for the dramatic reduction in DPX levels after
4 pre-exposure, even with the higher value of kioss used by Conolly et al. (2000). Therefore, in
5 addition to the gross increase in tissue weight, these data indicated either an induction in the
6 activity of enzymes that remove formaldehyde (aldehyde and formaldehyde dehydrogenase) or
7 other changes in the biochemical properties of the highly exposed tissue that must have occurred.
8 Given such a change, Subramaniam et al. (2007) concluded that the experimental results in
9 Casanova et al. (1994) were consistent with the smaller experimental value of kioss indicated by
10 the Quievryn and Zhitkovich (2000) data. In particular, they argued that if Vmax increased with
11 exposure (in a tissue region- and dose-specific manner), then it was possible to explain the naive
12 vs. pre-exposed data of Casanova et al. (1994), with the value of kioss effectively measured in
13 vitro by Quievryn and Zhitkovich (2000). Furthermore, this value was measured directly, rather
14 than obtained by indirect interpretation of measurements made at only two time points where
15 significant changes in the tissue had occurred. Therefore, Subramaniam et al. (2007) considered
16 the use of this lower value for kioss to be more appropriate. The same lower value of kioss was
17 also used by Georgieva et al. (2003). Consequently, they reimplemented and reoptimized the
18 Conolly et al. (2000) model with this modification and found that the fit so obtained to the acute
19 DPX data was excellent. The reimplemented model will be used in this assessment, and more
20 details can be found in Subramaniam et al. (2007).
21 It should be noted that this slower DPX repair rate was obtained in an in vitro study by
22 using human cell lines that were transformed and immortalized. However, it appears that DPX
23 repair in normal cells would be even slower. When nontransformed freshly purified human
24 peripheral lymphocytes were used instead, the half-life for DPX repair was about 50% longer
25 than in the cultured cells (Quievryn and Zhitkovich, 2000).
26
27 3.6.6.3.2. Statistical uncertainty in parameter estimates and extrapolation. Klein et al. (2010)
28 developed methods for deriving statistical inferences of results from PBPK models, and used the
29 structure of the Conolly et al. (2000) model for demonstrating their methods, specifically
30 because of the sparse time-course information in the above DPX data. However, they used the
31 value of kioss deduced from Quievryn and Zhitkovich (2000) and fitted the model simultaneously
32 to both the rat and rhesus monkey data, as opposed to the sequential fitting in Conolly et al.
33 (2000). They found that the predicted DPX concentrations were extremely sensitive to Vmax and
al. (2000) state that they used data on pre-exposed rats.
This document is a draft for review purposes only and does not constitute Agency policy.
3-53 DRAFT—DO NOT CITE OR QUOTE
-------
1 tissue thickness as was also concluded by Georgieva et al. (2003) and Cohen-Hubal et al. (1997).
2 Km was seen to be substantially different across species, a finding that was attributed plausibly
3 to the involvement of more than one enzyme (Klein et al., 2010; Georgieva et al., 2003). Klein
4 et al. (2010) concluded that the two efforts (Conolly et al. [2000] vs. Klein et al. [2010]) resulted
5 in substantially different predictions outside the range of the observed data over which the
6 models were calibrated.
7 The differences between these models occur in spite of the fact that both methods use all
8 the available DPX data in both species and the same model structures. At the 0.1 ppm exposure
9 concentration, in general these authors obtained three- to fourfold higher DPX concentrations
10 averaged over a 24-hour period after exposure. Furthermore, the standard deviations in Klein et
11 al. (2010) for Vmax and Km were an order of magnitude higher and that for kf was 35-fold lower
12 than the corresponding standard deviations reported in Conolly et al. (2000). The relatively
13 larger standard deviation for kf resulted in this parameter becoming negative in Conolly et al.
14 (2000) at half the standard deviation below the maximum likelihood estimate (MLE) value. Note
15 that, at a negative value of kf, formaldehyde would be produced as opposed to being cleared
16 through its intrinsic reactivity.
17 Klein et al. (2010) concluded that these "remarkable differences outside the range of the
18 observed data suggest caution in the use of these models in a predictive sense for extrapolating to
19 human exposures."
20
21 3.7.7. Uncertainty in Prediction of Human DPX Concentrations
22 Conolly et al. (2000) used both the rat and rhesus monkey data to predict human DPX
23 concentrations and constructed a PBPK model for the rhesus monkey along similar lines as for
24 the F344 rat. In the rhesus monkey model, they maintained the same values of kb, kioss, and kf as
25 in the rat model but optimized the values of Vmax and Km against the rhesus monkey data from
26 Casanova et al. (1994). The rat and rhesus monkey parameters were then used to construct a
27 human model (see Conolly et al. [2000] for a more detailed report of implementing the rhesus
28 monkey model and the extrapolating to humans).
29 For the human, the model used the value of Km obtained in the rhesus monkey model and
30 the epithelial thickness averaged over three regions of the rhesus monkey nose. The maximum
31 rate of metabolism, Vmax, which was estimated independently for the rat and rhesus monkey by
32 fitting to the DPX data available for these species, was then extrapolated to the human by
33 assuming a power law scaling with body weight (BW) (i.e., Vmax = a x BWb), and the coefficient
34 "a" and exponent "b" were derived from the independently estimated values of (Vmax)RAT and
This document is a draft for review purposes only and does not constitute Agency policy.
3-54 DRAFT—DO NOT CITE OR QUOTE
-------
1 (Vmax)MONKEY. Table 3-8 gives the values of Vmax and Km in the Conolly et al. (2000)
2 extrapolation.
3 Table 3-8. Extrapolation of parameters for enzymatic metabolism to the
4 human
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Parameter
Vmax (pmol/min-
mm3)
Km (pmol/mm3)
F344 rat
1,008.0
70.8
Rhesus monkey
91.0
6.69
Human
15.7
6.69
Source: Conolly et al. (2000).
The above scale-up procedure was an attempt to use both the rodent and primate DPX
data. However, laws for allometric scaling across species, such as how enzymatic metabolic
rates vary across organisms, are empirical regression relationships whose strength is that they are
based on data from multiple species and usually multiple sources of data points. For example,
3/4
West and Brown (2005) demonstrate that metabolic rates scale with mass using data from
organisms ranging over 27 orders of magnitude in mass (intracellular up to the largest
organisms). In Conolly et al. (2000) the power-law relationship is derived using two data points
(F344 rat and rhesus monkey for a single chemical) with log BW as x-axis and Vmax on y-axis.
Since such a regression does not have the power to delineate the curvature in the scaling
function, the empirical strength of the allometric relationship derived in Conolly et al. (2000) is
extremely weak for use in extrapolating from the rat to the human on the basis of body-weight.
The following observations point to the uncertainty in the values of the parameters Vmax
and Km in the Conolly et al. (2000) models for predicting DPXs. First, Km varies by an order of
magnitude across the rat and monkey models but is then considered invariant between the
monkey and human models (Conolly et al., 2000). Second, the values in Conolly et al. (2000)
for Vmax/Km, the low-dose limit of the rate of enzymatic metabolism, is roughly similar between
the rat and monkey but lower by a factor of six in the human.
Another factor that can substantially influence the above extrapolation of DPXs in the
human is that Conolly et al. (2000) assumed the tissue to be a well-mixed compartment with
regard to formaldehyde interaction with DNA and used the amount of formaldehyde bound to
DNA per unit volume of tissue as the DPX dose metric. Considering formaldehyde's highly
reactive nature, the concentrations of formaldehyde and DPX are likely to have a sharp gradient
with distance into the nasal mucosa (Georgieva et al., 2003). Given the interspecies differences
in tissue thickness, there is consequent uncertainty as to whether DPX per unit volume or DPX
This document is a draft for review purposes only and does not constitute Agency policy.
3-55 DRAFT—DO NOT CITE OR QUOTE
-------
1 per unit area of nasal lining is the more appropriate dose metric to be used in the extrapolation.
2 In particular, it may be assumed that the cells at risk for tumor formation are only those in the
3 epithelium and that measured DPX data (in monkeys and rats) are an average over the entire
4 tissue thickness. Since the epithelial DPXs in monkeys (and presumably humans) would then be
5 more greatly "diluted" by lower levels of DPX formation that occur deeper into the tissue than in
6 rats, it could be predicted that the ratio of epithelial to measured DPXs in monkeys and humans
7 would be much higher than the ratio in rats.
8
9 3.7.8. Modeling Interindividual Variability in the Nasal Dosimetry of Reactive and
10 Soluble Gases
11 Garcia et al. (2009) used computational fluid dynamics to study human variability in the
12 nasal dosimetry of reactive, water-soluble gases in 5 adults and 2 children, aged 7 and 8 years
13 old. The sample size in this study is too small to consider the results representative of the
14 population as a whole (as also recognized by the authors). Nonetheless, various comparisons
15 with the characteristics of other study populations add to the strength of this study (see
16 Appendix B). The authors considered two model categories of gases, corresponding to maximal
17 and moderate absorption at the nasal lining. We focus here only on the "maximum uptake"
18 simulations in Garcia et al. (2009). In this case, the gas was considered so highly reactive and
19 soluble that it was reasonable to assume an infinitely fast reaction of the absorbed gas with
20 compounds in the airway lining. Although such a gas could be reasonably considered as a proxy
21 for formaldehyde, these results cannot be fully utilized to inform quantitative estimates of
22 formaldehyde dosimetry (and it does not appear to have been the intent of the authors either).
23 This is because the same boundary condition corresponding to maximal uptake was applied on
24 the vestibular lining of the nose as well as on the respiratory and transitional epithelial lining on
25 the rest of the nose. This is not appropriate for formaldehyde as the lining on the nasal vestibule
26 is made of keratinized epithelium which is considerably less absorbing than the rest of the nose
27 (Kimbell et al., 2001b).
28 The Garcia et al. (2009) study and the results of their analyses have been further
29 described and evaluated in Appendix B. Overall uptake efficiency, average flux (rate of gas
30 absorbed per unit surface area of the nasal lining) and maximum flux levels over the entire nasal
31 lining did not vary substantially between adults (1.6-fold difference in average flux and much
32 less in maximum flux), and the mean values of these quantities were comparable between adults
33 and children. These results are also in agreement with conclusions reached by Ginsberg et al.
34 (2005) that overall extrathoracic absorption of highly and moderately reactive and soluble gases
35 (corresponding to Category 1 and 2 reactive gases as per the scheme in EPA [1994]) is similar in
This document is a draft for review purposes only and does not constitute Agency policy.
3-56 DRAFT—DO NOT CITE OR QUOTE
-------
1 adults and children. On the other hand, Figure 6 A of the paper (reproduced as Figure B-l in
2 Appendix B), provides a different perspective on variations between the adults in flux values at
3 specific points on the nasal walls. The plot indicates that local flux of formaldehyde may vary
4 among individuals by a factor of 3 to 5 at various distances along the septal axis of the nose;
5 such an evaluation of inter-individual variability in the spatial distribution of formaldehyde flux
6 over the nasal lining is important for a highly reactive and soluble gas whose regional absorption
7 is highly nonhomogeneously distributed (see text surrounding Figure 3-8).
This document is a draft for review purposes only and does not constitute Agency policy.
3-57 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
End of Volume I
This document is a draft for review purposes only and does not constitute Agency policy.
3-58 DRAFT—DO NOT CITE OR QUOTE
-------
DRAFT - DO NOT CITE OR QUOTE EPA/635/R-10/002A
^ ™™^j» www.epa.gov/iris
TOXICOLOGICAL REVIEW OF
FORMALDEHYDE -
INHALATION ASSESSMENT
(CAS No. 50-00-0)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
VOLUME II of IV
Hazard Characterization
June 2, 2010
NOTICE
This document is an External 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
-------
1 DISCLAIMER
2
3 This document is a preliminary draft for review purposes only. This information is
4 distributed solely for the purpose of pre-dissemination peer review under applicable information
5 quality guidelines. It has not been formally disseminated by EPA. It does not represent and
6 should not be construed to represent any Agency determination or policy. Mention of trade
7 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.
Il-ii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS IN BRIEF
VOLUME I: INTRODUCTION, BACKGROUND, AND TOXICOKINETICS
Chapter 1: Introduction
Chapter 2: B ackground
Chapter 3: Toxicokinetics
VOLUME II: HAZARD CHARACTERIZATION
Chapter 4: Hazard Characterization
VOLUME III: QUANTITATIVE ASSESSMENT, MAJOR CONCLUSIONS IN THE
CHARACTERIZATION OF HAZARD AND DOSE RESPONSE, AND
REFERENCES
Chapter 5: Quantitative Assessment: Inhalation Exposure
Chapter 6: Major Conclusions in the Characterization of Hazard and Dose-Response
References
VOLUME IV: APPENDICES
Appendix A: Summary of External Peer Review and Public Comments and Dispositions
Appendix B: Simulations of Interindividual and Adult-to-Child Variability in Reactive
Gas Uptake in a Small Sample of People (Garcia et al., 2009)
Appendix C: Lifetable Analysis
Appendix D: Model Structure & Calibration in Conolly et al. (2003, 2004)
Appendix E: Evaluation of BBDR Modeling of Nasal Cancer in the F344 Rat: Conolly
et al. (2003) and Alternative Implementations
Appendix F: Sensitivity Analysis of BBDR Model for Formaldehyde Induced
Respiratory Cancer in Humans
Appendix G: Evaluation of the Cancer Dose-Response Modeling of Genomic Data for
Formaldehyde Risk Assessment
Appendix H: Expert Panel Consultation on Quantitative Evaluation of Animal
Toxicology Data for Analyzing Cancer Risk Due to Inhaled Formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
II-iii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS—TOXICOLOGICAL REVIEW OF FORMALDEHYDE
(CAS No. 50-00-0)
VOLUME II
LIST OF TABLES II-viii
LIST OF FIGURES II-xv
4. HAZARD CHARACTERIZATION 4-1
4.1. HUMAN STUDIES 4-1
4.1.1. Noncancer Health Effects 4-1
4.1.1.1. Sensory Irritation (Eye, Nose, Throat Irritation) 4-1
4.1.1.2. Pulmonary Function 4-11
4.1.1.3. Asthma 4-50
4.1.1.4. Respiratory Tract Pathology 4-65
4.1.1.5. Immunologic Effects 4-69
4.1.1.6. Neurological/Behavioral 4-82
4.1.1.7. Developmental and Reproductive Toxicity 4-85
4.1.1.8. Oral Exposure Effects on the Gastrointestinal Tract 4-96
4.1.1.9. Summary: Noncarcinogenic Hazard in Humans 4-96
4.1.2. Cancer Health Effects 4-97
4.1.2.1. Respiratory Tract Cancer 4-98
4.1.2.2. Nonrespiratory Tract Cancer 4-134
4.1.2.3. Summary: Carcinogenic Hazard in Humans 4-188
4.2. ANIMAL STUDIES 4-191
4.2.1. Reflex Bradypnea 4-192
4.2.1.1. Tolerance 4-195
4.2.1.2. Cross-Species Differences in Inhaled Dose 4-195
4.2.1.3. Cross-Tolerance 4-198
4.2.1.4. Formaldehyde Binding and Activation of Trigeminal Nerve
Afferent Activity 4-199
4.2.2. Respiratory Tract Pathology 4-201
4.2.2.1. Mucociliary Clearance 4-202
4.2.2.2. Cell Proliferation 4-209
4.2.2.3. Short-Term Studies 4-226
4.2.2.4. Subchronic Studies 4-241
4.2.2.5. Chronic Inhalation Bioassays 4-261
4.2.2.6. Summary of Respiratory Pathology and Carcinogenic Potential.... 4-287
4.2.3. Gastrointestinal Tract Pathology 4-292
4.2.3.1. Short-Term and Subchronic Ingestion Studies 4-292
4.2.3.2. Chronic Ingestion Studies 4-295
4.2.3.3. Summary of Gastrointestinal Effects and Evaluation of
Carcinogenic Potential 4-301
4.2.4. Immune Function 4-301
4.2.5. Hypersensitivity and Atopic Reactions 4-310
This document is a draft for review purposes only and does not constitute Agency policy.
Il-iv DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.2.5.1. Inhalation Studies in Experimental Animals 4-312
4.2.5.2. Dermal Sensitization 4-332
4.2.5.3. Summary of Sensitization Studies 4-334
4.2.6. Neurological and Neurobehavioral Function 4-339
4.2.6.1. Inhalation Exposure 4-339
4.2.6.2. Oral Exposure 4-368
4.2.6.3. Summary 4-370
4.2.6.4. Other Considerations 4-370
4.2.7. Reproductive and Developmental Toxicity 4-371
4.2.7.1. Inhalation Studies Addressing Developmental and
Reproductive Toxicity 4-371
4.2.7.2. Oral Exposure Studies Addressing Developmental and
Reproductive Toxicity 4-388
4.2.7.3. Intraperitoneal Studies Addressing Developmental and
Reproductive Toxicity 4-391
4.2.7.4. Dermal Exposure Studies Addressing Developmental and
Reproductive Toxicity 4-393
4.2.7.5. Summary of Reproductive and Developmental Effects 4-394
4.3. GENOTOXICITY 4-407
4.3.1. Formaldehyde-DNAReactions 4-407
4.3.1.1. DNA-Protein Cross-Links (DPXs) 4-410
4.3.1.2. DNAAdducts 4-416
4.3.1.3. DNA-DNA Cross-Links (DDXs) 4-419
4.3.1.4. Single Strand Breaks 4-419
4.3.1.5. Other Genetic Effects of Formaldehyde in Mammalian Cells 4-421
4.3.2. In Vitro Clastogenicity 4-422
4.3.3. In Vitro Mutagenicity 4-425
4.3.3.1. Mutagenicity in Bacterial Systems 4-425
4.3.3.2. Mutagenicity in Other Nonmammalian Cell Systems 4-430
4.3.3.3. Mutagenicity in Mammalian Cell Systems 4-430
4.3.4. In Vivo Mammalian Genotoxicity 4-436
4.3.4.1. Genotoxicity in Laboratory Animals 4-436
4.3.4.2. Genotoxicity in Humans 4-439
4.3.5. Summary of Genotoxicity 4-449
4.4. SYNTHESIS AND MAJOR EVALUATION OF NONCARCINOGENIC
EFFECTS 4-450
4.4.1. Sensory Irritation 4-450
4.4.2. Pulmonary Function 4-458
4.4.3. Hypersensitivity and Atopic Reactions 4-462
4.4.4. Upper Respiratory Tract Histopathology 4-467
4.4.5. Toxicogenomic and Molecular Data that may Inform MO As 4-469
4.4.6. Noncancer Modes of Actions 4-470
4.4.7. Immunotoxicity 4-473
This document is a draft for review purposes only and does not constitute Agency policy.
II-v DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.4.8. Effects on the Nervous System 4-474
4.4.8.1. Irritant Threshold Detection 4-474
4.4.8.2. Behavioral Effects 4-474
4.4.8.3. Neurochemistry, Neuropathology, and Mechanistic Studies 4-475
4.4.8.4. Summary 4-476
4.4.8.5. Data Gaps 4-476
4.4.9. Reproductive and Developmental Toxicity 4-477
4.4.9.1. Spontaneous Abortion and Fetal Death 4-477
4.4.9.2. Congenital Malformations 4-480
4.4.9.3. Low Birth Weight and Growth Retardation 4-480
4.4.9.4. Functional Developmental Outcomes 4-481
4.4.9.5. Male Reproductive Toxicity 4-481
4.4.9.6. Female Reproductive Toxicity 4-483
4.4.9.7. Mode of Action 4-484
4.4.9.8. Data Gaps 4-486
4.5. SYNTHESIS AND EVALUATION OF CARCINOGENICITY 4-486
4.5.1. Cancers of the Respiratory Tract 4-486
4.5.2. Lymphohematopoietic Malignancies 4-491
4.5.2.1. Background 4-491
4.5.2.2. All LHP Malignancies 4-492
4.5.2.3. All Leukemia 4-496
4.5.2.4. Subtype Analysis 4-498
4.5.2.5. Myeloid Leukemia 4-499
4.5.2.6. Solid Tumors of Lymphoid Origin 4-502
4.5.2.7. Supporting Evidence from Animal Bio-Assays for
Formaldehyde-Induced Lymphohematopoietic Malignancies 4-506
4.5.3. Carcinogenic Mode(s) of Action 4-510
4.5.3.1. Mechanistic Data for Formaldehyde 4-512
4.5.3.2. Mode of Action Evaluation for Upper Respiratory Tract
Cancer (Nasopharyngeal Cancer, Sino-nasal) 4-523
4.5.3.3. Mode(s) of Action for Lymphohematopoietic Malignancies 4-529
4.5.4. Hazard Characterization for Formaldehyde Carcinogenicity 4-535
4.6. SUSCEPTIBLE POPULATIONS 4-537
4.6.1. Life Stages 4-537
4.6.1.1. Early Life Stages 4-538
4.6.1.2. Later Life Stages 4-542
4.6.1.3. Conclusions on Life-Stage Susceptibility 4-542
4.6.2. Health/Disease Status 4-543
4.6.3. Nutritional Status 4-544
4.6.4. Gender Differences 4-545
4.6.5. Genetic Differences 4-545
4.6.6. Coexposures 4-547
4.6.6.1. Cumulative Risk 4-547
This document is a draft for review purposes only and does not constitute Agency policy.
Il-vi DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.6.6.2. Aggregate Exposure 4-548
4.6.7. Uncertainties of Database 4-548
4.6.7.1. Uncertainties of Exposure 4-548
4.6.7.2. Uncertainties of Effect 4-549
4.6.8. Summary of Potential Susceptibility 4-550
This document is a draft for review purposes only and does not constitute Agency policy.
II-vii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function 4-14
Table 4-2. Maj or cohort studies of formaldehyde exposure and nasopharyngeal cancer... 4-99
Table 4-3. Case-control studies of formaldehyde exposure and nasopharyngeal cancer.. 4-102
Table 4-4. Case-control studies of formaldehyde exposure and nasal and paranasal
(sinonasal) cancers 4-122
Table 4-5. Epidemiologic studies of formaldehyde and pharyngeal cancer 4-136
Table 4-6. Comparison of SMRs in the exposed workers to the SMRs in the unexposed
workers from NCI cohort reported by Beane-Freeman et al. (2009) 4-146
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers... 4-152
Table 4-8. Respiratory effects of formaldehyde-induced reflex bradypnea in various
strains of mice 4-193
Table 4-9. Respiratory effects of formaldehyde-induced reflex bradypnea in various
strains of rats 4-194
Table 4-10. Inhaled dose of formaldehyde to nasal mucosa of F344 rats and B6C3F1
mice exposed to 15 ppm 4-197
Table 4-11. Exposure regimen for cross-tolerance study 4-198
Table 4-12. Summary of formaldehyde effects on mucociliary function in the upper
respiratory tract 4-208
Table 4-13. Cell proliferation in nasal mucosa, trachea, and free lung cells isolated from
male Wistar rats after inhalation exposures to formaldehyde 4-214
Table 4-14. The effect of repeated formaldehyde inhalation exposures for 3 months on c
ell count, basal membrane length, proliferation cells, and two measures of
cell proliferation, LI and ULLI, in male F344 rats 4-217
Table 4-15. Formaldehyde-induced changes in cell proliferation (ULLI) in the nasal
passages of male F344 rats exposed 6 hours/day 4-219
Table 4-16. Cell population and surface area estimates in untreated male F344 rats and
regional site location of squamous cell carcinomas in formaldehyde-
exposed rats for correlation to cell proliferation rates 4-220
This document is a draft for review purposes only and does not constitute Agency policy.
II-viii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-17. Summary of formaldehyde effects on cell proliferation in the upper
respiratory tract 4-222
Table 4-18. Concentration regimens for ultrastructural evaluation of male CDF rat
nasoturbinates 4-227
Table 4-19. Enzymatic activities in nasal respiratory epithelium of male Wistar rats
exposed to formaldehyde, ozone, or both 4-229
Table 4-20. Lipid analysis of lung tissue and lung lavage from male F344 rats exposed
to 0, 15, or 145.6 ppm formaldehyde for 6 hours 4-236
Table 4-21. Formaldehyde effects on biochemical parameters in nasal mucosa and
lung tissue homogenates from male F344 rats exposed to 0, 15, or
145.6 ppm formaldehyde for 6 hours 4-238
Table 4-22. Mast cell degranulation and neutrophil infiltration in the lung of rats
exposed to formaldehyde via inhalation 4-239
Table 4-23. Summary of respiratory tract pathology from inhalation exposures to
formaldehyde—short-term studies 4-242
Table 4-24. Location and incidence of respiratory tract lesions in B6C3F1 mice
exposed to formaldehyde 4-245
Table 4-25. Formaldehyde effects (incidence and severity) on histopathologic changes
in the noses and larynxes of male and female albino SPF Wistar rats
exposed to formaldehyde 6 hours/day for 13 weeks 4-247
Table 4-26. Formaldehyde-induced nonneoplastic histopathologic changes in male
albino SPF Wistar rats exposed to 0, 10, or 20 ppm formaldehyde and
examined at the end of 130 weeks inclusive of exposure 4-248
Table 4-27. Formaldehyde-induced nasal tumors in male albino SPF Wistar rats
exposed to formaldehyde (6 hours/day, 5 days/week for 13 weeks) and
examined at the end of 130 weeks inclusive of exposure 4-249
Table 4-28. Formaldehyde effects on nasal epithelium for various concentration-by-
time products in male albino Wistar rats 4-252
Table 4-29. Rhinitis observed in formaldehyde-treated animals; data pooled for male
and female animals 4-254
This document is a draft for review purposes only and does not constitute Agency policy.
Il-ix DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-30. Epithelial lesions found in the middle region of nasoturbinates of
formaldehyde-treated and control animals; data pooled for males and
females 4-254
Table 4-31. Cellular and molecular changes in nasal tissues of F344 rats exposed to
formaldehyde 4-256
Table 4-32. Percent body weight gain and concentrations of iron, zinc, and copper in
cerebral cortex of male Wistar rats exposed to formaldehyde via inhalation
for 4 and 13 weeks 4-257
Table 4-33. Zinc, copper, and iron content of lung tissue from formaldehyde-treated
male Wistar rats 4-258
Table 4-34. Total lung cytochrome P450 measurements of control and formaldehyde-
treated male Sprague-Dawley rats 4-259
Table 4-35. Cytochrome P450 levels in formaldehyde-treated rats 4-259
Table 4-36. Summary of respiratory tract pathology from inhalation exposures to
formaldehyde, subchronic studies 4-262
Table 4-37. Histopathologic findings and severity scores in the naso- and
maxilloturbinates of female Sprague-Dawley rats exposed to inhaled
formaldehyde and wood dust for 104 weeks 4-266
Table 4-38. Histopathologic changes (including tumors) in nasal cavities of male
Sprague-Dawley rats exposed to inhaled formaldehyde or HC1 alone and
in combination for a lifetime 4-270
Table 4-39. Summary of neoplastic lesions in the nasal cavity of F344 rats exposed to
inhaled formaldehyde for 2 years 4-274
Table 4-40. Apparent sites of origin for the SCCs in the nasal cavity of F344 rats
exposed to 14.3 ppm of formaldehyde gas in the Kerns et al. (1983) bioassay 4-
274
Table 4-41. Incidence and location of nasal squamous cell carcinoma in male F344
rats exposed to inhaled formaldehyde for 2 years 4-276
Table 4-42. Summary of respiratory tract pathology from chronic inhalation
exposures to formaldehyde 4-283
Table 4-43. Summary of lesions observed in the gastrointestinal tracts of Wistar rats
after drinking-water exposure to formaldehyde for 4 weeks 4-294
This document is a draft for review purposes only and does not constitute Agency policy.
II-x DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-44. Incidence of lesions observed in the gastrointestinal tracts of Wistar rats
after drinking-water exposure to formaldehyde for 2 years 4-297
Table 4-45. Effect of formaldehyde on gastroduodenal carcinogenesis initiated by
MNNG and NaCl in male Wistar rats exposed to formaldehyde in
drinking water for 8 weeks 4-300
Table 4-46. Battery of immune parameters and functional tests assessed in female
B6C3F1 mice after a 3 week, 15-ppm formaldehyde exposure 4-303
Table 4-47. Summary of the effects of formaldehyde inhalation on the mononuclear
phagocyte system (MPS) in female B6C3F1 mice after a 3-week, 15 ppm
formaldehyde exposure (6 hours/day, 5 days/week) 4-304
Table 4-48. Formaldehyde exposure regimens for determining the effects of
formaldehyde exposure on pulmonary S. aureus infection 4-306
Table 4-49. Summary of immune function changes due to inhaled formaldehyde
exposure in experimental animals 4-311
Table 4-50. Study design for guinea pigs exposed to formaldehyde through different
routes of exposure: inhalation, dermal, and injection 4-317
Table 4-51. Sensitization response of guinea pigs exposed to formaldehyde through
inhalation, topical application, or footpad injection 4-318
Table 4-52. Cytokine and chemokine levels in lung tissue homogenate supernatants in
formaldehyde-exposed male ICR mice with and without Der f sensitization . 4-325
Table 4-53. Correlation coefficients among ear swelling responses and skin mRNA
levels in contact hypersensitivity to formaldehyde in mice 4-334
Table 4-54. Summary of sensitization and atopy studies by inhalation or dermal
sensitization due to formaldehyde in experimental animals 4-336
Table 4-55. Fluctuation of behavioral responses when male AB mice inhaled
formaldehyde in a single 2-hour exposure: effects after 2 hours 4-344
Table 4-56. Fluctuation of behavioral responses when male AB mice inhaled
formaldehyde in a single 2-hour exposure: effects after 24 hours 4-344
Table 4-57. Effects of formaldehyde exposure on completion of the labyrinth test
by male and female LEW. IK rats 4-348
This document is a draft for review purposes only and does not constitute Agency policy.
Il-xi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-58. Summary of neurological and neurobehavioral studies of inhaled
formaldehyde in experimental animals 4-364
Table 4-59. Effects of formaldehyde on body and organ weights in rat pups from
dams exposed via inhalation from mating through gestation 4-374
Table 4-60. Formaldehyde effects on Leydig cell quantity and nuclear damage in
adult male Wistar rats 4-383
Table 4-61. Formaldehyde effects on adult male albino Wistar rats 4-384
Table 4-62. Formaldehyde effects on testosterone levels and seminiferous tubule
diameters in Wistar rats following 91 days of exposure 4-385
Table 4-63. Effects of formaldehyde exposure on seminiferous tubule diameter and
epithelial height in Wistar rats following 18 weeks of exposure 4-3 87
Table 4-64. Incidence of sperm abnormalities and dominant lethal effects in
formaldehyde-treated mice 4-387
Table 4-65. Body weights of pups born to beagles exposed to formaldehyde during
gestation 4-389
Table 4-66. Testicular weights, sperm head counts, and percentage incidence of
abnormal sperm after oral administration of formaldehyde to male Wistar
rats 4-390
Table 4-67. Effect of formaldehyde on spermatogenic parameters in male Wistar
rats exposed intraperitoneally 4-392
Table 4-68. Incidence of sperm head abnormalities in formaldehyde-treated rats 4-393
Table 4-69. Dominant lethal mutations after exposure of male rats to formaldehyde 4-394
Table 4-70. Summary of reported developmental effects in formaldehyde inhalation
exposure studies 4-396
Table 4-71. Summary of reported developmental effects in formaldehyde oral
exposure studies 4-402
Table 4-72. Summary of reported developmental effects in formaldehyde dermal
exposure studies 4-403
Table 4-73. Summary of reported reproductive effects in formaldehyde inhalation
studies 4-404
This document is a draft for review purposes only and does not constitute Agency policy.
II-xii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-74. Summary of reported reproductive effects in formaldehyde oral studies 4-408
Table 4-75. Summary of reported reproductive effects in formaldehyde
intraperitoneal studies 4-409
Table 4-76. Formaldehyde-DNA reactions (DPX formation) 4-411
Table 4-77. Formaldehyde-DNA reactions (DNA adduct formation) 4-417
Table 4-78. Formaldehyde-DNA interactions (single strand breaks) 4-420
Table 4-79. Other genetic effects of formaldehyde in mammalian cells 4-421
Table 4-80. In vitro clastogenicity of formaldehyde 4-423
Table 4-81. Summary of mutagenicity of formaldehyde in bacterial systems 4-427
Table 4-82. Mutagenicity in mammalian cell systems 4-431
Table 4-83. Genotoxicity in laboratory animals 4-437
Table 4-84. MN frequencies in buccal mucosa cells of volunteers exposed to
formaldehyde 4-441
Table 4-85. MN and SCE formation in mortuary science students exposed to
formaldehyde for 85 days 4-442
Table 4-86. Incidence of MN formation in mortuary students exposed to
formaldehyde for 90 days 4-443
Table 4-87. Multivariate repression models linking genomic instability/occupational
exposures to selected endpoints from the MN assay 4-448
Table 4-88. Genotoxicity measures in pathological anatomy laboratory workers
and controls 4-448
Table 4-89. Summary of human cytogenetic studies 4-452
Table 4-90. Summary of cohort and case-control studies which evaluated the
incidence of all LHP cancers in formaldehyde-exposed populations and
all leukemias 4-493
Table 4-91. Secondary analysis of published mortality statistics to explore alternative
disease groupings within the broad category of all lymphohematopoietic
malignancies 4-495
This document is a draft for review purposes only and does not constitute Agency policy.
II-xiii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-92. Summary of studies which provide mortality statistics for myeloid leukemia
subtypes 4-500
Table 4-93. Summary of mortality statistics for Hodgkin's lymphoma, lymphoma and
multiple myeloma from cohort analyses of formaldehyde exposed workers ..4-503
Table 4-94. Leukemia incidence in Fischer 344 rats surviving at least 21 months 4-508
Table 4-95. Available evidence for susceptibility factors of concern for formaldehyde
exposure 4-551
This document is a draft for review purposes only and does not constitute Agency policy.
II-xiv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
Figure 4-1. Delayed asthmatic reaction following the inhalation of formaldehyde
after "painting" 100% formalin for 20 minutes 4-51
Figure 4-2. McGwin Forest plot of relative risk estimates and 95% CIs from studies
included in a meta-analysis of formaldehyde exposure and asthma in
children based on the random effects models 4-60
Figure 4-3. Association between peak formaldehyde exposure and the risk of
lymphohematopoietic malignancy 4-148
Figure 4-4. Association between average intensity of formaldehyde exposure and
the risk of lymphohematopoietic malignancy 4-149
Figure 4-5. Formaldehyde effects on minute volume in naive and formaldehyde-
pretreated maleB6C3Fl mice andF344 rats 4-196
Figure 4-6. Sagittal view of the rat nose (nares oriented to the left) 4-202
Figure 4-7. Main components of the nasal respiratory epithelium 4-203
Figure 4-8. Decreased mucus clearance and ciliary beat in isolated frog palates
exposed to formaldehyde after 3 days in culture 4-207
Figure 4-9. Effect of formaldehyde exposure on cell proliferation of the respiratory
mucosaof rats and mice 4-211
Figure 4-10. Diagram of nasal passages, showing section levels chosen for
morphometry and autoradiography in male rhesus monkeys exposed
to formaldehyde 4-233
Figure 4-11. Formaldehyde-induced cell proliferation in male rhesus monkeys
exposed to formaldehyde 4-234
Figure 4-12. Formaldehyde-induced lesions in male rhesus monkeys exposed to
formaldehyde 4-235
Figure 4-13. Frequency and location by cross-section level of squamous metaplasia
in the nasal cavity of F344 rats exposed to formaldehyde via inhalation 4-273
Figure 4-14. Mortality corrected cumulative incidences of nasal carcinomas in the
indicated exposure groups 4-291
Figure 4-15. Alveolar MP Fc-mediated phagocytosis from mice exposed to 5 ppm
formaldehyde, 10 mg/m3 carbon black, or both 4-308
This document is a draft for review purposes only and does not constitute Agency policy.
II-xv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 4-16. Compressed air in milliliters as parameter for airway obstruction following
formaldehyde exposure in guinea pigs after OVA sensitization and OVA
challenge 4-320
Figure 4-17. OVA-specific IgGl (IB) in formaldehyde-treated sensitized guinea pigs
prior to OVA challenge 4-320
Figure 4-18. Anti-OVA liters in female Balb/C mice exposed to 6.63 ppm
formaldehyde for 10 consecutive days or once a week for 7 weeks 4-321
Figure 4-19. Vascular permeability in the trachea and bronchi of male Wistar rats
after 10 minutes of formaldehyde inhalation 4-323
Figure 4-20. Effect of select receptor antagonists on formaldehyde-induced vascular
permeability in the trachea and bronchi male of Wistar rats 4-324
Figure 4-21. The effects of formaldehyde inhalation exposures on eosinophil
infiltration and goblet cell proliferation after Der f challenge in the
nasal mucosa of male ICRmice after sensitization and challenge 4-326
Figure 4-22. NGF in BAL fluid from formaldehyde-exposed female C3Ft/He mice
with and without OVA sensitization 4-329
Figure 4-23. Plasma substance P levels in formaldehyde-exposed female C3H/He
mice with and without OVA sensitization 4-330
Figure 4-24. Motor activity in male and female rats 2 hours after exposure to
formaldehyde expressed as mean number of crossed quadrants ± SEM 4-341
Figure 4-25. Habituation of motor activity was observed in control rats during the
second observation period 4-342
Figure 4-26. Motor activity was reduced in male and female LEW. IK rats 2 hours
after termination of 10-minute formaldehyde exposure 4-343
Figure 4-27. The effects of the acute formaldehyde exposures on the ambulatory and
vertical components of SLMA 4-345
Figure 4-28. Effects of formaldehyde exposure on the error rate of female LEW. IK
rats performing the water labyrinth learning test 4-349
Figure 4-29. Basal and stress-induced trunk blood corticosterone levels in male
LEW. IK rats after formaldehyde inhalation exposures 4-354
Figure 4-30. NGF production in the brains of formaldehyde-exposed mice 4-359
This document is a draft for review purposes only and does not constitute Agency policy.
II-xvi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 4-31. DNA-protein cross-links (DPX) and thymidine kinase (tk) mutants in
TK6 human lymphoblasts exposed to formaldehyde for 2 hours 4-434
Figure 4-32. Developmental origins for cancers of the lymphohematopoietic system 4-491
Figure 4-33. Unscheduled deaths in female F344 rats exposed to formaldehyde for
24 months 4-508
Figure 4-34. Cumulative incidence of tumor bearing animals for lymphoma in female
B6C3F1 mice exposed to formaldehyde for 24 months (p < 0.05) 4-509
Figure 4-35. Effect of various doses of formaldehyde on cell number in (A) HT-29
human colon carcinoma cells and in (B) human umbilical vein epithelial
cells (HUVEC) 4-517
Figure 4-36. Integrated MOA scheme for respiratory tract tumors 4-529
Figure 4-37. Location of intraepithelial lymphocytes along side epithelial cells in the
human adenoid 4-532
This document is a draft for review purposes only and does not constitute Agency policy.
II-xvii DRAFT—DO NOT CITE OR QUOTE
-------
This page intentionally left blank.
This document is a draft for review purposes only and does not constitute Agency policy.
II-xviii DRAFT—DO NOT CITE OR QUOTE
-------
1 4. HAZARD CHARACTERIZATION
2
3
4 4.1. HUMAN STUDIES
5 This chapter discusses epidemiologic studies of site-specific cancers and other adverse
6 health effects that may be caused by exposure to formaldehyde. The primary focus is on the
7 literature describing inhalation exposure and its potential carcinogenic and noncarcinogenic
8 health risks. In addition, oral, dermal, and ocular exposures to formaldehyde are discussed.
9 The noncancer health effects section is organized by endpoint, beginning with sensory
10 irritation (SI) and followed by pulmonary function, asthma, respiratory tract pathology,
11 immunologic responses, neurological and behavioral responses, and, finally, developmental and
12 reproductive outcomes.
13 The carcinogenicity section is divided into two parts, respiratory tract and nonrespiratory
14 tract cancers. The first part discusses site-specific cancers that are chiefly located in the
15 respiratory tract where direct contact with formaldehyde occurs: nasopharyngeal cancers (NPCs),
16 nasal and paranasal cancers, other respiratory tract cancers, and lung cancers. The second part
17 on nonrespiratory tract cancer discusses those cancers at other sites with more distant exposure to
18 formaldehyde than respiratory epithelium—mainly, lymphohematopoietic (LHP) cancer, brain
19 and central nervous system (CNS) cancer, pancreatic cancer, and cancer at other sites.
20
21 4.1.1. Noncancer Health Effects
22 4.1.1.1. Sensory Irritation (Eye, Nose, Throat Irritation)
23 As a reactive gas, formaldehyde is a sensory irritant. Sensory irritation of the eyes and
24 respiratory tract by formaldehyde has been observed consistently in clinical and epidemiologic
25 studies in residential and occupational populations. Binding to sensory nerves at the portal of
26 entry (POE) results in direct sensory responses (e.g., detection of odor and tissue irritation) as
27 well as reflex responses to the sensory irritation and neurogenic sensitization. Reflex responses
28 result from CNS stimulation by the afferent sensory signals and include lacrimation, coughing,
29 sneezing, and bronchial constriction (BC). An additional reflex seen in rodents is reflex
30 bradypnea (RB) (also known as reflex apnea [RA]). Formaldehyde-induced sensory irritation
31 may be evident after acute exposures at average concentrations of 730 ppb (Kriebel et al., 1993),
32 as well as in chronically exposed individuals at lower concentrations (100-300 ppb) (Ritchie and
33 Lehnen, 1987). Formaldehyde-induced neurogenic sensitization and atopy may result in lifelong
34 health effects from short-term or transient exposures. For this discussion, sensory irritation will
35 include both direct sensory response to formaldehyde exposure and reflex responses
36 (lacrimation, coughing, sneezing, RB, and BC, and sensitization.
This document is a draft for review purposes only and does not constitute Agency policy.
4-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 Eye, nose, and throat irritation in response to formaldehyde inhalation exposure is well
2 documented (Doty et al., 2004). Broadly, studies examining these endpoints are either controlled
3 chamber studies with a defined population (e.g., healthy volunteers or sensitive individuals),
4 worker/student studies, or population (e.g., residential) studies. Chamber studies, by design, are
5 acute studies, although some researchers have investigated repeated exposures. Occupational,
6 student, and residential exposures are generally longer duration, although there is variability in
7 exposure and duration among subjects. Endpoints include both local effects and reflex effects of
8 sensory irritation. The endpoints for assessing irritation include self-reporting of adverse
9 symptoms (e.g., pain, burning, itching) as well as objective measures of irritation (e.g., eye-blink
10 counts, lacrimation) (Doty et al., 2004). The following review focuses on eye, nose, and throat
11 irritation but studies have documented other types of irritation, including dermal irritation
12 eczema and dermatitis.
13
14 4.1.1.1.1. Epidemiologic literature.
15 A wide variety of epidemiologic studies have assessed the potential effects of exposure to
16 formaldehyde on endpoints, indicating sensory irritation of the eye, nose, and throat. These
17 studies generally include three different types of exposure populations: (1) Residents and visitors
18 exposed to formaldehyde in homes and mobile buildings, where formaldehyde is present from
19 various sources, including building components, furniture and home furnishings, heating and
20 cooking combustion as well as active and passive smoking; (2) various occupational exposures
21 from industrial processes related to wood products, furniture making, and formaldehyde-based
22 resins; and (3) anatomy students who are exposed under well-defined conditions during
23 academic courses where they are examining formaldehyde-preserved cadavers.
24
25 4.1.1.1.1.1. Residential epidemiology.
26 Among the residential epidemiology studies of formaldehyde effects on sensory
27 irritation, one of the strongest studies based on study design, execution, analysis, and sample size
28 was the observational study undertaken by Ritchie and Lehnen (1987). In this cross-sectional
29 study of nearly 2,000 Minnesota residents living in 397 mobile and 494 conventional homes,
30 personal data and formaldehyde samples were collected from residents that had responded to an
31 offer by the state health department to test homes for formaldehyde. Technicians administered a
32 symptom questionnaire to participating residents at the time of formaldehyde sample collection.
33 Residents were asked to close doors and windows of their homes for 12 hours before testing was
34 conducted, and a standardized collection protocol was used for both sample collection and
35 analysis. Measurements of formaldehyde exposure were taken from two rooms of the home,
This document is a draft for review purposes only and does not constitute Agency policy.
4-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 usually the bedroom and living room, and the samples were kept refrigerated until analysis.
2 Respondents were not aware of the results of the formaldehyde analyses in their homes at the
3 time they responded to the symptom questionnaire. The results from Ritchie and Lehnen (1987)
4 provide a clear dose-response relationship in the percentage of residential occupants reporting
5 eye, nose, and throat irritation. Specifically, eye irritation responses increased from 1-2% in
6 homes with formaldehyde concentrations lower than 0.1 ppm to 12-32% in homes with
7 formaldehyde concentrations ranging from 0.1 to 0.3 ppm and 86-93% of residents reporting eye
8 irritation at > 0.3 ppm. These effects were found in the same concentration range for people
9 living in either mobile (n = 851) or conventional (n = 1,156) homes. Similar percentages were
10 found for nose/throat irritation. The authors reported that they controlled for smoking, age and
11 sex in logistic regression models. Thus, smoking is not likely to be a confounder of the observed
12 relation between formaldehyde and irritation. The percent reporting irritation increased with
13 increasing formaldehyde concentration category within all strata for smoking (active, passive and
14 nonsmokers) among residents in both mobile homes and conventional homes. While the
15 participants in this study were self-selected and not a random residential sample, a clear
16 concentration response was observed, and, even if participants sought testing because they
17 suspected that they were being exposed to formaldehyde in their homes, they could not know the
18 measured concentration of formaldehyde when reporting their irritation symptoms, so recall bias
19 cannot explain the concentration response.
20 The results of an adverse association of sensory irritation with formaldehyde reported by
21 Ritchie and Lehnen (1987) are corroborated by Hanrahan et al. (1984) who conducted a cross-
22 sectional survey by using a random sample of mobile homes from mobile home parks in
23 Wisconsin. Sixty-one teenage and adult residents participated. Health questionnaires were self-
24 administered by each occupant. Respondents were blinded to the results of their home
25 formaldehyde vapor measurements, which were sampled from two rooms in the homes following
26 instruction to close windows, refrain from smoking, and turn off gas appliances for 30 minutes
27 prior to air sampling. Logistic regression analyses were used to ascertain potential symptom risk
28 ratio dependency on each respondent's age, smoking status, gender, and formaldehyde
29 concentration measures in the home. Formaldehyde concentrations ranged from 0.1 ppm to
30 0.8 ppm with a median concentration of 0.16 ppm. Across this concentration range, a clear and
31 statistically significant concentration-response relationship was reported in graphical form,
32 controlling for age, gender, and smoking status. At 0.1 ppm, the regression model showed less
33 than 5% predicted prevalence of burning eyes. At 0.2 ppm, the midpoint of the exposure
34 category in Ritchie and Lehnen (1987) that was reported to be the lowest adverse effect level for
35 eye irritation with 12-32% reporting eye irritation, the regression model of Hanrahan et al.
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 (1984) showed approximately 17.5% predicted prevalence of burning eyes. The prevalence of
2 burning eyes rose linearly to approximately 65% prevalence at 0.5 ppm, with some diminishment
3 in the rate of rise up to approximately 80% prevalence at 0.8 ppm. While only 65 out of 208
4 randomly selected homes volunteered to complete the health questionnaires, the investigators
5 were able to complete home formaldehyde vapor measurements on all the homes and reported
6 nearly an identical distribution of formaldehyde concentrations in participating and
7 nonparticipating homes. Demographic characteristics of some of the nonrespondents were
8 available and were reported as nearly identical to those of participants. There was no indication
9 of selection bias. Confounding is unlikely to explain such a strong concentration response.
10 These findings of associations of sensory irritation with residential exposures to
11 formaldehyde are further supported by studies that did not examine concentration response but
12 nonetheless assessed the association of formaldehyde with sensory irritation. Similar findings to
13 those of Ritchie and Lehnen (1987) and Hanrahan et al. (1984) have been reported in other
14 residential studies of increased symptoms in association with formaldehyde exposure (Liu et al.,
15 1991; Thun et al., 1982; Dally et al., 1981). Dally et al. (1981) collected data in 100 "complaint
16 structures" (65% mobile homes, 27% conventional homes, 2% travel trailers, 2% office
17 buildings, etc). Of these, 60% were from home owners contacting the health department and
18 30% from physician referrals. Twenty percent of the buildings had concentrations below the
19 limit of detection (0.1 ppm), 20% had levels at or above 0.81 ppm, and overall the concentrations
20 ranged from below detection to above 3 ppm with an overall median of 0.35 ppm. The median
21 levels were 0.47 and 0.10 ppm for mobile and conventional homes, respectively. No other
22 contaminants were measured. Eye, nose, and throat irritation were reported in a high percentage
23 of occupants (eye irritation 68%, burning eyes 60%, runny nose 60%, dry or sore throat 57%,
24 cough 51%), but these were not reported as a function of dose or home type. Thus, there was no
25 control group to which rates of irritation could be compared. However, symptoms reportedly
26 stopped in 89% of occupants when they left the "complaint structure." The most recent
27 residential study was performed on over 1,000 mobile homes with 1,394 participants (Liu et al.,
28 1991). Home formaldehyde concentration ranged from below 0.01 to 0.46 ppm. Analyses used
29 logistic regression to control for potential confounders. Eye irritation was positively associated
30 with formaldehyde with a clear concentration response demonstrated with cumulative exposure.
31 During the summer and winter months, formaldehyde exposure was associated with burning
32 eyes. In the winter months, formaldehyde exposure was associated with sore throat. There was
33 no association of formaldehyde exposure with cough or running nose during either season. Liu
34 et al. (1991) also report a synergistic effect on irritation by formaldehyde exposure and chronic
35 disease prevalence. Thun et al. (1982) reported increased symptoms of itchy skin and "wheezing
This document is a draft for review purposes only and does not constitute Agency policy.
4-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 and difficulty breathing" in residents in 395 homes insulated with urea-formaldehyde foam
2 relative to nearby homes without urea foam formaldehyde insulation (UFFI); however, there
3 were no measurements of formaldehyde concentration taken in this study.
4 While not strictly a residential epidemiology study, Olsen and Dossing (1982) studied
5 occupational exposures within mobile and nonmobile daycare centers. They reported the mean
6 concentration in mobile and nonmobile day care centers were 350 (200-450) ppb and 65
7 (40-90) ppb, respectively. Adverse eye, nose, and throat irritation were significantly elevated in
8 the workers (n = 70) in the mobile units as compared with those in nonmobile units (n = 34).
9 The authors also state that a high percentage of workers in the mobile day cares reported that the
10 symptoms disappeared after working hours; however, the authors did not report any such
11 percentages among those working in nonmobile units.
12
13 4.1.1.1.1.2. Occupational epidemiology.
14 Horvath et al. (1988) compared irritation symptoms between 109 workers at a
15 particleboard manufacturing plant and 264 workers at food plants as a control group. The mean
16 8-hour time-weighted average (TWA) formaldehyde concentrations between these two groups
17 were 0.69 ppm (range 0.17-2.93) and 0.05 ppm (range 0.03-0.12), respectively. Eye, nose, and
18 throat irritation were more common among the former group (prevalence of symptoms during a
19 work shift: throat sore or burning—test 22.0%, controls 3.9%; cough—test 34.9%, controls
20 18.9%; burning of nose—test 28.4%, controls 2.0%; stuffy nose—test 33.9%, controls 14.2%;
21 itching of nose—test21.1%, controls 7.9%; eyes burning or watering—test39.5%, controls
22 9.1 %; eyes itching—test 19.3%, control s 7.1 %).
23 Similar results were reported for frequency of eye and nasal discomfort in a group of
24 workers involved in the manufacture of formaldehyde resins. These workers were exposed to a
25 mean concentration of 0.40 mg/m3. Alexandersson and Hedenstierna (1988) reported that the
26 frequency of eye, nose, and throat irritation was significantly greater in 38 workers exposed to
27 formaldehyde and solvents in lacquers (average employment duration 7.8 years) as compared
28 with 18 controls (nonexposed individuals working at the same factory). The frequency of eye
29 irritation was 65.8% among those exposed and 16.7% among controls. No controls reported
30 nose/throat irritation, but about 40% of those exposed did.
31 A Swedish study conducted at a chemical plant found that nasal and eye discomfort were
32 reported by 64 and 24%, respectively, of workers (n = 70) exposed to formaldehyde (range
33 0.05-0.50 mg/m3 with a mean of 0.26 mg/m3) versus 25 and 6%, respectively, of the control
34 group made up of clerks from the local government (n = 36). In addition, the majority of
35 workers exposed to formaldehyde reported that their symptoms were relieved during weekends
This document is a draft for review purposes only and does not constitute Agency policy.
4-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 and vacations (Holmstrom and Wilhelmsson, 1988). Another study by the same authors
2 (Wilhelmsson and Holmstrom, 1992) reported similar results. In this study irritation prevalent
3 among 66 workers from a formaldehyde-producing plant was compared with that seen among
4 36 community clerks. The workers were exposed to 0.26 mg/m3 of formaldehyde (range
5 0.05-0.6 mg/m3). The clerks were exposed to an average of 0.09 mg/m3. Nasal and eye
6 discomfort were reported at rates of 53 and 24%, respectively, among the workers. Among the
7 community clerks, 3 and 6%, respectively, reported discomfort.
8 Holness and Nethercott (1989) reported significant increases in eye irritation (42 vs.
9 21%) and nose irritation (44 versus 16%) among 84 funeral service workers as compared with
10 38 controls (students and individuals from a service organization). The former group had been
11 actively embalming for approximately 10 years and had nearly twice the pack-years smoked as
12 the controls. The exposure concentration in both groups was 0.36 and 0.02 ppm, respectively.
13
14 4.1.1.1.1.3. Epidemiology on laboratory students.
15 Several studies have monitored sensory irritation in medical/physical therapy students
16 exposed to formaldehyde during anatomy courses. These studies have particular advantages: the
17 student population generally has no former occupational exposure, and, oftentimes, preclass
18 survey data serve as the control, providing a better basis for assessing the effects of
19 formaldehyde exposure.
20 In a study of 24 formaldehyde-exposed anatomy students (personal breathing zone
21 samples 0.73 ppm, range 0.49-0.93), the prevalence of eye irritation before the start of a cadaver
22 dissection class was 16%, while after the class, the prevalence was 59%. The increase in
23 prevalence of eye irritation was most pronounced (43%), but increases were also observed in the
24 prevalence of irritation of the nose (21%) and throat (15%) (Kriebel et al., 1993). The authors
25 also reported a tendency for this increase in intensity between the beginning and end of class to
26 diminish over the 10-week course, especially for eye irritation. However, although the intensity
27 of the irritation diminished, eye irritation was still present among the students after 10 weeks of
28 intermittent exposure. The report of increase in post- versus preclass irritation symptoms in this
29 study was no greater for asthmatic students (n = 5) compared with nonasthmatic students.
30 Takahashi et al. (2007) showed that 143 medical students reported various symptoms
31 (including eye and throat irritation) and that the percentage of students reporting symptoms
32 increased between the beginning and end of the course 2 months later. After the first day of
33 class, approximately 35% of students reported eye soreness and about 15% reported throat
34 irritation. After the course ended, these rates were close to 70% for eye soreness and slightly
35 above 40% for throat irritation. The reported average room formaldehyde concentration was
This document is a draft for review purposes only and does not constitute Agency policy.
4-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 2.12 ppm (range 1.7-2.4), while the gas samplers worn on the students' chests averaged 2.4 ppm
2 (range 1.8-3.8). Another study of students in an anatomy laboratory class in Japan (Takigawa et
3 al., 2005) measured formaldehyde concentrations and irritation symptoms before and after the
4 installation of a ventilation system. This system reduced the median personal formaldehyde
5 exposure concentration from 2.7 to 0.72 ppm. Before installation of the ventilation system, the
6 students complained about exacerbation of all the sensory irritation symptoms on average. The
7 increase in 8 out of 25 symptoms was significantly reduced after installing general ventilation
8 (p< 0.05). After installation of the ventilation system, a dose-dependent relationship with
9 formaldehyde was seen for irritated eyes but not for itchy nose.
10 Akbar-Khanzadeh et al. (1994) detected mean personal area levels of formaldehyde at
11 1.24 ppm and a range of 0.1-2.94 ppm from personal air sampling devices. Almost 90% of the
12 students in this study reported eye irritation, 74% reported nose irritation, and close to 30%
13 reported throat irritation during or after exposure to formaldehyde during the laboratory period
14 after having completed at least 6 weeks of laboratory sessions with formaldehyde exposure. In
15 addition, Uba et al. (1989) demonstrated that symptoms of eye, nose, and throat irritation were
16 correlated with formaldehyde exposure among medical students by comparing students'
17 responses on a questionnaire completed after a lab with formaldehyde exposure to a
18 questionnaire completed after a lab with no formaldehyde exposure. The authors compared
19 questionnaires completed prior to students' first anatomy lab to a questionnaire completed
20 7 months later. Reports of cough were more frequent after the 7 months. These students were
21 exposed to a mean level of 1.9 ppm (range 0.1-5.0) while dissecting (measured using portable
22 infrared spectrophotometer), and a TWA from all laboratory activities ranged from below limits
23 of detection to 0.93 (measured using personal sampling devices in the students' breathing zones).
24
25 4.1.1.1.2. Acute studies: controlled chamber exposures.
26 Results from controlled human studies demonstrate eye, nose, and throat irritation in
27 association with formaldehyde exposure (Lang et al., 2008; Yang et al., 2001; Krakowiak et al.,
28 1998; Kulle, 1993; Green et al., 1989, 1987; Kulle et al., 1987; Sauder et al., 1987, 1986;
29 Schachter et al., 1987, 1986; Witek et al., 1987; Day et al., 1984; Bender et al., 1983; Anderson
30 et al., 1983; Weber-Tschopp et al., 1977; Andersen, 1979; Schuck et al., 1966). A key advantage
31 of chamber studies is the ability to monitor and closely control formaldehyde concentrations
32 during exposure. However, chamber studies may also be limited by other aspects of the study
33 design, including small number of participants, use of healthy volunteers, short exposure
34 durations (a few minutes), and often studies were conducted with only one exposure group and at
35 relatively high concentrations (>1 ppm). The lack of multiple exposure levels in many studies
This document is a draft for review purposes only and does not constitute Agency policy.
4-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 limits the understanding of exposure-response relationships. Additionally, numerous reports that
2 demonstrate multiple symptoms of eye, nose, and throat irritation at levels at or above 1 ppm did
3 not explore lower levels of exposure and can only be used for primary hazard identification
4 (Yang et al., 2001; Green et al., 1989, 1987; Sauder et al., 1987, 1986; Schachter et al., 1987,
5 1986; Witek et al., 1987; Day et al., 1984).
6 The National Aeronautical and Space Administration conducted experiments in closed-
7 environment living, including environmental monitoring and air quality. James et al. (2002)
8 quantified air pollutants, including formaldehyde, during 30, 60, and 90-day tests in a closed
9 chamber study of a Lunar-Mars life support chamber. Unfortunately, the detection methods used
10 during the 30-day test were not sensitive enough to detect formaldehyde at levels below 2
11 mg/m3. Thus, badge samples were obtained in the 60-day and 90-day tests and provided greater
12 detection sensitivity (to 0.02 mg/m3). Measured values of formaldehyde increased over time. In
13 the 60-day test, formaldehyde levels were well above accepted limits (data not shown). Health
14 effects data are limited since there were only four crew members. One crew member reported
15 eye and upper airway irritation at formaldehyde concentrations of 0.25 mg/m3 (203 ppb) on day
16 15. It should also be noted that astronauts are exceptionally healthy individuals, and these data
17 should be interpreted carefully when determining expected health effects in the general
18 population. The experimenters determined that formaldehyde levels increased as temperature
19 increased. Formaldehyde was also linked to murals lining the chamber and was subsequently
20 removed before executing the 90-day study. Between days 0 and 60, formaldehyde levels
21 remained between 0.02 and 0.04 mg/m3, with one sharp peak that occurred at day three to 0.07
22 mg/m3. Between days 60 and 90, formaldehyde concentrations increased to 0.07 to 0.09 mg/m3.
23 The increase was attributed to an incomplete oxidation of methanol in a catalytic bed rather than
24 in excessive off-gassing of formaldehyde. No crew members reported any adverse effects in the
25 90-day study.
26 A few studies have been conducted that specifically address sensitive populations
27 (asthmatics) and/or individuals during exercise, which can exacerbate asthma (further details of
28 these studies are in Section 4.1.1.3, Effects on Asthmatics). In Sauder et al. (1986), 8-minute
29 bicycle exercise was completed multiple times during the exposure period (3 hours). However,
30 irritation symptoms were only reported after 2 hours of exposure and do not address whether
31 changes occurred during the periods of exercise. Overall, reports of eye, nose, and throat
32 irritation increased with exposure to formaldehyde (3 ppm) compared with reports of irritation
33 with no exposure to formaldehyde. Green et al. (1987) report that eye, nose, and throat irritation
34 symptoms were greater immediately after exercise during exposure to 3 ppm formaldehyde.
35 Additionally, the response levels were similar between asthmatic (n = 16) and nonasthmatic
This document is a draft for review purposes only and does not constitute Agency policy.
4-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 (n = 22) subjects. Similar effects of exercise on certain symptoms, such as throat irritation, were
2 reported in 15 asthmatic subjects exposed to 2 ppm formaldehyde at rest and after exercise
3 (Witeketal., 1987).
4 Kulle (1993) and Kulle et al. (1987) enrolled 19 healthy volunteers and exposed them to a
5 range of formaldehyde concentrations. At 2 ppm, 53% reported mild or moderate eye irritation
6 (32% mild, 21% moderate). At 3 ppm, 100% of subjects exposed at this level (n = 9) reported
7 irritation. The reported increase in irritation was shown to correspond with increasing
8 formaldehyde concentration in a linear fashion. Mild nose/throat irritation was present among
9 37% of those exposed to 2 ppm of formaldehyde. Odor detection was very similar to the
10 distribution seen for eye irritation. Nineteen subjects performed light to moderate exercise while
11 exposed to 2 ppm; there was no increase in report of eye irritation, but nose/throat irritation did
12 increase. The data were reanalyzed (Kulle, 1993), and thresholds for irritation were found to be
13 0.5-1 ppm for eye irritation and 1 ppm for nose/throat irritation.
14 Yang et al. (2001) reported that eight individuals exposed to varying levels of
15 formaldehyde (1.65, 2.99, and 4.31 ppm) had mild to moderate eye irritation during the 5-minute
16 exposures. The increase in irritation was detected at 30 seconds with exposure to 1.65 ppm of
17 formaldehyde. The highest severity ratings at this concentration occurred between 60 and
18 90 seconds. Frequency of eye blinking was also measured. The peak in blinking rate occurred
19 after about 1 minute of exposure and then decreased almost back to a normal rate after 5 minutes
20 of exposure. Higher formaldehyde concentrations were associated with increased frequency of
21 blinking compared with the 1.65 ppm exposure.
22 Other studies have examined responses across multiple exposure levels. For example,
23 Weber-Tschopp et al. (1977) used two different methods of studying irritation resulting from
24 formaldehyde exposure. For one, they exposed subjects (n = 33) to an increasing level of
25 formaldehyde (maximum exposure was 3.2 ppm). This design precluded evaluation of distinct
26 effects at different exposure levels. The researchers addressed this by examining another group
27 of subjects (n = 48) that were exposed to 0, 1, 2, 3, or 4 ppm five times for 90 seconds. Levels of
28 nasal and throat irritation for this discontinuous exposure were slightly higher than the irritation
29 levels reported among those with continuous exposure. However, this was reversed for eye
30 irritation; those with continuous exposure reported higher levels of irritation than those with
31 discrete exposures. An objective measure, eye-blinking rate, was measured for those with
32 continuous exposure and was found to have a statistically significant increase at 1.7 ppm.
33 Bender et al. (1983) conducted a study that enrolled individuals who "responded" to
34 formaldehyde at 1.3 and 2.2 ppm and did not report irritation to the clean air control. They
35 found that, among these subjects, exposure to 1 ppm of formaldehyde (n = 27) resulted in the
This document is a draft for review purposes only and does not constitute Agency policy.
4-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 reporting of eye irritation with a median response time of 78 seconds. Reports of irritation were
2 given as less than slightly irritating for formaldehyde concentrations of 0.3-0.9 ppm.
3 Assessment of sensory irritation for pain and discomfort often relies on self-reporting,
4 using symptom questionnaires and severity ratings (e.g., mild, moderate, severe). In the case of
5 formaldehyde, subjective ratings of eye irritation correlate positively with eye-blinking
6 frequency (Lang et al., 2008). Lang et al. (2008) saw an increase in eye blinking after
7 195 minutes of exposure to formaldehyde at 0.5 ppm with four peak exposures of 1 ppm. After
8 this amount of time and formaldehyde exposure, there was also an increase in moderate eye
9 redness. Weber-Tschopp et al. (1977) reported that, among concentrations ranging from 0.03 to
10 3.2 ppm, eye-blinking frequency was increased at 1.7 ppm; similarly Yang et al. (2001) reported
11 increased blinking at >1.5 ppm (the lowest concentration examined). There are studies that
12 suggest that psychological factors (e.g., anxiety) can impact the perception of irritation—and
13 perhaps more so at lower concentrations (Lang et al., 2008; Ihrig et al., 2006; Dalton, 2003).
14 However, when Lang et al. (2008) controlled for mood prior to exposure, subjective symptoms
15 of eye, nasal, and olfactory irritation were significantly related to exposure (0.5 ppm)
16 Schuck et al. (1966) performed a study that also examines self-reported eye irritation as
17 well as blinking rate. Fourteen individuals were exposed to formaldehyde concentrations
18 ranging from 0 to 1 ppm. Increased irritation was reported with increasing formaldehyde
19 concentration. One subject, judged to be the least sensitive, was still able to detect formaldehyde
20 levels as low as 0.01 ppm. In addition, the authors examined the blinking rate of participants,
21 which they found was related to irritation intensity.
22 Andersen (1979) and Anderson and M01have (1983) reported on a controlled experiment
23 in which 16 individuals were exposed to varying levels of formaldehyde for five hours and rated
24 their level of discomfort over the exposure period. Discomfort occurred within 1 hour at
25 formaldehyde exposure levels of 1 and 2 mg/m3 (Andersen and M01have, 1983) After 2 hours,
26 increasing discomfort was reported among the groups exposed to 0.3 and 0.5 mg/m3. Subject
27 reported that discomfort was mainly conjunctival irritation and dryness in the nose and throat.
28 Subjects complained at all four concentrations of formaldehyde: 0.3, 0.5, 1.0, and 2.0 mg/m3 and
29 of 16 subjects, 3, 5, 15, and 15 subjects complained at each respective exposure concentration
30 (Andersen and M01have, 1983).
31 Controlled chamber studies have also been conducted on various populations of
32 previously exposed individuals to determine if formaldehyde exposure potentiates an
33 individual's response to acute exposures. Schachter et al. (1987) reported on 15 laboratory
34 workers "frequently exposed to formaldehyde" (no quantification of exposure is given; however,
35 the workers report being exposed for 1 to 7 days per week from a range of 1 to 21 years). Tests
This document is a draft for review purposes only and does not constitute Agency policy.
4-10 DRAFT—DO NOT CITE OR QUOTE
-------
1 performed at the start of the study found that these individuals had pulmonary function similar to
2 that seen in healthy individuals. The workers in this study reported subjective measures of eye,
3 nose, and throat irritation after 40 minutes of exposure to 2 ppm of formaldehyde. However, the
4 2 ppm acute exposure in this study may be sufficiently high to induce significant irritation in
5 most individuals. Krakowiak et al. (1998) reported that 10 asthmatics with occupational
6 exposure to formaldehyde (via formaldehyde solutions or pure gaseous formaldehyde) exhibited
7 similar symptom scores to healthy controls (never exposed to formaldehyde in the workplace)
8 exposed to 0.4 ppm formaldehyde for 2 hours. The mean symptom scores and standard
9 deviation (SD), which included information on sneezing, rhinorrhea, mucosal edema, and
10 itching, were 4.6 ±1.6 (mean ± SD) for asthmatics and 4.3 ±1.2 for healthy subjects
11 immediately after inhalation. These dropped to 1.8 ± 1.2 and 1.2 ± 1.3, respectively, 4 hours
12 after the exposure. It is unclear if sensitive individuals may not be represented in either of these
13 groups, as the workers were tolerating their exposures during the work shift "healthy worker"
14 effect. However, residents (n = 9) exposed to formaldehyde in their homes, who complained
15 about adverse effects from the material, but with no occupational exposure reported eye, nose,
16 and throat irritation at a similar rate as controls (individuals in homes without formaldehyde or
17 individuals in homes with formaldehyde but not reporting adverse effects [n = 9]) after a
18 90-minute exposure to 1 ppm (Day et al., 1984). The number of individuals reporting eye
19 irritation, nasal congestion, and throat irritation were seven, three, and two among sensitive
20 individuals and eight, four, and three among controls, respectively. These individuals may be
21 considered a sensitive population since they had "previously complained of various
22 nonrespiratory effects from the UFFI in their homes" (household concentrations unknown).
23
24 4.1.1.2. Pulmonary Function
25 Pulmonary function is assessed using spirometry which measures the volume and speed
26 of air that is exhaled or inhaled. Multiple parameters can be measured during spirometric
27 testing. Forced vital capacity (FVC) measures the volume of air that can be exhaled. This
28 volume can be partitioned into the volume that is exhaled in the first second (FEVi) or that
29 which is exhaled during the middle of a breath between the 25th and 75th percentiles called the
30 forced expiratory flow 25-75% (FEF25-7s) and is also called the maximal mid-expiratory flow
31 (MMEF). Other common metrics of lung function are the ratio of FEVi to FVC (FEVi/FVC)
32 and the peak expiratory flow rate (PEF or PEFR). Spirometric results can be important
33 diagnostic criteria for physician-diagnosed asthma. Changes in lung function are an important
34 health endpoint with potentially long-term consequences. The observed consequences of early
This document is a draft for review purposes only and does not constitute Agency policy.
4-11 DRAFT—DO NOT CITE OR QUOTE
-------
1 life exposure to adverse levels of air pollutants include diminished lung function, increased
2 susceptibility to acute respiratory illness and asthma (Bateson and Schwartz, 2008).
3 Absolute values for lung function parameters are likely to vary by gender, age, height,
4 and smoking status and are best compared when normalized to the expected lung function based
5 on these variables (Schoenberg et al., 1978). Therefore, these well-known predictors of lung
6 function should be controlled for in an evaluation of spirometric data. Individual variation can
7 also be addressed by each subject serving as his/her control with measurements taken before,
8 during, and after exposure. Analysis of the percent change in various parameters in this context
9 may have greater sensitivity to detect exposure-related changes in function. In addition to
10 individual variation in baseline lung function, there is also individual variation in bronchial
11 responsiveness. Reduced lung function parameters in response to methacholine challenge is a
12 standard test for bronchial constriction, and this can be used to define responsive, sensitive, or
13 susceptible individuals. Since formaldehyde-induced bronchial constriction is measured with
14 these lung function tests, variability in bronchial responsiveness may impact interpretation of
15 formaldehyde-induced changes. Depending upon the proportion of susceptible individuals in a
16 study population, the group-mean change in lung function parameters may or may not reflect any
17 effect of exposure. Studies that exclude sensitive or responsive individuals may not detect
18 changes in lung function. Studies based on random population samples may include some
19 sensitive individuals who respond to exposures with large changes in lung function parameters
20 that may be difficult to detect if only the group mean change in lung function is examined.
21 Experiments that report individual-level changes in lung function parameters or that focus on
22 sensitive individuals can help address this question. An additional complication in the
23 interpretation of pulmonary function experiments is that 'sensitivity' may be specific to timing of
24 exposure in relation to the potential allergen and the individuals' atopic status.
25 Formaldehyde may itself be an allergen or it may potentiate the ability of other allergens
26 to cause atopic switching or increase the sensitivity of atopic individuals. Thus formaldehyde
27 exposure among nonatopic individuals could theoretically cause atopic switching in the presence
28 or absence of allergens possibly resulting in a diagnosis of asthma. Formaldehyde could also
29 cause an asthma attack or potentiate the influence of other stimuli on the risk of asthma attacks.
30 Demonstration of a clear association of formaldehyde exposure, or the lack of an association, at
31 one particular time prior to or following the onset of asthma does not necessarily imply that
32 exposure to formaldehyde is causing or not causing adverse outcomes at other times.
33 Workers chronically exposed to formaldehyde have exhibited signs of reduced lung
34 function consistent with bronchial constriction, inflammation, or chronic obstructive lung
35 disease. Worker exposures that report cross-shift differences in spirometric values are consistent
This document is a draft for review purposes only and does not constitute Agency policy.
4-12 DRAFT—DO NOT CITE OR QUOTE
-------
1 with formaldehyde-induced sensory irritation. Additionally, concordance has been reported
2 between subjective irritant response and measured changes in pulmonary function, further
3 supporting the possibility that cross-shift and short-term evidence of bronchial constriction may
4 be a reflexive response to sensory irritation.
5 In occupational studies of formaldehyde exposure, lung function deficits have been
6 reported both in preshift versus postshift measurements and as a result of long-term exposures
7 (Pourmahabadian et al., 2006; Herbert et al., 1994; Malaka and Kodama, 1990; Alexandersson
8 and Hedenstierna, 1989; Alexandersson et al., 1982). Decreases in spirometric values, including
9 peak expiratory flow, vital capacity, forced expiratory volume, forced vital capacity, and
10 FEV/FVC have been reported. Studies of long-term exposure also report increased respiratory
11 symptoms, such as cough, increased phlegm, asthma, chest tightness, and chest colds, in exposed
12 workers (Pourmahabadian et al., 2006; Herbert et al., 1994; Malaka and Kodama, 1990;
13 Alexandersson and Hedenstierna, 1989; Alexandersson et al., 1982). Similar findings have been
14 reported for low-level residential formaldehyde exposure, including decreased PEF rates in
15 children (Krzyzanowski et al., 1990).
16
17 4.1.1.2.1. Epidemiologic literature.
18 The potential adverse effects of formaldehyde exposure on pulmonary function in
19 humans can be examined on several time scales of interest. The epidemiologic literature
20 supports the assessment of exposures among exposed anatomy medical students where all
21 participants have well-defined and similar duration of exposure (i.e., a semester-long class)
22 (Kriebel et al., 2001, 1993; Akbar-Khanzadeh and Mlynek, 1997; Akbar-Khanzadeh et al., 1994;
23 Uba et al., 1989; Fleisher, 1987); among individuals living or working in buildings with
24 formaldehyde exposure (Franklin et al., 2000; Krzyzanowski et al., 1990; Main and Hogan,
25 1983); and among workers (industrial, manufacturing, mortuary, hospital staff, etc.) (Ostojic et
26 al., 2006; Herbert et al., 1994; Khamgaonkar and Fulare, 1991; Malaka and Kodama, 1990;
27 Nunn et al., 1990; Alexandersson and Hedenstierna, 1989; Holness and Nethercott, 1989;
28 Holmstrom and Wilhelmsson, 1988; Horvath et al., 1988; Kilburn et al., 1985; Alexandersson et
29 al., 1982). These studies are summarized in Table 4-1.
30 Studies of anatomy students provided information about acute effects related to
31 formaldehyde exposures experienced in the laboratory (Kriebel et al., 2001, 1993) as well as
32 special insight into the intermediate stages of possible sensitization (Kriebel et al., 1993).
33 Kriebel and colleagues (1993) examined the prelaboratory and postlaboratory PEF using Mini-
34 Wright peak flowmeters in 24 students attending 3-hour anatomy classes to dissect cadavers
35 once per week over 10 weeks. Formaldehyde concentrations collected in the breathing zone
This document is a draft for review purposes only and does not constitute Agency policy.
4-13 DRAFT—DO NOT CITE OR QUOTE
-------
i
& s
II
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function
Reference
Ubaetal., 1989
Study design
Panel study of symptoms and
respiratory function among 103
medical students related to
formaldehyde exposure during a
7 month anatomy class meeting
twice a week for 4 hours
(September 1984-April 1985).
Pre- and postlab spirometric
measures were taken before the
class began, after the first 2
weeks, and after 7 months.
Complete data was available for
96 students. Cross-shift change
in pulmonary function analyzed
using repeated measures
ANOVA.
Exposure assessment
Personal sampling monitors
(impingers) in the breathing
zone measured time-
weighted average
formaldehyde exposure
during the gross anatomy
laboratory. A total of 32
samples were taken during 7
months. Short-term samples
were taken (N = 16) for peak
concentrations using a
portable infrared
spectrophotometer.
formaldehyde ranged from
below LOD (0.05 ppm) to
0.93 ppm for TWA and 0. 1
to 5.0 ppm during dissection.
Concentrations declined over
the 7 months.
Results, effect estimate, statistical significance
Test Day 1
Test Day 2
Test Day 3
Test Day 1
Test Day 2
Prelab
FVC (L)
FEVj (L)
FEF25_75 (L/sec)
FEVj/FVC
FVC (L)
FEVj (L)
FEF25_75 (L/sec)
FEVj/FVC
FVC (L)
FEVj (L)
FEF25.75 (L/sec)
FEVj/FVC
Cross-Lab Change
FVC (L)
FEVj (L)
FEF25_75 (L/sec)
FEVj/FVC
FVC (L)
FEVj (L)
FEF25.75 (L/sec)
FEVj/FVC
Value (SD)
5.246 (1.025)
4.379 (0.846)
4.492(1.216)
0.835
5.277 (1.027)
4.409 (0.824)
4.484(1.151)
0.836
5.308 (1.027)
4.399 (0.823)
4.392(1.198)
0.829
Mean change (%)
-0.012 (-0.23)
-0.031 (-0.71)
-0.079 (-1.76)
-0.004 (-0.48)
-0.0423 (-0.80)
-0.046b (-1.04)
-0.089 (-1.99)
-0.003 (-0.36)
s ,
fe
H
o
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Ubaetal., 1989
(continued)
Kriebel et al.
(1993)
Kriebel etal..
(2001)
Study design
Panel study of symptoms and
respiratory function related to
formaldehyde exposure among
24 clinical anatomy students
during a 10 week anatomy class
meeting once a week for 3
hours. PEF measured by trained
students pre- and postlab and 1-
3 times during lab using Mini-
Wright peak flowmeters. Mean
prelab and cross-lab change in
pulmonary function analyzed
using random effects models.
Panel study of symptoms and
respiratory function and
formaldehyde exposure among
5 1 gross anatomy students
during a 12 week class meeting
once per week for 2.5 hours.
Pre- and postlab measurements
obtained for at least one week
for 38. Individual pre- and
Exposure assessment
Personal samples in the
breathing zone formaldehyde
Concentrations ranged from
0.49-0.93 ppm, 8 samples.
No trend in concentrations
over semester.
Formaldehyde concentrations
were monitored continuously
in six homogenous sampling
zones in the lab (LOD = 0.05
ppm). Work location every
12 minutes was recorded and
12 minute work-zone
concentrations were
calculated for each student.
Results, effect estimate, statistical significance
Test Day 3
Weeks 1-2
Weeks 9-10a
Weeks 24-25
FVC (L)
FEVj (L)
FEF25.75 (L/sec)
FEVj/FVC
-0.0423 (-0.79)
-0.021C(-0.48)
0.003 (0.07)
0.002 (0.24)
aDay 2&3 vs Day l,p< 0.001
bDay2vsDay I,/? = 0.03)
cDay 3 vs Day I/? = 0.01)
Prelab
PEF (L/min)
PEF (L/min)
PEF (L/min)
Mean(SD)(N=20)
538.9 (86.9)
529.4 (88.4)
536.6 (86.2)
GLM of prelab decrement over 10 week courseb
B = -2.7± 1.1 L/min per week; ^ = 0.01
aEnd of course
bModel included asthma, asthma*week, eye symptoms, nose
symptoms
Weeks 1-2
Recent exposure
Recent exposure
*ln(wk)
Past exposure
Cold on lab day
Cross-lab change
PEF (L/min)
Change (% of
prelab)
-10.8 (-2.0)
PEF as fraction of baseline (before 1st lab)
B(SE)
-1.05 (0.33)
0.69 (0.24)
-0.52 (0.30)
-1.67(0.41)
p value
0.002
0.004
0.08
0.001
o
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Kriebeletal..
(2001)
(continued)
Akbar-
Khanzadeh et al.
(1994)
Study design
postlab data analyzed together
in generalized estimating
equations.
Comparison of pulmonary
function (spirometry) among 34
nonsmoking exposed medical
students and instructors and 12
nonmedical unexposed students
before and after their work in
the gross anatomy laboratory or
at predetermined times for the
unexposed (approximately 3
hours) over five consecutive
weeks.
Exposure assessment
Three exposure metrics were
developed. Recent exposure:
Mean concentration during
2.5 hour lab
Cumulative exposure: PPM-
minutes for all previous
weeks
Past average exposure:
Cumulative exposure divided
by the total number of
minutes of exposure
Geometric mean
concentration 0.7 ppm
(GSD:2.13ppm). Peak 12
min concentration was 10.91
ppm. Average concentration
1. 1 ppm (SD = 0.56 ppm).
Formaldehyde concentrations
decreased over the 12 week
semester.
TWA personal breathing
zone samples were obtained
for each exposed subject
over 9 days and 1 unexposed
subject over 6 days. The
TWA exposure from
personal sampling ranged
from 0.07-2.94 ppm. More
than 94% of the subjects
were exposed to >0.3 ppm
and 3 1.7% were exposed to
an 8-hour time-weighted
average of >0.5 ppm.
Results, effect estimate, statistical significance
FVC
FEVj
FEV3
FFF
rJ-*r25-75
FEVj/FVC
Study group (n = 34)
Percent cross-lab
change (SD)
-1.4 (4.4)a
-0.03 (3.4)
-1.2 (4.2)
2.5 (8.7)
1.6 (3.8)b
Referent (n = 12)
-0.3 (4.6)
1.0 (4.0)
1.3 (3.29)c
2.31(2.71)
0.6 (2.9)
"p < 0.1, > < 0.05, "independent group,
^<0.1
& s
II
3 ,
o
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Akbar-
Khanzadeh et al.
1997
Krzyzanowski et
al. (1990)
Krzyzanowski et
Study design
Comparison of pulmonary
function (spirometry) among 50
exposed first-year medical
students and 36 unexposed
second-year physiotherapy
students during a 3 -hour
anatomy lab. At least 2 exposed
and 2 unexposed students, close
in age to the exposed, were
evaluated on each test day for
the duration of the course.
Lung function over one 3 -hour
lab session was evaluated for
each participant. Prelab
spirometric variables, expressed
as a percentage of reference
values accounting for height,
weight, age, sex, and race.
Cross-lab change analyzed
within and between groups.
Cross-sectional study of
residential formaldehyde
exposure. A stratified random
sample of households of
municipal employees was
selected based on information
about potential exposure (age of
housing) and potential
susceptibility obtained from an
initial screening questionnaire.
Households with children aged
5-15 years (613 adults and 298
children) were eligible for
inclusion. Trained subjects
measured peak expiratory flow
Exposure assessment
Personal (breathing zone) (n
= 44) and area (n = 76)
formaldehyde samples weer
collected. Average
formaldehyde concentrations
measured in the breathing
zone of the students for an
average of 157 minutes was
1.88 (SD = 0.96) ppm with a
range of 0.30-4.45 ppm.
Residential exposures to
formaldehyde were based on
two one-week samples from
each individual's kitchen,
living area, and bedroom
using passive sampling
tubes. The average
formaldehyde concentration
was 26 ppb, with a maximum
sample value of 140 ppb.
The majority of subjects
(83%) lived in homes with 2-
week average concentrations
below 40 ppb.
Results, effect estimate, statistical significance
FVC
FEVj
FEV3
FEF25-75
Mean percent cross-lab change over 3
hours (SD)
Exposed (n = 50)
2.5 (5.4)a
2.4(5.1)a
2.7 (4.6)b
2.2 (9.4)
Referent (n = 36)
4.6 (6.4)b
6.2 (7.0)bc
5.2 (6.5)bd
9.3(11.9)bc
ap < 0.01, V < 0.001, "independent group,
p < 0.01, Independent group, p<0.l
Random effects model, ages < 15 (N =
208; 3021 observations)
Factor
Formaldehyde
(household mean)
Morning FA (vs
bedtime)
Bedroom FA
*morning
Bedroom FA
squared*morning
Morning*asthma
B(SD)
-1.28 (0.46)a
-6.1 (3.0)a
0.09(0.15)
0.003 l(0.0015)a
4.59 (9.60)
& s
II
3 ,
o
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
al. (1990)
(continued)
Study design
rates (PEFRs) using mini-
Wright peak flow meters four
times daily, in the morning, at
noon, in the early evening, and
before bed, for 2 weeks. The
largest of three tests was
recorded for each test period.
PEFR was analyzed using a
random effects model adjusting
for asthma status, smoking
status, SES, NO2 levels,
episodes of acute respiratory
illness, and time of day.
Analysis performed separately
for ages below and over 15
years of age.
Exposure assessment
Results, effect estimate, statistical significance
Bedroom
FA*morning*asth
ma
Bedroom FA
sq*morning*asth
ma
Constant
"p < 0.05
-1.45 (0.53)a
0.031(0.006)a
349.6(13.2)
Random effects model, ages > 15 (N =
526; 8463 observations)
Formaldehyde
(household mean)
Morning FA (vs
bedtime)
Bedroom FA
*morning
Morning* smoking
Bedroom
FA*morning*smo
king
Bedroom FA sq
*morning
* smoking
Constant
*p < 0.05,
b0.05<^<0.10
0.09 (0.27)
-5.9(l.l)a
-0.07 (0.04)b
-7.4 (2.6)a
0.59(0.13)a
-0.007 (0.00 l)a
491.7(8.5)
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
s ,
oo
O
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Franklin et al.
(2000)
Main and Hogan
(1983)
Study design
Cross-sectional study of
residential exposure among 224
children (116 girls, 108 boys)
with no current, or history of,
upper or lower respiratory tract
disease were included from
responses to a respiratory health
questionnaire and household
inventory distributed through
local primary schools. Clinical
respiratory measures were
obtained at the children's
hospital. Exhaled nitric oxide
and skin prick tests for 7
common allergens were
measured.
Cross-sectional comparison of
21 individuals working in two
mobile trailers for 34 months
(mean age 38 ± 9 years, 76%
male, 19% nonsmokers) and 18
individuals who did not work in
the trailers (mean age 30 ± 6,
50% male, 22% nonsmokers).
Percent predicted FEVi and
FVC stratified by smoking
status (unadjusted group means
compared using t tests).
Exposure assessment
Three to four-day passive
samples were collected in the
child's bedroom and the
main living area of the house
and formaldehyde levels
were recorded as a time-
weighted average.
Three 1-hour area samples
using impingers were taken
on 4 occasions (August,
September, December, April)
always on a Monday. At
least 1 sample was taken
from each office in both
trailers. Concentrations
ranged from 0.12 to 1.6 ppm.
Results, effect estimate, statistical significance
FVC
FEVj
Formaldehyde
>50 ppb
<50 ppb
No association
No association
eNO (ppb)
15.5
8.7a
ap = 0.002, adjusted for age, atopic status
FEVj
FVC
FEF50
FEF75
%AFEF50
Data not presented
Data not presented
Range
10.5-22.9
7.9-9.6
Mean percent predicted
Exposed (N= 14)
98
94
93
69
55
Unexposed(jV= 17)
99
97
90
70
43
3 ,
VO
o
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Alexandersson et
al. (1982)
Alexandersson
and
Hedenstierna,
1989
Study design
Cross-sectional occupational
study at a carpentry works. A
total of 47 exposed workers
employed at the plant for > 1
year were included (mean age
35 years, mean duration 5.9
years) and compared to 20
unexposed employees.
Spirometric measurements were
obtained Monday morning
preshift and after work for
exposed. Lung function was
measured in the unexposed in
the morning or the afternoon.
Prospective occupational study
of cabinetry workers first
reported by Alexandersson et
al., 1982. Of 47 exposed
workers and 20 unexposed
workers examined in 1980, 34
exposed and 18 unexposed were
examined again in 1984. Of the
34 originally exposed, 13 had
been reassigned to other
unexposed jobs. The exposed
Exposure assessment
TWA formaldehyde
concentration, measured
using personal sampling in
the working zone over a
working day, was 0.36 ppm
(0.04-1.25).
Personal exposure monitored
during 3-4 15 -minute periods
during the work day. The
time-weighted formaldehyde
concentration, measured
using personal sampling in
the working zone over a
working day, was 0.42 ±
0.27 mg/m3 (0.34 ppm) in
1980 and 0.50 ± 0.12 mg/m3
(0.4 ppm) in 1984.
Results, effect estimate, statistical significance
FVC (L)
FEVj (L)
FEV%
MMF (L/sec)
CV%
Preshift lung function (SD)
Exposed
5.73 (0.14)
4.52(0.12)a
792 (1.0)
4.94 (0.2)
16.7 (1.07)
Referent
6.0 (0.2)
4.86(0.15)
80.7(1.32)
5.08(0.31)
17.1(1.5)
Difference from reference value, p = 0.08
FVC (L)
FEVj (L)
FEV%
MMF (L/sec)
CV%
ap< 0.001,
V < 0.05
FVC (L)
FEVj (L)
FEVj/FVC (%)
FEV25-75
Cross-shift change
-0.05
-0.173
-2.1b
-0.32b
3.4a
Annual change (1980-1984), Mean (SD)
Smokers (N= 10)
-15 (24)
-15 (21)
-0.1 (0.4)
-60 (69)
Nonsmokers (N =
11)
-10 (26)
-31(20)
-0.4 (0.2)a
-212 (66)a
to
o
fe
H
o
O
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Alexandersson
and Hedenstierna
(1989)
(continued)
Kilburn et al.
(1985)
Study design
and transferred workers had
been exposed to formaldehyde
an average of 1 1 years.
Spirometric measures were
compared with reference values
for sex, age, height, and weight.
The 5 -year change was
corrected for age-dependent
change. Results were presented
by smoking status.
Occupational study of 45
fiberglass batt makers (out of
110); aged 21-55 years (40%
Hispanic, 60% white) Another
exposed group of 18 male
histology technicians, aged 25-
48 years (4 Hispanic, 2 Oriental,
12 white). Reference group was
hospital employees, 20-62 years
of age (3 5% Hispanic, 65%
white). Spirometry
measurements were taken
before and after an 8-hour work
shift.
Exposure assessment
Formaldehyde exposure
categorized as high or low
based on reported work
assignments.
Results, effect estimate, statistical significance
CV%
ap< 0.001
FVC (L)
FEVj (L)
FEVj/FVC (%)
FEV25-75
cv%
"p < 0.05
Batt Makers
Lung function (%
cutoff)
FVC (L) (5%)
FEVj (L) (10%)
FEV25-75 (L/sec)
(15%)
FEV75-85 (L/sec)
(15%)
-0.6 (0.3)
Cross-shift change
(All AT =21)
-0.03 (0.09)
-0.05 (0.09)
-0.3 (0.18)a
-0.1 (0.18)
-0.8(0.14)a
0.2 (0.4)
Percent with specified percent decrease or
more relative to preshift value
Smokers (N= 3 5)
8.6a
11.4a
11.4a
40.0 a
><0.01
Nonsmokers (N = 9)
22.2
33.3
33.3
22.2
& s
II
s ,
to
fe
H
o
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
Reference
Horvath et al.,
1988
Holmstrom and
Wilhelmsson
(1988)
Holmstrom and
Study design
Occupational study of wood
products workers compared
pulmonary function between
109 exposed (workers at a
particleboard and molded
products operation, 68.6% of all
exposed) and 254 unexposed
(workers from nearby food
processing facilities). The
exposed workers had been
employed an average of 10.3
years (1-20 years). Spirometry
was conducted before and after
the work shift. Lung function
as percent of predicted normal
was compared between exposed
and unexposed (unpaired t-test).
Cross-sectional occupational
study of 70 individuals (87%
male) from a chemical plant
where formaldehyde and
formaldehyde products were
made. Exposure levels varied
from 0.05-0.5 mg/m3. A group
of 100 furniture workers was
exposed to formaldehyde and
wood dust with mean
concentrations of 0.25 mg/m3.
A comparison group of 36
persons (56% male) was mostly
comprised of clerks for the local
government in an office with
mean formaldehyde
concentrations of 0.09 mg/m3.
Mean duration of employment
Exposure assessment
8-hour TWA formaldehyde
was measured using
individual passive monitors
on the day of the exam (LOD
0. 1 ppm). Area levels were
measured with an active
sampling train (impingers).
TWA formaldehyde
averaged 0.69 ppm (
0.17-2.93 ppm) and 0.05
ppm (0.03-0.12 ppm) in the
exposed and unexposed
industries, respectively.,
Mean annual exposure to
formaldehyde was estimated
for each participant from the
beginning of employment.
Data on formaldehyde
concentrations was available
between 1979 1984 and
from 1-2 hour personal
sampling in breathing zone at
different workstations in
1985. Dose-years were
calculated for each worker.
Results, effect estimate, statistical significance
Preshift Lung
Function
FEVj (L)
FVC(L)
FEVj/FVC
PEFR (L/sec)
FEV25-75 (L/sec)
% Predicted (SD)
Exposed
103 (13)
105 (12)
96(8)
100 (23)
83 (22)
^>0.05
Referent
105 (13)
107(13)
95(8)
103 (22)
85 (25)
Cross-shift change in lung function compared between exposed
and referent. Statistically significant differences reported for
FVC, FEVj/FVC, FEV25-75, FEF50 and FEF75 (Data not
presented).
FVC
Observed
Expected
FEV%
Observed
Expected
FA exposed (N =
70)
4.979a
5.556
80.8
80.6
FA-dust exposed (N
= 98)
4.929a
5.593
78.3
79.5
Referent (N =
36)
4.539
4.718
81.4
80.7
apaired t-test comparing observed to expected, p < 0.001
to
to
fe
H
O
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Wilhelmsson
(1988)
(continued)
Holness and
Nethercott (1989)
Holness and
Study design
was 10.4 years for exposed and
11.4 for referent group.
Spirometric measures were
analyzed as percent of expected
normal based on age, sex,
smoking, height and weight.
Cross-sectional study of funeral
workers, including 67 currently
active embalmers and 17 who
were no longer active, were
recruited through a list of
funeral homes developed by the
Metropolitan District Funeral
Directors Association in
Toronto, Canada (86.6%
participation). An unexposed
group (N= 38) was recruited
from a large service
organization and paid student
volunteers. Information on
symptoms, past and family
medical history, and work
practices was collected by
questionnaire, and pulmonary
function and skin tests were
administered. Lung function
tests were performed on 22
embalmers just prior to and
following an embalming
procedure, and on 13 controls 2-
3 hours after for first test.
Funeral workers had performed
the embalming procedure for an
average of 10 years. Lung
function as percent predicted
Exposure assessment
The average formaldehyde
concentration from 2 area
samples (impingers),
measured during embalming
procedures lasting from 30 to
180 minutes, was 0.36 ± 0.19
ppm (U.Uo U.ot ppm).
Unexposed participants were
stated to be exposed to an
average concentration of
0.02 ppm.
Results, effect estimate, statistical significance
Lung function (%
predicted) (SD)
FVC
FEVj
FEVj/FVC
FEF50
FEF75
FVC
FEVj
FEVj/FVC
FEF50
FEF75
% Change in
Lung Function
during embalming
FVC
FEVj
FEF50
FEF75
Exposed (N = 84)
100.5 (12.3)
99.2 (12.9)
98.4 (7.9)
104.8 (29.7)
76.2 (32.9)
Active (N= 67)
100.7 (12.2)
100.8(12.19)
98.9 (7.8)
107.5 (28.7)
80.8(33.1)
Unexposed (N= 38)
100.9(11.5)
100.7 (12.9)
99.4 (8.7)
110.3(34.5)
86.6 (36.0)
Inactive (N= 17)
95.8 (12.0)a
93.1(14.1)b
96.6 (8.0)
94.1 (32.3)
57.1 (24.7)
ap = 0.0385, bp = 0.0652
Exposed (N= 22)
+0.88 (2.95)
-0.03 (2.4)
-2.28(13.43)
-8.55 (15.09)
Unexposed (TV = 13)
+1.13(3.98)
+1.45 (4.43)
+1.23 (12.44)
+1.93 (27.54)
J^.
K>
fe
H
O
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Nethercott (1989)
(continued)
Nunn et al.
(1990)
Malaka and
Study design
was compared between exposed
and unexposed using multiple
regression correcting for age,
height, and pack-years smoked.
Prospective, occupational study
of workers, aged 25 or older, at
a chemical factory in Duxford,
England, manufacturing urea
formaldehyde resin. A total of
164 workers exposed to free
formaldehyde in 1980 and a
group of 129 workers from the
bonded structures division at the
same factory in 1980 were
followed over a 6 year period
from 1980-1985. DataonFEV:
and FVC (highest of two
readings within 5% of each
other) were obtained from
routine annual health screenings
conducted by the same nurse
throughout the study period.
Follow-up was complete for
76% of the exposed and 74% of
the unexposed workers. FEVi
values (FEVi/height3), adjusted
for height, were regressed on
time of screening visit for each
worker, adjusting for age in
1980, smoking status in 1980
and at final assessment,
maximum and mean exposure
assessment level and total
duration of exposure.
Cross-sectional occupational
Exposure assessment
Data on formaldehyde
concentrations from area
samples (1-6 hours sample
collections) taken
periodically between 1979
and 1985, and from personal
samplers attached to
representative exposed
workers from 1985 to 1987
were used to categorize the
workers' employment
experience into low, medium
and high formaldehyde
groups corresponding to a 8-
hour TWA of 0.1-5.0 ppm,
0.6-2.0 ppm, and >2 ppm,
respectively. Exposure
assessments prior to 1976
were based on subjective
determinations and
knowledge of process
changes and industrial
hygiene measures.
Exposed and unexposed
Results, effect estimate, statistical significance
Decline in FEVi with age by smoking history (Mean slope, ml/year (95% CI)
Smoking status
Never
Ex-smoker
Current
Total
Exposed
45 (28-62)
33 (20-46)
46 (33-59)
42(34-51)
N
26
34
57
117
Mean Baseline Spirometric Values
Unexposed
29(7-51)
40 (26-54)
46 (32-61)
41 (32-50)
N
13
31
36
80
& s
II
s ,
J^.
K>
fe
H
O
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Kodama, 1990
Khamagaonkar
Study design
study at a plywood company
comparing a group of
formaldehyde exposed with
plant workers not exposed to
formaldehyde, matched for age,
ethnicity, and smoking status.
Exposed workers (N= 100)
were randomly selected with
stratification by smoking status
and years of occupation (< 5
and >5 years). The unexposed
group (N- 100) worked in
areas where formaldehyde was
not used (range 0.003-0.07
ppm) and had no previous or
current exposure to
formaldehyde based on
occupational histories. A total
of 93 exposed and 93
unexposed participants
completed the protocol (93%
participation). Baseline and
cross-shift spirometric
measurements were taken.
Lung function (percentage of
expected function) was
analyzed using analysis of
covariance adjusting for
sampled dust levels and
stepwise regression. Unexposed
workers had been employed
slightly longer than those in the
exposed group (6.7 ± 2.3 versus
6.2 ±2.4 years, p< 0.05).
Cross-sectional occupational
Exposure assessment
workers were identified
using area formaldehyde
measurements and personal
monitoring. Formaldehyde
levels ranged between 0.22
and 3.48 ppm (average 1.13
ppm). Exposure was
evaluated using a cumulative
measure using area
concentrations and duration
in current job (mean 6.29
ppm-year, SD 2.72).
Multiple thirty minute air
Results, effect estimate, statistical significance
FEVj/FVC (%)
FEVj (L)
FVC (L)
FEF25-75% (L/sec)
(adjusted for dust) (SD)
Exposed
84.7 (6.5)
2.78 (0.41)
3.28 (0.44)
3.04 (0.76)
*p< 0.001
Referent
86.9 (4.9)a
2.82 (0.30)a
3.37 (0.36)
3.44 (0.78)a
Multiple regression model of pulmonary function3
FEVj/FVC (%)
FEVj (L)
FVC (L)
FEF25-75% (L/sec)
B (per ppm-yr FA)
-0.347b
-0.015 b
NS
-0.043 b
Adjusted for age, height, weight, cigarettes/day, and dust
bp < 0.05
Mean lung function
J^.
to
fe
H
U
O
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
andFulare, 1991
Herbert et al.
(1994)
Herbert et al.
Study design
study of 74 individuals working
in anatomy and histopathology
departments at 3 colleges in
India and exposed to
formaldehyde. A comparison
group was matched by age and
sex (N = 74) (individuals not
working in laboratories with
formaldehyde). Persons with a
history of lung disease before
their present occupation were
excluded. Lung function tests
were performed on a subset of
37 exposed and 37 comparison
individuals on a Monday
morning after days of no
exposure.
Cross-sectional occupational
study comparing 99 oriented
strand board workers (exposed
to formaldehyde) (98%
participation rate) with 165
oil/gas field plant workers (not
exposed to formaldehyde) from
the same geographic area (82%
participation rate). Duration of
employment was a mean of 5. 1
and 10 years for OSB and
oilfield workers, respectively.
Spirometric testing (best of 5
satisfactory maneuvers) was
conducted at start of work shift
and after 6 hours. Lung
function was analyzed using
ANCOVA controlling for age,
Exposure assessment
samples were collected in the
breathing zone in both the
unexposed (N= 18) areas for
formaldehyde analysis.
Mean formaldehyde
concentrations were
1.00 ppm (range 0.036-2.27)
and 0. 102 ppm (range
0-0.52) among the exposed
and referent groups,
respectively.
Time weighted average
formaldehyde concentrations
based on 21 hour continuous
sampling in the breathing
zone at 5 work sites on 2
separate days ranged
between 0.07 and 0.27 ppm.
Results, effect estimate, statistical significance
FVC (L)
MMEFR (L/sec)
FEVj (%)
FEVj (ml)
FVC (ml)
FEVj/FVC (%)
FEVj (ml)
FVC (ml)
FEVj/FVC (%)
Exposed (N= 37)
2.18
1.55
60.68
Referent (N= 37)
2.63a
2.71b
78.74a
*p< 0.01, V< 0.05
Preshift lung function (Mean)
OSB
4.203
5.364
78.6
ap = 0.028
Oilfield
4.223
5.257
80.3a
Cross-shift difference in lung function
39b
47 b
0.1
V < 0.05
to
a\
fe
H
O
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
(1994)
(continued)
Pourmahabadian
et al. (2006)
Study design
height, and smoking.
Cross-sectional study of
formaldehyde exposure among
staff working in different
departments in 7 large hospitals
of Tehran University of Medical
Sciences in Iran. Pre- and
postshift lung function testing
was conducted for 124 workers
from the pathology labs (38),
surgery (65), endoscopy (21)
departments and 56
administrative affairs staff
during February 2002 to
February 2003.
Exposure assessment
Formaldehyde concentrations
were highest in the pathology
lab (8-hour average [SD]:
0.96 [0.74] ppm) compared
to the surgery rooms (0.25
[0.18] ppm) and endoscopy
departments (0.25 [0.22]
ppm). Highest one-hour spot
samples recorded
concentrations of 6.5 ppm,
2.7 ppm and 0.6 ppm in
pathology, surgery and
endoscopy, respectively.
Formaldehyde levels in
administrative offices were
not reported.
Results, effect estimate, statistical significance
VC(L)
FVC (L)
FEVj (L)
FEVj/VC (%)
FEVj/FVC (%)
FEF25-75 (L/sec)
VC(L)
FVC (L)
FEVj (L)
FEVj/VC (%)
FEVj/FVC (%)
FEF25_75 (L/sec)
Preshift lung function, Mean (SD)
Exposed
3.34(1.07)
3.46 (1.04)
2.45 (1.02)
73.35(21.53)
70.81 (17.8)
2.79(1.4)
Referent
3.92 (1.26)a
4.03 (1.23)a
2.90(1.21)b
73.98(21.84)
71.96 (24.4)
3.38(1.7)b
*p< 0.002, V< 0.01
Cross-shift difference, Mean (SD)
Exposed
0.14(0.15)
0.16(0.03)
0.15(0.02)
1.48 (6.67)
1.11(2.9)
0.13 (0.10)
Referent
0.02 (0.04) a
0.03 (0.03) b
0.00 (0.01) b
0.38(0.16)
0.54 (0.20)
0.03 (0.10)
ap< 0.001, V< 0.04
& s
II
s ,
J^.
to
fe
H
o
o
-------
Table 4-1. Epidemiologic studies of formaldehyde and pulmonary function (continued)
Reference
Ostojic et al.
(2006)
Study design
Cross-sectional occupational
study of 16 male health-service
professionals (8 medical
doctors, 8 lab technicians) at a
university in the pathoanatomic
laboratory for at least 4 years
with daily exposure to
formaldehyde. Exposed were
compared to 16 males matched
by age ad stature. All were
nonsmokers. The timing of the
spirometric tests was not stated.
Exposure assessment
Assessment of formaldehyde
exposure was not described.
Results, effect estimate, statistical significance
FVC
FEVj
FEVj/VC
PEF
MEF50
MEF25
Lung function (percent of expected, SD)
Exposed
111(14)
112(13)
108 (6)
99(19)
104 (22)
102 (22)
Referent
106(11)
102 (9)
96(6)
92(13)
110(20)
105 (25)
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
to
oo
fe
H
o
o
-------
1 ranged between 0.49-0.93 ppm with a geometric average concentration of 0.73 (SD 1.22) ppm.
2 Students, trained in the use of the flowmeters, took measurements at the beginning and end of
3 each session, and an additional 1 to 3 times during the sessions. The average of the 3 largest
4 values in a set was used in the analyses. The strongest pulmonary response was observed for the
5 average cross-laboratory decrement in PEF in the first 2 weeks of the study. Overall, the
6 students exhibited a 2% decrement in PEF during the laboratory session in the first 2 weeks,
7 while 5 students with any history of asthma showed a 7.3% decrement in PEF. The numbers in
8 the two groups were too small to evaluate differences in response statistically. The Kriebel et al.
9 (1993) study also shows how the acute effects of formaldehyde exposure were altered following
10 several weeks of weekly episodic exposure. By the fifth week of class, the pre- and
11 postlaboratory measurements of PEF were no longer reflecting a clearly demonstrated acute
12 effect, but, following the seventh week of episodic exposure, both pre- and postlaboratory PEF
13 continued to drop steadily until the class adjourned after 10 weeks. While the acute effects of
14 formaldehyde exposure appeared to diminish after several weeks of exposure, the intermediate
15 effect across 10 weeks was a 2.7 ± 1.1 L/minute per week drop in PEF that was statistically
16 significant (p < 0.01) in a model adjusting for random person effects, asthma, interaction
17 between time and asthma, and eye as well as nose symptoms of irritation. Prevalence of eye and
18 nose symptoms was associated with decreased PEF (p < 0.02).
19 These findings were partially corroborated in a subsequent study, which used a similar
20 study design and more thorough exposure assessment (Kriebel et al., 2001). This study,
21 involving 38 students with data for at least one week, measured formaldehyde in 6 zones in the
22 laboratory and linked students locations, recorded every 12 minutes, to the zone measurements to
23 estimate a mean formaldehyde exposure over the entire 2.5 hour lab. The geometric mean
24 formaldehyde concentration was 0.70 (GSD 2.13) ppm and the highest 12 minute exposure was
25 10.91 ppm. Generalized estimating equation models included an interaction term for recent
26 exposure (mean formaldehyde concentration during the lab) and time (natural log weeks) which
27 indicated an association with post lab PEF (percent of prelab PEF at first session) and recent
28 exposure that attenuated over several weeks. Past average exposure (average exposure over all
29 previous weeks) was associated with prelab PEF. The model indicated an approximate overall
30 decrease in maximum PEF of 1% per ppm formaldehyde associated with formaldehyde levels
31 during the laboratory in the early weeks of the course and an additional 0.7% (SE 0.3, p = 0.02)
32 decrease per ppm associated with past average formaldehyde levels during the weeks previous to
33 the current week. The addition of having a cold on the day of the lab reduced the coefficient for
34 past average exposure somewhat (B = -0.52 ±0.3, p = 0.08). Spirometry measurements (FEVi
35 and FVC) before and after the course were not related to formaldehyde exposure. Class
This document is a draft for review purposes only and does not constitute Agency policy.
4-29 DRAFT—DO NOT CITE OR QUOTE
-------
1 attendance decreased markedly over the semester and it is not clear whether attendance could
2 have been related to perceived tolerance of formaldehyde exposure in which case only the results
3 based on the first weeks of the class can be construed to corroborate the earlier results by Kriebel
4 etal. (1993).
5 A study of 103 medical students was performed over a period of 7 months in which the
6 students were exposed to formaldehyde in the breathing zone at a time weighted average that
7 ranged from 0.93 ppm to below the limit of detection (0.05 ppm) during anatomy laboratory
8 sessions meeting twice a week (Uba et al., 1989). Mean concentrations during dissection were
9 1.9 ppm (0.1-5 ppm). Monthly average formaldehyde concentrations were highest during the
10 first two months of the course (0.6 and 0.8 ppm) and declined over the next 5 months to 0.1 ppm.
11 Twelve students were asthmatics. Pulmonary function tests were conducted according to criteria
12 published by the American Thoracic Society before the start and at the end of the anatomy
13 laboratory on three occasions, before the laboratory began, at 2 weeks and after 7 months. The
14 largest of all acceptable spirometry values were used. Unlike the studies by Kriebel et al. (2001,
15 1993), these researchers did not find a prelab decrement in pulmonary function over the course
16 of 7 months. However, mean change for pulmonary function before and after the exposure,
17 expressed as percent of the prelab value, showed a similar pattern; at 2 weeks, all parameters
18 were decreased and, except for FEV1/FVC, the changes were larger compared to the first test
19 day when students were not exposed to formaldehyde (FVC -0.80 versus -0.23%,
20 p value < 0.001; FEV1 -1.04 versus -0.71%,p value 0.03; FEF 25-75 -1.99 versus-1.76%,
21 p value > 0.05; FEV1/FVC -0.36 versus -.48%, p value > 0.05). At 7 months, the changes were
22 less marked, with measures showing decreases in function at the end of the laboratory session
23 (measurements taken at the 7-month time point: FVC -0.79% (change greater than day 1,
24 p> 0.001), FEVi -0.48% (change less than day \,p< 0.01), FEF25-75% 0.07%, FEVi/FVC
25 0.24%). Symptoms of eye, nose and throat irritation, sneezing, rhinorrhea and chest tightness
26 (p = 0.05) were more frequent during the exposure (p < 0.01). While formaldehyde
27 concentrations appear to be comparable to those measured by Kriebel et al., Uba et al., used
28 spirometry to assess lung function at 3 separate points in time while Kriebel et al. measured peak
29 expiratory flow among subjects using multiple measures during each laboratory session
30 providing greater statistical power to detect a response to formaldehyde. Also, the lung function
31 parameters used by the two studies may have measured effects in different parts of the lung.
32 Similar studies among medical students have been performed. In one study, 34
33 nonsmoking exposed medical students and instructors and 12 nonmedical unexposed students
34 completed pulmonary function tests according to criteria published by the American Thoracic
35 Society before and after their work in the laboratory or at predetermined times for the unexposed
This document is a draft for review purposes only and does not constitute Agency policy.
4-30 DRAFT—DO NOT CITE OR QUOTE
-------
1 (approximately 3 hours) over five consecutive weeks (Akbar-Khanzadeh et al., 1994). The
2 medical students had been exposed to formaldehyde for at least 6 weeks prior to the study. The
3 time-weighted average exposure from personal sampling ranged from 0.07-2.94 ppm. More than
4 94% of the subjects were exposed to >0.3 ppm and 31.7% were exposed to an 8-hour time-
5 weighted average of >0.5 ppm. The best of 4 spirometric measurements were used in the
6 analysis. Comparing pre- and postexposures among the exposed students, on average FVC
7 decreased by 1.4%, FEV3 decreased by 1.2%, FEVi/FVC increased by 1.6%, and FVC25-75%
8 increased 2.5%. The average percent changes in the unexposed group were -0.3, 1.30, 2.31, and
9 0.6%, respectively. The researchers also calculated correlation coefficients by examining the
10 relationship between lung function and formaldehyde concentration, but no association was
11 found. However, the pulmonary function changes were not evaluated with statistical models and
12 baseline parameters were not adjusted for height, gender, or age.. Further, age and race
13 composition varied between exposed and unexposed. Therefore, it is not clear that the results of
14 this study are informative. Akbar-Khanzadeh and Mlynek (1997) performed another study with
15 50 exposed first-year medical students and 36 unexposed second-year physiotherapy students to
16 evaluate pulmonary function changes during a 3-hour anatomy lab that met during the morning.
17 Average formaldehyde concentrations measured in the breathing zone of the students was 1.88
18 (SD = 0.96) ppm with a range of 0.30-4.45 ppm. At least 2 exposed and 2 unexposed students,
19 close in age to the exposed, were evaluated on each test day for the duration of the course. Lung
20 function over one 3-hour lab session was evaluated for each participant. Spirometric
21 measurements were conducted according to criteria published by the American Thoracic Society.
22 Prelab spirometric variables, expressed as a percentage of reference values accounting for height,
23 weight, age, sex, and race, were comparable between the two exposure groups. Although lung
24 function (percent change from prelab value) increased in both groups during the first hour, the
25 authors reported a larger increase among the unexposed students when compared with exposed
26 students during the second two hours of exposure (FVC 3.0 versus 0.9%, FEVi 4.1 versus 1.2%,
27 FEV? 3.3 versus 0.8%, forced expiratory flow during the middle of the FVC [FEF25-?5%] 6.1
28 versus 0.7%). These differences between exposed and unexposed remained for FEVi, FEVs, and
29 FEF25-75% after 3 hours. Since lung function has been shown to have diurnal variation,
30 increasing during the morning, the authors suggested that formaldehyde exposure may have
31 inhibited a normal increase in respiratory function over the 3 hour period.
32 Finally, Fleisher (1987) gave self-administered questionnaires to 204 medical students
33 one month after completing an anatomy laboratory course (formaldehyde concentrations <1
34 ppm) and again after a subsequent pathology/microbiology laboratory course (no formaldehyde
35 exposure). Area samples taken in the anatomy labs on one day during the semester measured <1
This document is a draft for review purposes only and does not constitute Agency policy.
4-31 DRAFT—DO NOT CITE OR QUOTE
-------
1 ppm formaldehyde. Of the 38 students who completed the questionnaire after both courses, over
2 8% reported experiencing shortness of breath during the laboratory with formaldehyde exposure,
3 but none of the students reported shortness of breath in the laboratory session with no exposure.
4 No objective measurements of formaldehyde exposure were used and spirometric tests were not
5 performed. A higher proportion of students reported other symptoms after the anatomy class
6 compared to the pathology/microbiology class including eye and nose irritation, sneezing,
7 headache, nausea, cough, throat irritation and sinus problems (p value < 0.01).
8 Three studies have been performed that examine formaldehyde exposure from the
9 buildings in which individuals live or work. Krzyzanowski et al. (1990) studied the health
10 effects of formaldehyde exposure in a residential population in Arizona. A stratified random
11 sample of households of municipal employees was selected based on information about potential
12 exposure (age of housing) and potential susceptibility obtained from an initial screening
13 questionnaire. Households with children aged 5-15 years (613 adults and 298 children) were
14 eligible for inclusion in this cross-sectional study of home exposures. Residential exposures to
15 formaldehyde were based on two one-week samples from each individual's kitchen, living area,
16 and bedroom using passive sampling tubes. The average formaldehyde concentration was
17 26 ppb, with a maximum sample value of 140 ppb. The majority of subjects (83%) lived in
18 homes with 2-week average concentrations below 40 ppb.
19 Subjects' peak expiratory flow rates (PEFRs) were determined four times daily, in the
20 morning, at noon, in the early evening, and before bed, for 2 weeks. Subjects were trained to use
21 the mini-Wright peak flow meters and the largest of three tests was recorded for each test period.
22 PEFR data from the first 2 days of observations were excluded to account for a learning effect.
23 A statistically significant linear relationship between increased household mean formaldehyde
24 exposure and decreased PEFR was reported in children but not adults (B = -1.28 ± 0.46 L/minute
25 per ppb formaldehyde). A curvilinear (concave) relation was observed among asthmatic children
26 for increasing formaldehyde levels between zero and 50 ppb and decreasing morning PEFR.
27 Higher prevalence rates of physician-diagnosed asthma and chronic bronchitis were also shown
28 at higher concentrations of formaldehyde (60-140 ppb), an effect that was exacerbated by
29 environmental tobacco exposures. Among children with physician-diagnosed asthma, the
30 observed effects of increased formaldehyde exposure on decreased PEFR were more pronounced
31 (p < 0.05). All statistical models controlled for socioeconomic status, tobacco smoking (current
32 active or environmental tobacco smoking), and nitrogen dioxide concentrations.
33 Among adult smokers, there was a statistically significant nonlinear relationship with
34 decreasing morning PEFR at formaldehyde concentration > 40 ppb. In addition, the
35 investigators also demonstrated statistically significant interaction between formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-32 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposures, smoking status, and prevalence of chronic cough among adults. That is, a
2 formaldehyde concentration that caused decreased pulmonary function at residential levels also
3 caused chronic cough among nonsmokers.
4 The formaldehyde monitors were prepared by the Lawrence Berkeley Laboratories and
5 were considered to be precise and highly reliable. The 7-day passive formaldehyde monitors
6 generally provide the lowest limit of formaldehyde detection. This study found decrements in
7 lung function among children in a large, representative sample of the community associated with
8 formaldehyde concentrations in their homes. The investigators specifically tested an a priori
9 hypothesis and demonstrated to a high level of statistical significance that increased residential
10 formaldehyde exposures were associated with decreased pulmonary function as measured by
11 PEFR in children.
12 Franklin et al. (2000) studied children 6-13 years of age (median age = 9.5 years) and
13 measured the levels of formaldehyde in their homes. A total of 224 children (116 girls,
14 108 boys) with no current, or history of, upper or lower respiratory tract disease were included
15 from responses to a respiratory health questionnaire and household inventory distributed through
16 local primary schools. The length of time the children had lived in their homes was not reported.
17 Three to four-day passive samples were collected in the child's bedroom and the main living area
18 of the house and formaldehyde levels were recorded as a time-weighted average. Clinical
19 respiratory measures were obtained at the children's hospital. Exhaled nitric oxide was
20 measured during a single-breath exhalation and the average of three plateaus varying by less than
21 10% was used for the NO concentration value. Spirometry was conducted on all children
22 according to ATS guidelines and skin prick tests were measured for 7 common allergens. There
23 was no association between FVC or FEV1 and the indoor concentrations of formaldehyde,
24 although there were signs of lower airway inflammation as measured by levels of exhaled nitric
25 oxide (NO) in children exposed to average formaldehyde levels >0.05 ppm compared to <0.05
26 ppm (Franklin et al., 2000). The multiple regression models controlled for age and atopic status.
27 Other housing factors evaluated in bivariate analyses, were not associated with eNOS or
28 spirometry measures, and therefore were not included in the regression models (p > 0.1);
29 however, the effect of environmental tobacco smoke was not evaluated. The measurement of
30 lung volume at one point in time by Franklin et al. may have been less sensitive to detect an early
31 change in small airways than the measurement of flow rate used by Krzyzanowski et al.
32 Main and Hogan (1983) reported on a group of individuals (n = 21) working in two
33 mobile trailers for 34 months and exposed to levels of formaldehyde ranging from 0.12 to
34 1.6 ppm (mean age 38 ± 9 years, 76% male, 19% nonsmokers). The unexposed population was
35 comprised of individuals who did not work in the trailers (n = 18; mean age 30 ± 6, 50% male,
This document is a draft for review purposes only and does not constitute Agency policy.
4-33 DRAFT—DO NOT CITE OR QUOTE
-------
1 22% nonsmokers). There were no differences between the exposed and unexposed groups'
2 percent predicted FEVi or FVC regardless of smoking status (unadjusted group means compared
3 using t tests). Although pulmonary function measures were analyzed stratified by smoking status
4 in this small study, exposure to environmental tobacco smoke was more common among the
5 unexposed (44%) compared to the exposed group (29%). Since ETS is associated with lung
6 function, this parameter may have acted as a confounder.
7 Several studies allowed for the examination of potential chronic effects of formaldehyde
8 exposure. These included an occupational study at a plywood company by Malaka and Kodama
9 (1990) that reported preshift pulmonary function as a percentage of expected function among the
10 formaldehyde exposed compared with plant workers not exposed to formaldehyde, matched for
11 age, ethnicity, and smoking status. Exposed and unexposed workers were identified using area
12 formaldehyde measurements. Exposed workers (N= 100) were randomly selected with
13 stratification by smoking status and years of occupation (< 5 and >5 years). Formaldehyde levels
14 ranged between 0.22 and 3.48 ppm. The unexposed group (N= 100) worked in areas where
15 formaldehyde was not used (range 0.003-0.07 ppm) and had no previous or current exposure to
16 formaldehyde based on occupational histories. Among the 93 exposed and 93 unexposed
17 participants who completed the protocol (93% participation), an average formaldehyde exposure
18 of 1.13 ppm from area samples was associated with statistically significant decrements in FEVi,
19 FEVi/FVC, and FEF25-75o/0 using analysis of covariance adjusting for sampled dust levels. The
20 strongest response was for FEF25-75o/0! which showed a 12% drop in observed function compared
21 with expected function in the unexposed. Workers in the unexposed group had been employed
22 slightly longer than those in the exposed group (6.7 ± 2.3 versus 6.2 ± 2.4 years,/? < 0.05).
23 When the exposed group was categorized into groups of low and high exposure based on a
24 cumulative measure calculated from area concentrations and length of employment in the current
25 job, a decrease was reported for one measure, FEVI/FVC (percent of expected 86.9 ± 4.9,
26 85.3 ± 6.4, and 84.4 ± 6.5 among unexposed, low and high formaldehyde groups, respectively
27 (statistical tests not reported). In multiple regression models adjusting for age, height, weight,
28 cigarettes/day, and dust, formaldehyde as a continuous variable was a significant predictor for
29 FEVi, FEVi/FVC, and FEF25-75<>/0. Each unit increase in formaldehyde (ppm-years) was
30 associated with a decrease of 0.015 liters, 0.347%, and 0.043 1/s in FEVh FEVi/FVC, and
31 FEF25-75o/0, respectively. Changes in spirometric measures across the shift were not associated
32 with formaldehyde exposure in a subgroup of 55 exposed and 50 unexposed participants. A
33 higher prevalence of cough, phlegm, chronic bronchitis, asthma, occupational asthma, shortness
34 of breath and chest colds were reported by the exposed group (p < 0.04). This was a carefully
This document is a draft for review purposes only and does not constitute Agency policy.
4-34 DRAFT—DO NOT CITE OR QUOTE
-------
1 conducted occupational study with a high participation rate, relatively thorough exposure
2 estimates, appropriate reporting of methods, and control for potential confounders.
3 A cross-sectional study comparing 99 oriented strand board workers (exposed to
4 formaldehyde) (98% participation rate) with 165 oil/gas field plant workers (not exposed to
5 formaldehyde) from the same geographic area (82% participation rate) demonstrated a difference
6 in pulmonary function between the two groups (Herbert et al., 1994). Time weighted average
7 formaldehyde concentrations based on 21 hour continuous sampling in the breathing zone at 5
8 work sites on 2 separate days ranged between 0.07 and 0.27 ppm. The groups were similar in
9 regard to measured FVC and FEVi (controlled for age, height, and smoking), but the workers
10 exposed to formaldehyde had lower FEVi/FVC (78.6% and 80.3% for oriented strand board and
11 oil workers, respectively (p = 0.028)). In addition, those exposed to formaldehyde showed a
12 decrease in FVC and FEVi after their shift, with an average pre- and postshift difference of
13 47 mL (p = 0.022) and 39 mL (p = 0.044) for FVC and FEVi, respectively (however change
14 could not be compared with the controls of this study because no postshift measurements were
15 taken). Strand board workers were more likely to report respiratory symptoms in analyses
16 controlling for age and smoking status including cough, phlegm, shortness of breath, wheeze,
17 and chest tightness. Odds ratios were statistically significantly elevated.
18 Another occupational study of wood products workers found no differences in preshift
19 percent predicted pulmonary function between 109 exposed (workers at a particleboard and
20 molded products operation, 68.6% of all exposed) and 254 unexposed (workers from nearby
21 food processing facilities) (Horvath et al., 1988). Regression models controlled for height, sex,
22 age, and smoking. Formaldehyde measured using individual monitors averaged 0.69 ppm (
23 0.17-2.93 ppm) and 0.05 ppm (0.03-0.12 ppm) in the exposed and unexposed industries,
24 respectively., The exposed workers had been employed an average of 10.3 years (1-20 years).
25 Although there was no difference in preshift measurements, the authors found a cross-shift
26 decline in FEVi; FEVi/FVC, FEF25-75%, FEF25%- and FEF75o/0 among exposed workers and in
27 FVC and FEVi among the unexposed group (paired t-test,/> < 0.05). Thus, while acute declines
28 in large airway function over the shift were observed in both groups, the exposed group also
29 demonstrated declines in small airway function. The authors also evaluated cross-shift changes
30 between the groups and reported a significant difference for FVC, FEVi/FVC, FEF25-75o/0,
31 FEF25%- and FEF75o/0 (data were not presented). When the investigators assigned all unexposed
32 workers a formaldehyde exposure value of 0.05 ppm and evaluated formaldehyde associations
33 among exposed and unexposed combined, a correlation was detected in pre-and postshift
34 pulmonary function changes and formaldehyde levels, though no specific details on regression
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 5 DRAFT—DO NOT CITE OR QUOTE
-------
1 analyses were provided. Symptom prevalence related to cough, chest tightness and eye, nose
2 and throat irritation during the work shift was higher among the exposed workers (p < 0.01).
3 No association between formaldehyde and lung function was observed among funeral
4 workers and an unexposed control group (Holness and Nethercott, 1989). Funeral workers,
5 including 67 currently active embalmers and 17 who were no longer active, were recruited
6 through a list of funeral homes developed by the Metropolitan District Funeral Directors
7 Association in Toronto, Canada (86.6% participation). An unexposed group (N= 38) was
8 recruited from a large service organization and paid student volunteers. Information on
9 symptoms, past and family medical history, and work practices was collected by questionnaire,
10 and pulmonary function and skin tests were administered. The average formaldehyde
11 concentration, measured during embalming procedures lasting from 30 to 180 minutes, was 0.36
12 ± 0.19 ppm (0.08-0.81 ppm). Funeral workers had performed the embalming procedure for an
13 average of 10 years. Unexposed participants were stated to be exposed to an average
14 concentration of 0.02 ppm. The authors reported no difference in percent predicted pulmonary
15 function of the two groups at baseline, although 10% of the funeral workers compared to 3% of
16 the unexposed group had an FVC or FEV1 value less than 80% of predicted. Exposed and
17 unexposed were of similar age, height and years worked, although embalmers had higher
18 weights, a higher proportion of current smokers, and a higher number of pack-years smoked.
19 The 17 funeral workers who were no longer active had lower FVC (95.8 ± 12.0 versus 101.7 ±
20 12.2,/> = 0.04) and FEVI (93.1 ± 14.1 versus 100.8 ± 12.1,p = 0.07) compared to the 67 active
21 embalmers, when corrected for age, height, and pack-years smoked. The authors reported that
22 active workers were younger and had worked less time in the industry. After exposure, there
23 was no change in lung function for the exposed or unexposed when comparing lung function
24 tests done immediately before and after an embalming procedure (for controls the repeat
25 measures were taken approximately 2-3 hours after the first measure) (changes in percentage
26 predicted FVC and FEVi were 0.88 ± 2.95 and -0.03 ± 2.40 for exposed and 1.13 ± 3.98 and
27 1.45 ± 4.43 for unexposed). Further analysis showed no association between formaldehyde
28 levels and changes in lung function during the embalming procedure. Although this study did
29 not find lung function decrements among embalmers compared to an unexposed group,
30 decrements in FVC and FEVi among those who had stopped embalming suggests a possible
31 response among more sensitive individuals. Sample sizes in the exposure group were small, and
32 the null association with formaldehyde exposure may have been influenced by differences
33 between a healthy occupational population and the service organization volunteers.
34 A cross-sectional study of formaldehyde exposure among staff working in different
35 departments in 7 large hospitals of Tehran University of Medical Sciences in Iran reported
This document is a draft for review purposes only and does not constitute Agency policy.
4-36 DRAFT—DO NOT CITE OR QUOTE
-------
1 differences in spirometric measures related to formaldehyde (Pourmahabadian et al., 2006).
2 Lung function testing was conducted for 124 workers from the pathology labs (38), surgery (65),
3 endoscopy (21) departments and 56 administrative affairs staff during February 2002 to February
4 2003. Formaldehyde concentrations were highest in the pathology lab (8-hour average [SD]:
5 0.96 [0.74] ppm) compared to the surgery rooms (0.25 [0.18] ppm) and endoscopy departments
6 (0.25 [0.22] ppm). Highest one-hour spot samples recorded concentrations of 6.5 ppm, 2.7 ppm
7 and 0.6 ppm in pathology, surgery and endoscopy, respectively. Formaldehyde levels in
8 administrative offices were not reported. Pulmonary function measured before the work shift
9 was lower in the exposed group compared to the administrative staff. Average VC (liters, SD),
10 FVC, FEVi, andFEF25-75 were 3.34 (1.07), 3.46 (1.04), 2.45 (1.02) and 2.79 (1.4) in the exposed
11 groups and 3.92 (1.26), 4.03 (1.23), 1.21 (2.59) and 1.7 (2.5) among the unexposed group
12 (p < 0.01). FEVi/FC and FEVi/FVC were not significantly different between exposed and
13 unexposed. In addition, the change in pulmonary function across the shift was significantly
14 greater among the exposed workers compared to the unexposed (p < 0.04). Over 80% of the
15 exposed staff reported eye irritation and the prevalence of nose and throat irritation ranged
16 between 57 and 81%. The differences between the exposure groups is difficult to interpret
17 because the pulmonary function measures were not adjusted for age, gender, and height while
18 distribution by age, sex and smoking status varied between the groups.
19 A study performed in India (Khamgaonkar and Fulare, 1991) examined 74 individuals
20 working in anatomy and histopathology departments at 3 colleges in India and exposed to
21 formaldehyde (mean 1.00 ppm, range 0.036-2.27). A comparison group matched by age and sex
22 (N= 74) (individuals not working in laboratories with formaldehyde) was exposed to an average
23 of 0.102 ppm formaldehyde (range 0-0.52). Persons with a history of lung disease before their
24 present occupation were excluded. Multiple thirty minute air samples were collected in the
25 breathing zone in both the exposed (N= 43) and unexposed (N= 18) areas for formaldehyde
26 analysis. Lung function tests were performed on a subset of 37 exposed and 37 comparison
27 individuals on a Monday morning after days of no exposure in order to examine chronic effects.
28 The FVC, FEVi%, and maximum mid-expiratory flow rate of the exposed group, respectively,
29 were 17.12 (p < 0.01), 22.94 (p < 0.01), and 42.81% (p < 0.05) lower compared with the
30 unexposed. Mean height was comparable in the two groups. The exposed group also had a
31 higher prevalence of symptoms including productive cough, breathlessness and tightness of chest
32 (p < 0.01). However, while the pool of exposed and unexposed were matched on age and
33 gender, there was no mention by the investigators of normalizing the pulmonary function metrics
34 by gender and height, which would have made for more appropriate comparisons. Kilburn et al.
35 (1985) also demonstrated reduced pulmonary function (lower percent predicted FVC, FEVi, and
This document is a draft for review purposes only and does not constitute Agency policy.
4-37 DRAFT—DO NOT CITE OR QUOTE
-------
1 FEF25-75%) among 45 male fiberglass bat makers exposed to formaldehyde when compared with
2 26 hospital employees, including respiratory therapists, gardeners, and attendants, without
3 formaldehyde exposure. However, formaldehyde levels in work areas were not measured and
4 group differences were not analyzed statistically in this small study.
5 Two occupational studies found no association between formaldehyde exposure and
6 deficits in pulmonary function (Ostojic et al., 2006; Holmstrom and Wilhelmsson, 1988).
7 Ostojic et al. (2006) examined nonsmoking male health service professionals working in
8 pathoanatomic laboratories with 8 hours of formaldehyde exposure per day at an unspecified
9 concentration for at least 4 years (n = 16). The source of the comparison group, comprised of
10 sixteen age- and stature-matched nonsmoking male controls, was not described. There was no
11 difference in mean percent predicted FVC or FEVi between exposed and unexposed. The
12 researchers also examined values for diffusing lung capacity for carbon monoxide and membrane
13 diffusion capacity, which were similar between the exposed and control groups. However, blood
14 volume of pulmonary capillaries was found to be higher in the exposed group. Holmstrom and
15 Wilhelmsson (1988) recruited individuals from a chemical plant where formaldehyde and
16 formaldehyde products were made (n = 70). Exposure levels varied from 0.05-0.5 mg/m3. A
17 comparison group, mostly comprised of clerks for the local government worked in an office with
18 mean formaldehyde concentrations of 0.09 mg/m3 (n = 36). FEV% was not different from the
19 predicted value among either the exposed or comparison groups. Mean FVC was lower than
20 expected among the exposed group (expected values were based on age, sex, smoking habits,
21 height, and weight) but not among the comparison group. Spirometric measures between the
22 exposure groups were not statistically compared. The comparison group was older (39.9 versus
23 36.9 years of age) and contained more women (44% versus 13%). The investigators measured
24 changes in pulmonary function for those employed more than 5 years and reported no signs of
25 increasing restrictivity after 5 years. Cumulative exposure to formaldehyde was estimated using
26 mean annual formaldehyde concentrations based on sampling conducted between 1979 and 1985
27 summed over the number of years employed. There was no correlation between pulmonary
28 function and cumulative dose of formaldehyde (Holmstrom and Wilhelmsson, 1988). The
29 studies by Ostojic et al. and Holmstrom and Wilhelmsson are less informative because the
30 adequacy of the comparison group cannot be evaluated, sample sizes were small, and analytic
31 methods were not adequate. There have been two studies that have reported on the longitudinal
32 follow-up of workers exposed to formaldehyde (Nunn et al., 1990; Alexandersson and
33 Hedenstierna, 1989). The Alexandersson and Hedenstierna (1989) investigation examined
34 Monday morning preshift lung function and the acute effects of exposure across shift at a
35 carpentry works in central Sweden in 1980 and then again in 1984 (Alexandersson et al., 1982).
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 8 DRAFT—DO NOT CITE OR QUOTE
-------
1 The time-weighted formaldehyde concentration, measured using personal sampling in the
3
2 working zone over a working day, was 0.42 ± 0.27 mg/m (0.34 ppm) in 1980 and
3 0.50 ± 0.12 mg/m3 (0.4 ppm) in 1984. Of 47 exposed workers and 20 unexposed workers
4 examined in 1980, 34 exposed and 18 unexposed were examined again in 1984. Of the
5 34 originally exposed, 13 had been reassigned to other unexposed jobs. The exposed and
6 transferred workers had been exposed to formaldehyde an average of 11 years. Cross-
7 sectionally, a decreased preshift FEV1 relative to reference values (according to sex, age, height)
8 was suggested among exposed workers in 1980 (p < 0.1). In 1984, preshift FEV1, FVC, and
9 FEF25-75o/0 were significantly different from predicted among the 21 currently exposed workers.
10 Statistically significant annual decreases in FEVi/FVC and FEF25-?5% were noted over the
11 intervening 5 years in nonsmokers after correction for normal aging and reference lung function
12 spirometry values. The decrease in FEF25-75o/0 was 0.212 ± 0.066 L/second (mean ± SD) for each
13 year of exposure and was significant (p < 0.01). The decrease in FEVI/FVC was -0.4 ± 0.2
14 (p<0.01).
15 For comparison with the 12% drop in the same pulmonary metric reported by Malaka and
16 Kodama (1990) over an estimated 6.5 years, the extrapolated percentage decrease in FEF25-75<>/0
17 was computed for the Alexandersson and Hedenstierna (1989) study by using the reported yearly
18 decrement applied to the preshift values at the time of the initial study period. From the
19 predicted value of 4.26 L/second, a decrease of 0.168 L/second for each year of exposure
20 regardless of smoking status was calculated. For 6.5 years of exposure, this would result in a
21 24% drop in FEF25-75%. Significant across shift decreases in FEVI/FVC and FEF25-?5% were
22 reported in 1980 and cross-shift decreases in FEV1, FVC%, and MMF were report in 1984,
23 particularly for nonsmokers.
24 The study by Nunn et al. (1990) assessed the decrease in FEVi among workers, aged 25
25 or older, at a chemical factory in Duxford, England manufacturing urea formaldehyde resin
26 followed over a 6 year period from 1980. A total of 164 workers exposed to free formaldehyde
27 in 1980 and a group of 129 workers from the bonded structures division at the same factory in
28 1980 were included. Data on formaldehyde concentrations from area samples (1-6 hours sample
29 collections) taken periodically between 1979 and 1985, and from personal samplers attached to
30 representative exposed workers from 1985 to 1987 were used to categorize the workers'
31 employment experience into low, medium and high formaldehyde groups. Exposure
32 assessments prior to 1976 were based on subjective determinations and knowledge of process
33 changes and industrial hygiene measures. FEVI and FVC measurements (highest of two
34 readings within 5% of each other) were obtained during annual health screenings by the same
35 nurse throughout the study. Follow-up was complete for 76% of the exposed and 74% of the
This document is a draft for review purposes only and does not constitute Agency policy.
4-39 DRAFT—DO NOT CITE OR QUOTE
-------
1 unexposed workers. FEV1 values, adjusted for height, for each worker were regressed on time
2 of screening visit. The mean decrease in FEVi was 42 mL/year among workers exposed to
3 formaldehyde and 41 mL/year for workers who were not exposed to formaldehyde, however a
4 larger decline appeared to be evident among 36 nonsmoking exposed (45 ml/year, 95% CI:
5 28-62) compared to 13 nonsmoking unexposed (29 ml/year, 95% CI: 7-51). Other covariates
6 stated to be analyzed in relation to FEVI decline were FEVI level, age in 1980, smoking state in
7 1980 and final visit, maximum and mean exposure assessment level (1~5), and duration of
8 exposure but analytic details were not presented. The exposed were on average older and had
9 worked longer in the factory than the unexposed. It is difficult to interpret the findings of this
10 study because, although a small subset (N= 20) among the unexposed group were assessed and
11 not found to be exposed to other potential irritants above recommended limits, this may not have
12 been the case in the past. This group was exposed to asbestos, carbon and glass fibers,
13 silicaceous fillers, acid anhydrides, aliphatic amines, phenol formaldehyde (not free
14 formaldehyde) and urea formaldehyde (not free formaldehyde).
15 To summarize, the epidemiologic literature contains studies of formaldehyde exposure
16 and both acute and chronic effects on pulmonary function. Several studies with adequate
17 analytical methods and reporting have assessed acute exposures among naively exposed anatomy
18 graduate students (Kriebel et al., 1993; 2001), anatomy graduate students with several weeks of
19 episodic exposure (Kriebel et al., 1993; Uba et al., 1989), and postshift versus preshift worker
20 pulmonary function among those with regular occupational exposure (Malaka and Kodama,
21 1990; Herbert et al., 1994; Alexandersson et al., 1982; Alexandersson and Hedenstierna, 1989;
22 Holness and Nethercott, 1989; Horvath et al., 1988). Depending on whether the exposures are
23 naive or not, the epidemiologic studies that assessed the pulmonary effects after acute exposures
24 to formaldehyde are assessing different biological responses, namely, the acute effect alone or
25 the acute effect(s) in people who may have already been sensitized to different and unknown
26 degrees. Generally, these studies found that intermittent laboratory or occupational shift
27 exposures to high formaldehyde concentrations (time weighted averages between 0.3-3 ppm)
28 were associated with deficits in lung function across the shift or laboratory session and, in some
29 studies, over time (Kriebel et al., 1993; 2001; Uba et al., 1989; Herbert et al., 1994; Horvath et
30 al., 1988; Alexandersson et al., 1982; Alexandersson and Hedenstierna, 1989. Pourmahadadian
31 et al. (2006) reported cross-shift reductions in pulmonary function but results were not adjusted
32 for age, height or gender. Acute exposure associated with eye, nose and throat irritation and
33 other respiratory symptoms including chest tightness and cough also was reported by some
34 studies (Kreibel et al., 1993; Uba et al., 1989; Fleisher, 1987; Horvath et al., 1988;
35 Pourmahadadian et al., 2006).
This document is a draft for review purposes only and does not constitute Agency policy.
4-40 DRAFT—DO NOT CITE OR QUOTE
-------
1 Several adequately conducted and reported studies evaluated chronic effects of
2 occupational (Malada and Kodama, 1990; Herbert et al., 1994; Horvath et al., 1988; Holness and
3 Nethercott, 1989; Alexandersson et al., 1982; Alexandersson and Hedenstierna, 1989) or
4 residential exposure to formaldehyde (Krzyzanowski et al., 1990; Franklin et al., 2000).
5 Decreased preshift lung function relative to unexposed groups was reported in several different
6 occupational groups including plywood, oriented strand board and carpentry (Malaka and
7 Kodama, 1990; Herbert et al., 1994; Alexandersson et al., 1982; Alexandersson and
8 Hedenstierna, 1989). The studies included appropriate comparison groups, formaldehyde
9 measurements, adequate reporting of methods, and adjustment for potential confounding
10 variables including age and smoking status (in some studies weight and ethnicity). All studies
11 presented lung function values as percent of expected based on age, height, and sex or adjusted
12 for these variables in the analysis. Some studies reported lower mean values for several preshift
13 measures of lung function among formaldehyde exposed workers (university laboratories,
14 fiberglass bat makers) suggestive of chronic decrements in pulmonary function but results were
15 not adjusted for age, height or gender, or statistical analyses were not reported complicating their
16 interpretation (Khamgaonkar and Fulare, 1991; Pourmahadadian et al. (2006); Kilburn et al.,
17 1985). Studies of long-term exposure also reported increased respiratory symptoms such as
18 cough, increased phlegm, asthma, chest tightness and chest colds in exposed workers (Malaka et
19 al., 1990; Herbert et al., 1994; Pourmahabadian et al., 2006, Alexandersson et al., 1982;
20 Alexandersson and Hedentierna 1989). Some of the studies reporting null findings suffered from
21 small sample size, methodological, analytical or reporting deficiencies complicating the
22 interpretation of the findings (Main and Hogan, 1983; Ostojic et al., 2006; Holmstrom and
23 Wilhelmsson, 1988; Nunn et al., 1990). Also, occupationally exposed groups compared to
24 nonoccupational referent groups may have exhibited a healthy worker effect masking any
25 formaldehyde related lung function decrements (Holness and Nethercott, 1989).
26 The longitudinal study by Alexandersson and Hedenstierna (1989) provides support for
27 the association of lung function decrements over time with occupational formaldehyde
28 concentrations of 0.42-0.5 mg/m3 (340-400 ppb) over 5 years. Kryzanowski et al. (1990), a
29 well-designed and executed cross-sectional study of residential formaldehyde exposure in a
30 large, representative sample, provides clear evidence of a linear relationship between increased
31 formaldehyde exposure and decreased peak expiratory flow rate among children. The average
32 formaldehyde concentration was 26 ppb, with a maximum sample value of 140 ppb. Decrements
33 in PEFR associated with increasing formaldehyde concentrations were observed among adults
34 beginning at an average concentration of 40 ppb. While Franklin et al. (2000) did not observe an
35 association between a one-time measurement of FVC or FEVi among children aged 6-13 years
This document is a draft for review purposes only and does not constitute Agency policy.
4-41 DRAFT—DO NOT CITE OR QUOTE
-------
1 and indoor concentrations of formaldehyde, levels of exhaled nitric oxide (NO) were higher in
2 children exposed to average formaldehyde levels >0.05 ppm compared to < 0.05 ppm. These
3 findings indicate that formaldehyde may increase lower airway inflammation at concentrations
4 associated with effects on pulmonary function. Malaka and Kodama (1990) also observed an
5 exposure-response pattern for a cumulative formaldehyde measure and declines in FEVi/FVC.
6 The pulmonary function measures associated with formaldehyde exposure are consistent
7 with bronchial constriction, inflammation, or chronic obstructive lung disease. Decreases in
8 spirometric values, including vital capacity (VC), forced expiratory volume (FEV), forced vital
9 capacity (FVC) and FEV/FVC have been documented. Decreases in lung volume (FEVi, FVC)
10 indicate possible pulmonary obstruction (narrowing of the airways during exhalation) (Pellegrino
11 et al., 2005). Early changes in small airways are observed as reductions in expiratory flow in the
12 terminal portion of the spirogram (PEF, FEF75%, MEF25o/0_75o/0). These changes may be observed
13 even if FEVi is not affected.
14 Worker and laboratory exposures associated with cross-shift differences in spirometric
15 values are consistent with formaldehyde-induced sensory irritation. Concordance has also been
16 reported between subjective irritant response and measured changes in pulmonary function
17 further supporting the possibility that cross-shift and short-term evidence of bronchial
18 constriction may be a reflexive response to sensory irritation (Kriebel et al., 1993). Similar
19 findings have been reported following low-level residential formaldehyde exposure including
20 decreased PEFRs (Krzyzanowski et al., 1990).
21
22 4.1.1.2.2. Acute studies: controlled chamber exposures.
23 Pulmonary effects of acute formaldehyde exposure have been studied in both healthy
24 volunteers and sensitive populations under controlled conditions. Controlled chamber studies
25 have the advantage of measured controlled exposures, but other factors can limit the usefulness
26 of the studies, especially when study populations are small and there is high variability in the
27 measured parameters.
28 Anderson and M01have (1983) described a 5 hour controlled exposure study with
29 16 healthy students (5 female and 11 male). The students were an average of 23 years (20-33)
30 and included 5 smokers. After a 2 hour exposure to clean air, subjects in groups of four, were
31 exposed to four 5-hour formaldehyde exposures (0.3, 0.5, 1.0, and 2.0 mg/m3) on 4 consecutive
32 days in an order determined by a Latin square design. Measurements were taken prior to, after
33 2-3 hours, and 4-5 hours of exposure to formaldehyde. Only average standard deviations from
34 the mean were presented for vital capacity (0.28), FEF25.75o/0, (0.28), and FEVI .0 (0.24). The
35 authors concluded that no significant changes were measured in the airway resistance measures.
This document is a draft for review purposes only and does not constitute Agency policy.
4-42 DRAFT—DO NOT CITE OR QUOTE
-------
1 Day et al. (1984) exposed 18 volunteers to formaldehyde (1.0 ppm for 90 minutes)
2 generated either from formalin in chambers or offgasing from urea formaldehyde foam insulation
3 (1.2 ppm for 30 minutes) in individual hoods. Pulmonary function tests were conducted before
4 formaldehyde exposures in room air (0.02 ppm formaldehyde), immediately after exposures, and
5 8 hours after the beginning of the exposure. No differences in absolute values for FVC, FEV1 or
6 FEF25_75o/0 before and after formaldehyde exposures, analyzed using paired t-tests, were observed
7 among 9 subjects who claimed to experience adverse effects to UFFI and 9 who had not. For
8 example, FEV1 before, immediately after, and 6.5 hours after exposure was 3.31 ± 0.79,
9 3.32 ± 0.81, and 3.41 ± 0.77 L among those with complaints and 3.81 ± 1.0, 3.75 ± 1.04, and
10 3.71 ± 1.02 L among those with no complaints or not living in homes with UFFI. Mean change
11 in FEV1 after methacholine challenge was -3.2% (-13-7.4%) and 0.04% (-6.0-5.0%) among
12 those with prior complaints and those with no prior complaints, respectively. Since demographic
13 and other characteristics of the subjects were not presented, and only absolute values were
14 analyzed, the null findings in this small study cannot be adequately interpreted.
15 A study of healthy nonsmokers evaluated a dose-response relation using 2 groups (mean
16 age 26.3 ± 4.7 years, 10 males and 9 females) exposed at rest and during exercise (Kulle et al.,
17 1987). Subjects had no history of allergy, asthma, hay fever, or upper respiratory infection in the
18 6 weeks before the study began. One group of 10 subjects were exposed to formaldehyde
19 concentrations of 0.0, 0.5, 1.0, and 2.0 ppm in random order and the other group was exposed to
20 0.0, 1.0, 2.0, and 3.0 ppm in random order. Both groups were also exposed to 2.0 ppm with
21 exercise. Each session was separated by 1 week. Spirometric measurements were taken at the
22 beginning and end of each exposure, at several points during the 3 hour session, and 24 hours
23 post exposure. Airway resistance and thoracic gas volume were measured at the beginning and
24 end of each exposure. Nonspecific airway reactivity, measured by methacholine challenge, was
25 assessed at the end of the exposure and after 24 hours. No differences between dose levels and
26 no trend with increasing dose was found in analyses using lung function values at each time
27 point as a ratio of the value measured before exposure began. Lung function was not expressed
28 as a percent of predicted according to age, sex or height and dose level means were presented by
29 time point without standard deviations to characterize individual variability. In addition, no
30 differences in bronchial reactivity were observed.
31 A study of sensory irritation and subjective symptoms among 21 healthy volunteers with
32 4 hour exposures to 0, 0.15, 0.3 and 0.5 ppm formaldehyde over 2 weeks did not observe
33 differences in pulmonary function measured at baseline, before the first exposure and after the
34 last exposure (Lang et al., 2007). The differences between an individual's postexposure and
35 pre-exposure values were analyzed in relation to formaldehyde. The data were not presented.
This document is a draft for review purposes only and does not constitute Agency policy.
4-43 DRAFT—DO NOT CITE OR QUOTE
-------
1 Schachter et al. (1986) measured lung function among 15 subjects, aged 25.4 ± 4 years,
2 exposed for 40 minutes to 0 or 2 ppm formaldehyde at rest or during exercise. Each exposure
3 session was conducted on 4 separate days. Bronchial reactivity assessed by methacholine
4 inhalation challenge was measured in 6 subjects. Subjects were healthy nonsmokers with no
5 history of asthma. Lung function parameters, expressed as a percent change from the baseline
6 measurement, showed slight improvements at the end of 40 minutes some of which were
7 statistically significant. For example, FEV1 at 40 minutes was 1.65 ± 4.5 and 4.56 ± 5.3 liters
8 for subjects exposed to 2 ppm formaldehyde at rest and during exercise, respectively. FEV1
9 after exposure to room air only was -1.14 ± 4.8 and 1.6 ±7.7, respectively. Maximum expiratory
10 flow at 50% of expired vital capacity (MEF 50%) was 7.4 ±5.0 and 8.8 ± 8.1 L/second at rest
11 and during exercise. MEF50% after exposure to room air only was 2.74 ± 4.4 and 8.72 ± 12.6.
12 Standard deviations indicate large variability in individual responses to exposure. Differences in
13 responses related to formaldehyde versus room air exposures were not evaluated statistically.
14 In a subsequent report the research team described a study using the same exposure
15 protocol examining lung function among 15 healthy laboratory workers frequently exposed to
16 formaldehyde in their occupation (Schachter et al., 1987). The group of 5 men and 10 women
17 ranged in age from 19 to 60 years and included 3 current smokers. The frequency of
18 formaldehyde exposure varied considerably (between 1 and 7 days per week for 1 to 21 years.
19 Chamber exposures of 2 ppm formaldehyde at rest or with exercise did not induce lung function
20 changes (percent of baseline) during the 40 minute exposure and up to 30 minutes afterward.
21 While individual responses were not presented, the standard deviations for the mean percent
22 change from baseline were large.
23 Witek et al. (1986) performed the same 40 minute exposure protocol with 0 and 2 |ig/L
24 formaldehyde exposures at rest and during exercise among 15 healthy subjects and 15 subjects
25 with asthma. Again, lung function parameters, as a percent change from the baseline
26 measurement, showed slight improvements at the end of 40 minutes some of which were
27 statistically significant. For example, FEV1 at 40 minutes was 1.65 ± 4.5 and 4.56 ± 5.3 liters
28 for subjects exposed to 2 ppm formaldehyde at rest and during exercise, respectively. FEV1
29 after exposure to room air only was -0.41 ± 5.0 and 4.87 ± 8.3, respectively. Again, standard
30 deviations indicated large variability in individual responses to exposure.
31 Other acute controlled studies including asthmatics also reported no changes in
32 pulmonary function associated with formaldehyde exposure (Sheppard et al., 1984, Ezratty et al.,
33 2007; Harving et al., 1990; Green et al., 1987; Sauder et al., 1987; Witek et al., 1987, 1986),
34 including studies of individuals thought to have formaldehyde-induced bronchial asthma
35 (Krakowiak et al., 1998). Specific airway resistance was not significantly increased among
This document is a draft for review purposes only and does not constitute Agency policy.
4-44 DRAFT—DO NOT CITE OR QUOTE
-------
1 7 volunteers (2 females, 5 males) aged 18 to 37 years old with physician-diagnosed asthma after
2 exposure to 1 or 3 ppm formaldehyde for 10 minutes at rest or with moderate exercise (Sheppard
3 et al., 1984). Subjects were exposed via mouthpiece. Two of the subjects experienced large
4 increases (over 3 liters) in airway resistance after formaldehyde exposures of 1 and 3 ppm,
5 however these two individuals also exhibited increased airway resistance after air-only exposure.
6 Witek et al. (1986, 1987) evaluated lung function and bronchial reactivity among 15 nonsmoking
7 subjects with asthma (18-35 years of age) using the same exposure protocol as that used with
8 healthy subjects described in Witek et al. (1986). Similar to the response noted in healthy
9 subjects, the mean percent change from baseline in FEV1 during formaldehyde exposure among
10 asthmatics was increased. FEV1 values at 40 minutes were 4.59 ± 4.9 and 6.63 ±11.2 liters for
11 subjects exposed to 2 ppm formaldehyde at rest and during exercise, respectively. FEV1 after
12 exposure to room air only was 2.85 ± 4.6 and 5.81 ± 8.0, respectively. Differences in responses
13 related to formaldehyde versus room air exposures were not evaluated statistically. The
14 threshold dose of methacholine required to produce a 20% decrease in FEV1 was determined
15 immediately following an additional 40 minute exposure to formaldehyde among 12 subjects
16 with asthma. Mean PD2oFEV1.0 after exposure was decreased (13.6 ± 20.5) compared to the
17 response assessed at baseline (24.0 ± 15.7) (p = 0.12).
18 Harving et al. (1990) evaluated pulmonary function and bronchial reactivity among
19 15 nonsmoking volunteers with asthma (8 female and 7 male) exposed to filtered air (0.008
20 mg/m3 formaldehyde), 0.12 mg/m3 and 0.85 mg/m3 formaldehyde for 90 minutes. The subjects,
21 15 to 36 years of age, were assigned at random to 1 of 3 groups and each group (5 subjects per
22 group) was exposed to formaldehyde or filtered air in random order over a 3 week period. FEV1
23 as a percent of baseline was not significantly changed after a 90 minute exposure. Values were
24 104%, 103% and 103% for formaldehyde levels of 0.85, 0.12, and 0 mg/m3, respectively. In
25 addition, no significant changes in the concentration of inhaled histamine required for a 20% fall
26 in peak expiratory flow measured immediately after exposures were noted in relation to
27 formaldehyde levels. The PC2ofor subjects exposed to 0, 0.12, and 0.85 mg/m3 formaldehyde
28 was 0.29 ±0.3, 0.36 ± 0.53, and 0.26 ± 0.31, respectively.
29 Sauder et al. (1987) exposed 9 nonsmoking volunteers with asthma to 3 hour exposures
30 of clean air or 3 ppm formaldehyde on 2 days separated by 1 week. Spirometric measurements
31 were obtained at 0, 15, 30, 60, 120 and 180 minutes and a methacholine inhalation challenge was
32 conducted at 180 minutes immediately after the lung function test. Pulmonary function tests
33 including FVC, FEV1, FEF25_75o/0, specific airway conductance, and functional residual capacity
34 indicated no difference between formaldehyde and clean air exposures. Paired t-tests were used
35 to evaluate the ratio of each time point to the measurement at time = 0 for the formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-45 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure subtracted by the ratio for the clean air exposure at the same time point. Response to
2 methacholine challenge also was not affected by formaldehyde exposure.
3 Krakowiak et al. (1998) exposed 10 subjects exposed to formaldehyde occupationally
4 (aged 23-52 years) and 10 volunteers with no occupational exposure to formaldehyde (aged
5 19-49 years) to 0.5 mg/m3 formaldehyde for 2 hours. The group with formaldehyde exposure
6 had bronchial asthma diagnosed by the doctor in the workplace and included 7 males and 3
7 females, some of whom smoked. The nonasthmatic group included only nonsmoking males.
8 Exposure to formaldehyde did not result in changes in FEV1, PEF or the dose of histamine
9 causing a 20% fall in FEV1 measured before, immediately after and up to 24 hours after the
10 exposure period in either group.
11 Small but statistically significant deficits in pulmonary function due to acute
12 formaldehyde exposure (2 or 3 ppm) have been reported in healthy volunteers during exercise
13 (Green et al., 1987, 1989; Sauder et al., 1986). Although changes in lung function parameters
14 averaged over experimental groups were generally small, some individuals exhibited clinically
15 significant deficits, even after only 2 hours of exposure (Green et al., 1987). Nine healthy
16 nonsmoking subjects, aged 26 ± 3.6 years, were exposed to clean air for 3 hours on day 1 and
17 then 3 ppm formaldehyde for 3 hours on day 2 (Sauder et al., 1986). In addition, exercise for
18 8 minutes on a bicycle was included prior to each spirometry measurement during the exposure
19 at 30, 60, 90, 120, 150, and 180 minutes. Deficits in FEVi and FEF25-75o/0 after the first 30
20 minutes of formaldehyde exposure compared to clean air were 2% (p < 0.05) and 7% (p < 0.01),
21 respectively. Spirometric measures were not different after 60 and 180 minutes of exposure,
22 even when assessed as absolute rather than relative measurements. The authors reported that the
23 range of individual responses was -5% to +1% for FEVI and -14% to +2% for FEF25-75%.
24 Another study compared the responses of healthy and asthmatic subjects exposed to clean
25 air and 3 ppm formaldehyde for 1 hour with 15 minute exercise segments at 15 and 45 minutes
26 using the same analytical methods as described by Sauder et al. (1986, 1987) (Green et al.,
27 1987). Among the 22 healthy subjects, small but statistically significant decrements were
28 observed during 3 ppm formaldehyde exposures, compared to the clean air exposures, at 47 and
29 55 minutes. Decreases of 2.1 to 3.8% were observed for FEVI, FVC, and FEV3. Decreases in
30 FEF25-75o/0 also were observed (5.7 to 10.9%), but the changes were not statistically significant.
31 No changes in mean specific conductance or airway reactivity measured by methacholine
32 challenge were observed in either the healthy subjects or 16 subjects with a clinical history of
33 asthma. No changes in pulmonary function measures were observed in the asthmatic group.
34 Individual variability in responsiveness was noted by the authors. Thirteen percent (5 of
This document is a draft for review purposes only and does not constitute Agency policy.
4-46 DRAFT—DO NOT CITE OR QUOTE
-------
1 38 subjects) demonstrated formaldehyde-induced clinically significant deficits when exposed at
2 3 ppm during exercise (defined by Green et al. (1987) as a decrease in FEVi > 10% of control).
3 The authors followed up with an additional study to evaluate the effects of formaldehyde
4 alone or in combination with respirable carbon particles (mass median aerodynamic diameter
5 [GSD] 1.4 um [1.8 um]). A total of 29 healthy subjects were exposed according to a randomized
6 block design for 2 hours each to clean air, 3.0 ppm formaldehyde, 0.5 mg/m3 activated carbon
7 aerosol and a mixture of 3 ppm formaldehyde and 0.5 mg/m3 activated carbon aerosol.
8 Exposures occurred at the same time of day and were separated by one week. Spirometric
9 measurements were obtained prior to the beginning of exposure, and at 20, 50, 80, and
10 110 minutes during the exposure. In addition, a 15 minute bicycle ergometer exercise was
11 completed at 15, 45, 75, and 105 minutes. The subject also measured peak flow at the end of the
12 2 hour exposure, on the hour for 8 hours, and 12 and 16 hours post exposure using Wright peak
13 flow meters. A statistically significant decrease in FEF25-75<>/0 (< 6% mean decrease) related to
14 formaldehyde was observed among the 24 subjects who were able to complete the exercise
15 protocol at 50 (p < 0.05) and 80 minutes (p < 0.01), and a decrease in peak flow was evident at
16 110 minutes/? < 0.03). Formaldehyde exposure also resulted in decreased specific airways
17 conductance at 120 minutes (p < 0.01). No formaldehyde effects on FVC, FEVi, or FEV? were
18 observed for formaldehyde alone, however statistically significant decreases in FVC (mean
19 2.5-4.5% decrease) and FEVs (mean decrease 3%) were reported at 20 and 50 minutes, and peak
20 flow at 110 minutes. The combined exposure also significantly increased coughing at 20 and
21 80 minutes (p < 0.05). Formaldehyde alone did not have an effect on cough but statistically
22 significant increases in headache, eye, nose and throat irritation, and chest discomfort occurred at
23 all time points. The authors concluded that formaldehyde exposure during exercise resulted in
24 an acute, transient effect on both large and small airways among healthy individuals. In addition,
25 a combined exposure to formaldehyde and carbon particles increased coughing and small
26 decrements in pulmonary function, suggesting that adhesion to particles increased delivery of
27 formaldehyde to the lower respiratory tract.
28 Casset et al. (2006) evaluated the effect of formaldehyde exposure on the bronchial
29 response to dust mite allergen in sensitized asthma patients. The study included 19 nonsmoking
30 subjects (12 women and 7 men) ages 19-35 years with mild asthma. Subjects had positive skin
31 prick tests and specific IgE to Dermatophagoides pteronyssinus (dust mites). Subjects had not
32 had a respiratory tract infection in the two weeks prior to the testing. Individuals sensitized to
33 pollen were studied outside the relevant pollen season and people sensitive to pet allergens did
34 not have pets at home. The 30 minute crossover exposures to 100 ug/m3 formaldehyde and air
35 were randomly assigned with a three-week washout period in between. Subjects underwent mite
This document is a draft for review purposes only and does not constitute Agency policy.
4-47 DRAFT—DO NOT CITE OR QUOTE
-------
1 allergen challenge immediately after exposure to formaldehyde or air to determine the dose of
2 allergen that resulted in a 20% reduction in FEVi (PD2o). The dose which induced early-phase
3 bronchial response was significantly lower in the subjects exposed to formaldehyde (34.3 ng vs.
4 45.4 ng: p < 0.05). Late-phase bronchial response was measured over the 6 hours following the
5 dust mite challenge by comparing FEVi. The maximum percentage FEVI reduction observed
6 was significantly higher after exposure to formaldehyde 15% vs. 11% (p < 0.05) (Casset et al.,
7 2006).
8 While the study by Casset et al. (2006) clearly showed that acute formaldehyde exposure
9 (100 |ig/m3) enhanced both early-phase and late-phase bronchial responsiveness to mite allergen
10 in mite-allergen sensitized people with asthma, a subsequent study with a similar protocol did
11 not duplicate this finding. Ezratty et al. (2007) evaluated the response of asthmatics to inhaled
12 allergen after a 60 minute exposure to 500 ug/m3 (0.4 ppm) formaldehyde. The 12 subjects
13 (7 men and 5 women) were 18-44 years old, were nonsmokers, and were diagnosed with
14 intermittent asthma and allergy to pollen. No subjects had contracted an upper respiratory
15 infection for at least 4 weeks before the study. The crossover exposures (60 minutes) to filtered
16 air and 500 ug/m3 formaldehyde occurred 2 weeks apart in random order. Allergen inhalation
17 challenge, using an extract of 5 grass pollen allergens, was conducted immediately after each
18 exposure and the dose producing a 15% decrease in FEVI was determined (PDisFEVl).
19 Responsiveness to methacholine was determined 8 hours after the allergen inhalation
20 challenge ended. Lung function measurements were taken using a spirometer before, during and
21 8 hours after the end of the allergen challenge. In addition, PEF and FEVI were measured with a
22 portable spirometer every 15 minutes during exposures to filtered air or formaldehyde and every
23 hour until the methacholine challenge. The authors reported that pulmonary function, expressed
24 as percent predicted, was not affected by the formaldehyde exposure, although the data were not
25 presented. The median PDisFEVl for the allergen challenge was 0.80 (0.15-2.0) index of
26 reactivity (IR) after formaldehyde exposure and 0.25 (0.10-2.0) IR after the filtered air exposure
27 (p = 0.06). The ratio of response after formaldehyde exposure compared to filtered air exposure
28 was 1 in 7 subjects and higher in 5 subjects. The PD20 for methacholine challenge was 0.23
29 (0.01-3.6) and 0.17 (0.03-4) (p = 0.42). The PD20 for methacholine for formaldehyde compared
30 to filtered air was lower in 3 subjects, higher in 4 subjects, and not changed in 5 subjects.
31 There are multiple potential explanations for these seemingly conflicting findings in
32 Casset et al. (2006 and Ezratty et al. (2007) which include the small sample size of these studies,
33 the differencing in the particular allergen tested and difference in the protocols. It is possible
34 that the 12 subjects studied by Ezratty et al. (2007) may not have included individuals who were
35 especially susceptible to formaldehyde. In the Casset et al. (2006) study, the investigators made
This document is a draft for review purposes only and does not constitute Agency policy.
4-48 DRAFT—DO NOT CITE OR QUOTE
-------
1 specific mention of the size of the particles used for the allergen challenge. They commented in
2 the discussion that dosimetric models of inhaled formaldehyde show that the flux is very large in
3 the first bronchial generations and then decreased rapidly. Casset et al. (2006) used an aerosol
4 with large particles (Mass Median Aerodynamic Diameter of 11.1 jim) specifically because these
5 large particles are deposited in the large airways where formaldehyde flux is higher. Ezratty et
6 al. (2007) did not report the size of the particle used in their challenge but both studies did report
7 the types of dosimeters j et nebulizers that were used and these were different.
8 In a study unrelated to formaldehyde, Praml et al. (2005) compared the physical and
9 biologic doses of methacholine for different nebulizers including the type used by Ezratty et al.
10 (2007). Praml et al. (2005) compared the airway responsiveness of 34 subjects using two types
11 of nebulizers and found that in 17 subjects, neither system caused a 20% decrease in FEV1,
12 while among 8 subjects, both systems were able to provoke a 20% decrease. The remaining
13 9 subjects responded to only one type of nebulizer. Using the same protocol, the same
14 methacholine agent produced different results based on the type of nebulizer. It may be that the
15 results of Casset et a. (2006) and Ezratty et al. (2007) which did have other differences can be
16 explained thusly. It may also be that a difference in particle size plays a role if the pollen
17 allergens used in Ezratty et al. (2007) were smaller and penetrated the lung beyond the upper
18 lung where formaldehyde exposures are greater.
19 In general, acute formaldehyde exposures (0.5-3 ppm) have not induced significant
20 pulmonary deficits in healthy, nonexercising volunteers (Kulle et al., 1987; Schachter et al.,
21 1986; Schachter et al., 1987; Witek et al., 1986; Day et al., 1984; Andersen and Molhave, 1983).
22 However, it is unclear whether the data analysis in these reports had the statistical power to
23 substantiate the small deficits reported in occupational and student studies. All five reports had
24 relatively small study groups of healthy individuals (n = 19 [Kulle et al., 1987], n = 16
25 [Andersen and Molhave, 1983], n = 15 [Schachter et al., 1986], n = 15 [Schachter et al., 1987],
26 n = 15 [Witek et al., 1986], and n = 9 [Day et al., 1984]). The studies exposed a small number of
27 diverse individuals, often including males and females of varying age, and some included current
28 smokers. Three studies report the absolute values of the lung function parameters without
29 adjustment to individual expected function or the unexposed baseline for each individual (Kulle
30 et al., 1987; Andersen and Molhave, 1983; Day et al., 1984). As discussed, this decreases the
31 power of the study to detect formaldehyde-induced changes in pulmonary function. In contrast,
32 Witek et al. (1986) and Schachter et al. (1986, 1987) report lung function as a percent of baseline
33 (although not normalized for age, gender and height). Each study showed an increase in FEVi in
34 formaldehyde-exposed individuals at rest and increases in maximal expiratory flow (MEF) at
35 50% of expired vital capacity (MEF50%) (Witek et al., 1986; Schachter et al., 1986). However,
This document is a draft for review purposes only and does not constitute Agency policy.
4-49 DRAFT—DO NOT CITE OR QUOTE
-------
1 in both reports the SDs of changes in lung function parameters are quite large, nearly equaling
2 the reported value and exceeding it in several cases. The absence of normalized raw data,
3 combined with large individual variation, limit the interpretation of these studies.
4 Small but statistically significant deficits in pulmonary function (e.g., decreased FEVi,
5 FVCi, FEVs, specific airways conductance) due to formaldehyde exposure (2 or 3 ppm) have
6 been reported in healthy volunteers in controlled human exposure studies using exercise (Green
7 et al., 1987, 1989; Sauder et al., 1986;). Although changes in lung function parameters averaged
8 over experimental groups were generally small, some individuals exhibited clinically significant
9 deficits, even after only 2 hours of exposure (Green et al., 1987). This differential response
10 suggests susceptibility in certain subjects (Green et al., 1987). Other studies that included an
11 exercise component did not report a difference in response among healthy volunteers (Schachter
12 et al., 1986; Kulle et al., 1987). Acute controlled studies that evaluated responses among
13 asthmatics reported no changes in pulmonary function associated with formaldehyde exposure
14 (Sheppard et al., 1984, Ezratty et al., 2007; Harving et al., 1990; Green et al., 1987, 1989; Sauder
15 et al., 1987; Witek et al., 1987, 1986; Krakowiak et al., 1998). These findings suggest that a
16 brief exposure to formaldehyde may not trigger a response in the airways of asthmatic
17 individuals in the absence of allergen. However, the large variation in pulmonary response
18 among the individuals (healthy and asthmatic) that participated in the experimental exposure
19 studies suggests that some individuals may be more sensitive to formaldehyde.
20
21 4.1.1.3. Asthma
22 A large number of studies have investigated the potential association between
23 formaldehyde exposure and a continuum of adverse health effects ranging from decrements in
24 pulmonary function to asthma. In general, epidemiologic studies of adults have reported varied
25 results between null findings and positive findings but have not consistently distinguished
26 between studies in which formaldehyde may be causing an increase in the incidence of asthma
27 (e.g., phenotypic switching), increasing the prevalence of asthma, initiating an asthma attack or
28 worsening the severity of an attack. Formaldehyde may itself be an allergen or it may potentiate
29 the ability of other allergens to cause atopic switching or increase the sensitivity of atopic
30 individuals. Thus formaldehyde exposure among nonatopic individuals could theoretically cause
31 atopic switching in the presence or absence of allergens possibly resulting in a diagnosis of
32 asthma. Formaldehyde could also cause an asthma attack or potentiate the influence of other
33 stimuli on the risk of asthma attacks. Demonstration of a clear association of formaldehyde
34 exposure, or the lack of an association, at one particular time does not necessarily imply that
35 exposure to formaldehyde is causing or not causing adverse outcome at other times.
This document is a draft for review purposes only and does not constitute Agency policy.
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
The National Research Council concluded in its report on Formaldehyde that,
"Formaldehyde has been shown to cause bronchial asthma in humans" (NRC, 1981), citing
numerous studies demonstrating the induction of asthma following exposure to formaldehyde
(Hendrick and Lane, 1975, 1977; Laffont and Noceto, 1961; Nova and Touraine, 1957; Paliard et
al., 1949;Popaetal., 1969; Sakula, 1975; Schoenberg and Mitchell, 1975; Turiar, 1952;
Vaughan, 1939). In a subsequent review article on formaldehyde and the health effects that have
been associated with it, Stenton and Hendrick (1994) reported on formaldehyde and asthma in
occupational settings and starkly describe the ".. .first detailed case report of formaldehyde
asthma confirmed by specific inhalation challenge test occurring in a nursing sister on a renal
dialysis unit. Her symptoms were suggestive of late asthmatic reactions occurring 4 to 5 hours
after heavy exposures. The occurrence of late reactions was confirmed in a series of challenge
tests" (Stenton and Hendrick, 1994; Hendrick 1997). The results of the challenge tests are
illustrated in Figure 4-1.
140 -
120 -
I100'
Ul
u BO -
W 60 -
IE
a.
20 -
-8C
" v'
1 -40 -20 0
^-^t^-7^7^
» • /
* • '"
\l /
fwmilin 25« tor 15 mini (1)
, , , n , j | | ] 1
20 40 60 2 4 9 8 10 12
nmuiM hourt
TIME FROM ONSET OF EXPOSURE
Figure 4-1. Delayed asthmatic reaction following the inhalation of
formaldehyde after "painting" 100% formalin for 20 minutes. Challenge 2
was premedicated with inhaled betamethasone 200 jig.
Source: Stenton and Hendrick (1994).
Five years later, the two nurses were re-challenged with the nurse who had left the
dialysis unit having no response to the subsequent challenge while the nurse who had remained
working in the unit developed mild late asthmatic response with peripheral blood eosinophilia
This document is a draft for review purposes only and does not constitute Agency policy.
4-51 DRAFT—DO NOT CITE OR QUOTE
-------
1 (Stenton and Hendrick, 1994; Hendrick et al., 1982). Stenton and Hendrick (1994) concluded
2 that these studies "provide clear evidence of formaldehyde's ability to induce asthma" but no
3 indication of the exposure concentrations to induce it. In a follow-up study of dialysis unit
4 staffers exposed to formaldehyde as a sterilizing agent, 8/28 people reported respiratory
5 symptoms and a prolonged increase in circadian rhythm of peak expiratory flow rate was seen in
6 one subject (Hendrick and Lane, 1983) implying an increase in airway responsiveness (Stenton
7 and Hendrick, 1994). It should be noted, however, that while there did appear to be a clear
8 response to formalin, it is not known what contribution to the response was attributable to
9 formaldehyde and what contribution might have been attributable to methanol. Other cases of
10 formaldehyde asthma have been described. Nordman et al. (1985) describe 12 cases and refers
11 to several other case reports (Popa et al., 1965; Sakula 1975; Alanko et al. 1977). While the
12 evidence of a causal association between formaldehyde and asthma is clear, the above studies do
13 not offer information on the concentrations at which adverse effects would be expected in a
14 population. While formaldehyde exposure is generally considered an etiologic factor for the
15 development of asthma in occupational settings it appears to be a rare occurrence.
16 Numerous epidemiologic studies have investigated adverse effects in populations.
17 Decreased peak expiratory flow rates (PEFR) are an important component in the diagnosis of
18 asthma and there is extensive evidence of formaldehyde-induced decrements in PEFR (see
19 Section 4.1.1.2). However, the diagnosis of asthma is both a more serious health condition and
20 diagnostically more complex than decreased PEFR alone and is evaluated here as a distinct
21 endpoint. While epidemiologic studies have investigated the potential association between
22 formaldehyde exposure and a continuum of adverse health effects from pulmonary function to
23 asthma, few nonoccupational studies have evaluated the potential effects of formaldehyde
24 exposure on the risk of asthma onset (Delfino 2002).
25 However, residential formaldehyde exposure was reported to be associated with an
26 increased risk of incident asthma in a population-based case-control study of 192 children aged 6
27 months to 3 years (Rumchev et al., 2002). The study was comprised of 88 children discharged
28 from the emergency department of a children's hospital in Perth, Australia, with a primary
29 diagnosis of asthma and 104 controls from the same community identified through the health
30 department. Information about the child's respiratory condition and risk factors for asthma was
31 obtained via a questionnaire compiled by the parent. Seasonal (winter, summer) in-home
32 formaldehyde measurements taken in the living room and subject's bedroom were used to assess
33 exposure (8-hour passive sampler). The odds ratios (ORs) for risk of asthma diagnosis by
34 formaldehyde exposure level category (10-29, 30-49, 50-59 and >60 ug/m3) were adjusted for
35 measured indoor air pollutants, allergy levels of house dust mite, relative humidity, indoor
This document is a draft for review purposes only and does not constitute Agency policy.
4-52 DRAFT—DO NOT CITE OR QUOTE
-------
1 temperature, family history of asthma, atopy, age, sex, socioeconomic status, smoking, presence
2 of pets, air conditioning, humidifier and gas appliances. Of these, age, allergic sensitization to
3 common allergens, and family history of allergy were independent risk factors for asthma (ORs
4 of 1.09, 2.57, and 2.66, respectively). Coexposures to other indoor air pollutants were also
5 controlled for including benzene, toluene and ethylbenzene (Rumchev et al., 2004).
6 Categorical analysis of the data indicates the ORs for asthma were increased in the two
7 highest formaldehyde exposure groups, reaching statistical significance for household exposures
8 > 60 |ig/m3 (49 ppb) (OR of 1.39) (Rumchev et al., 2002). Analysis of the data with
9 formaldehyde as a continuous variable provides a statistically significant increase in the risk of
10 asthma (3% increase in risk per every 10 |ig/m3 increase in formaldehyde level. The paper states
11 this effect as OR 1.003 (95% CI 1.002-1.004) which appears at odds with a 3% increased in risk
12 per every 10 |ig/m3 but this must be the effect per 1 |ig/m3 and can be confirmed by comparing
13 the per unit effect to the plotted results1. All analyses controlled for other indoor air pollutants,
14 allergen levels, relative humidity, and indoor temperature as well as other risk factors.
15 While the study by Rumchev et al. (2002) focused on formaldehyde controlling for other
16 indoor air pollutants, a subsequent report described the specific effects of those indoor air
17 pollutants (Rumchev et al., 2004). This paper evaluated the risk of asthma incidence with
18 10 VOCs. The highest odds ratios were increased risks of asthma diagnosis associated with
19 benzene, toluene, and ethylbenzene and were statistically significant associations. Compared to
20 the effects observed for formaldehyde, the strength of the associations appear to be stronger on a
21 per 10 |ig/m3 basis. The strength of these effects is an important consideration as the relative
22 strength of the VOC effects appears to be larger than that attributable to formaldehyde if the
23 effects of the measured indoor air pollutants had not been controlled for in the formaldehyde
24 analysis (Rumchev et al., 2002). However, as these indoor air pollutants had been controlled for,
25 the reported effect of formaldehyde should be independent of the effect of benzene and other
26 VOCs in the absence of residual confounding. If two factors both cause the same outcome and
27 are statistically associated, then they may mutually confound. In Rumchev et al. (2004) on page
28 750, the investigators assessed whether the effect of the VOCs were confounded by
29 formaldehyde and stated that the results showed that exposure to VOCs still had a highly
30 significant effect on asthma even when formaldehyde was controlled for. This finding further
31 substantiates the formaldehyde finding since mutual confounding was not identified.
1 In order to confirm that the effect size is 1.003 per unit change in exposure and 1.03 per 10 units, EPA compared
these results to the plotted results in Rumchev et al. (2002). A line drawn across the plot at OR = 1.003 per one unit
change in exposure estimates the non-linear categorical results well. At 60 ug/m3, the extrapolated linear effect
would be OR = 1.2.
This document is a draft for review purposes only and does not constitute Agency policy.
4-53 DRAFT—DO NOT CITE OR QUOTE
-------
1 Several other nonoccupational studies have evaluated the association between
2 formaldehyde exposure and the prevalence of asthma among children (Garrett et al., 1999;
3 Tavernier et al., 2006; Gee et al., 2005; Krzyzanowski et al., 1990; Palczynski et al., 1999).
4 Three studies (Tavernier et al., 2006; Gee et al., 2005; Garrett et al., 1999) were performed by
5 matching children with and without asthma and comparing the levels of formaldehyde in their
6 homes. Gee et al. (2005) selected 100 cases with current asthma and 100 controls from
7 2 primary care facilities in an area of England with low socioeconomic status. Cases were
8 identified through a screening questionnaire that had been validated with diagnoses by
9 physicians. Cases and controls (aged 4-16 years) were matched by age and sex. Median
10 formaldehyde levels were 0.03 ppm in living rooms and 0.04 ppm in bedrooms. Univariate
11 comparisons found no differences in formaldehyde levels between cases of current asthma and
12 controls without asthma. Notably, no association was observed for pollutant indicators of
13 environmental tobacco smoke and current asthma, a recognized risk factor. A subsequent study
14 of the same children in the same homes conducted a more thorough evaluation of risk factors
15 (Tavernier et al. (2006). Again, a one-week average formaldehyde concentration in the living
16 room or bedroom was not found to be associated with current asthma in multivariate analyses
17 adjusted for several indoor variables. Respirable particulates, tobacco specific particles, volatile
18 organic compounds, and nitrogen dioxide also were not associated with current asthma.
19 Tavernier et al. (2006) did not report the measured levels of formaldehyde, but gave the OR for
20 the highest tertile of exposure in the bedroom compared with the lowest tertile of exposure as
21 0.99 (95% CI: 0.39-2.50). The odds ratio for the second tertile compared to the lowest tertile
22 was 1.22 (95% CI: 0.49-3.07). The width of the confidence intervals indicates that study did not
23 have adequate statistical power to detect low level risks and suggests that these findings would
24 still be consistent with a two-fold increase in risk.
25 Garrett et al. (1999) reported on the risk of allergy and asthma-like respiratory symptoms
26 due to formaldehyde exposure in a cross-sectional survey of 80 households in rural Victoria,
27 Australia with children, aged 7-14 years, with (n = 53) or without (n = 88) doctor-diagnosed
28 asthma. Households were recruited via schools, medical centers, and advertisements in the local
29 press. The study was designed to include asthmatic children in half of the households and the
30 study recruited 43 households with at least one child with asthma diagnosed by a doctor and
31 37 households with no asthmatic children. Formaldehyde exposure was characterized by
32 4 seasonal in-home sampling events in 1994 and 1995 (4-day passive samples) in bedrooms of
33 all participating children and in living rooms, kitchens, and outdoors. Median indoor
34 formaldehyde concentrations were 15.8 ug/m3 (12.6 ppb) with a maximum of 139 ug/m3
35 (111 ppb). The median outdoor concentration was 0.7 ug/m3 with a range of < 0.3-15.3 ug/m3.
This document is a draft for review purposes only and does not constitute Agency policy.
4-54 DRAFT—DO NOT CITE OR QUOTE
-------
1 Information on asthma respiratory symptoms during the previous year was obtained through an
2 interview with a parent after sampling was completed. An erratum to the original paper reported
3 that the column headers in two tables were switched but that the summary statistical and
4 conclusions in the 1999 report were correct as published. The proportion of asthmatic children
5 by the highest formaldehyde level measured over four seasons was 0.16, 0.39, and 0.44 for
6 <20 ug/m3, 20-50 ug/m3, and >50 ug/m3, respectively (test for trend, p< 0.02). However, in
7 logistic regression models, the ORs for the association did not remain statistically significant
8 after controlling for parental allergy and asthma (ORs and 95% CIs were not provided).
9 A large, representative study of 202 households (mean formaldehyde level of 26 ppb)
10 found that among children aged 6-15 years old and exposed to environmental tobacco smoke,
11 the prevalence of physician-diagnosed asthma was 45.5% for those with measured levels of
12 formaldehyde in the kitchen >60 ppb (N = 11). The prevalence of asthma dropped to 0% for
13 levels 41-60 ppb (N= 12) and 15.1% for levels <40 ppb (N= 106) (chi-squared trend test
14 p< 0.05). No trend in asthma prevalence was seen for children who were not exposed to
15 environmental tobacco smoke (Krzyzanowski et al., 1990).
16 A study performed by Tuthill (1984) measured formaldehyde exposure for children
17 grades K through 6 by using a combination of proxy variables. Overall, there was no
18 association, but some individual variables did show an increased risk. For example, the reported
19 risk ratio for having new construction or remodeling performed in the house in the past 4 months
20 was 2.5 (95% CI: 1.7-3.9). The risk ratio for having new or upholstered furniture in the house
21 (brought into the house within the past 4 months) was 2.2 (95% CI: 1.2-3.9).
22 A study in Poland randomly selected 120 households with children 5-15 years of age in
23 10 year old apartment houses (Palczynski et al., 1999). Using self-reported asthma prevalence as
24 an outcome, study investigators found no association with levels of formaldehyde (mean
25 25.9 ug/m3, range 2.0-66.8 ug/m3) measured using 24-hour samples in the children. Among
26 adults, the authors reported a higher prevalence of allergic diseases in the highest formaldehyde
27 exposure group but that the group was too small for statistical evaluation. However, the
28 prevalence of allergic asthma was higher among adults exposed to 25.1-50 ug/m3 compared to
29 <25 ug/m3 and exposed to environmental tobacco smoke (p = 0.03).
30 Delfino et al. (2003) conducted a panel study of 22 Hispanic children with a minimum
31 one year history of doctor diagnosed asthma, aged 10-16 years, and living in Los Angeles. The
32 participants were nonsmokers from nonsmoking households, and lived and went to school within
33 3 miles of a central site monitor. The children recorded the severity of asthma symptoms in daily
34 diaries for 3 months. The mean outdoor 24-hour levels of formaldehyde were 7.21 ppb (range
35 4.27-14.02 ppb). A positive association between asthma symptom scores (comparing children
This document is a draft for review purposes only and does not constitute Agency policy.
4-55 DRAFT—DO NOT CITE OR QUOTE
-------
1 who reported bothersome or more severe symptoms, including those that interfered with their
2 daily activities, versus those with no symptoms or symptoms that were not bothersome) and
3 increasing levels of formaldehyde measured on the previous day was observed (OR 1.37 [95%
4 CI: 1.04-1.89]). The generalized estimating equation models adjusted for the occurrence of
5 respiratory infections. Notably a test for effect modification revealed a stronger association with
6 formaldehyde among children who did not regularly take anti-inflammatory medications.
7 Other studies of residential exposure to formaldehyde have investigated asthma
8 prevalence and symptoms in adults (Norback et al., 1995; Matsunaga et al., 2008). A cross-
9 sectional study by Norback et al. (1995) reported mean levels of formaldehyde, measured on one
10 day for 2 hours, were 29 |ig/m3 (range <5-l 10 |ig/m3) in the bedrooms of individuals
11 experiencing nocturnal breathlessness compared with formaldehyde levels of 17 |ig/m3
12 (<5-60 |ig/m3) among those without nocturnal breathlessness. The study sample was
13 88 individuals from a community in Sweden who responded to a screening questionnaire, had
14 lived in the same home during the study period, and agreed to a medical interview and home
15 sampling (58% of recruited). The eligible population was residents of Uppsala, Sweden who had
16 answered yes (N= 74) to one of 3 questions regarding attacks of asthma in the previous
17 12 months, nocturnal breathlessness in the past 12 months, or current use of asthma medications
18 or a random sample of those who had answered no (N= 80) to all three questions. The OR for
19 nocturnal breathlessness was 12.5 (95% CI: 2.0-77.9) per a 10-fold increase in the indoor
20 concentration in logistic regression models adjusted for age, sex, current smoking, wall to wall
21 carpets and presence of house dust mites. The effect was substantially stronger in magnitude
22 than the associations observed for toluene, terpenes, and volatile organic compounds. Therefore,
23 the association with formaldehyde is likely not entirely explained by confounding by the volatile
24 organic compounds
25 A recent cross sectional study (Matsunaga et al., 2008) found no association between a
26 24-hour personal sample for formaldehyde and prevalence of asthma when pregnant women with
27 an exposure >47 ppb were compared to those with exposure to <18 ppb. The adjusted odds ratio
28 from the logistic regression model was 2.65 (95% CI: 0.63-11.11). The small number of women
29 identified with asthma resulted in wide confidence intervals and low statistical power. The
30 authors analyzed baseline data collected in 2001 and 2003 from 998 participants in the Osaka
31 Maternal and Child Health Study in Japan. The prevalence of current asthma, defined as a self
32 report of medical treatment for asthma in the last 12 months, was 2.1% (N= 21). A prior
33 diagnosis or treatment prior to the last 12 months was not assessed. However, they did report an
34 increased risk of atopic eczema treated in the last 12 months. Median formaldehyde levels were
35 24 ppb with a maximum of 131 ppb. The odds ratio from atopic eczema for women exposed to
This document is a draft for review purposes only and does not constitute Agency policy.
4-56 DRAFT—DO NOT CITE OR QUOTE
-------
1 >47 ppb formaldehyde was 2.25 (95% CI 1.01-5.00) in multiple logistic regression models
2 adjusted for age, gestation, parity, family history of asthma, family income, education, mite
3 antigen level in house dust, and season. The association was significantly higher among those
4 with no family history of allergy.
5 The association between formaldehyde and asthma also has been studied by examining
6 occupational exposures (Fransman et al., 2003; Malaka and Kodama, 1990) and school-related
7 exposures (Zhao et al., 2008; Smedje and Norback, 2001; Norback et al., 2000). The two
8 occupational studies examined the respiratory health of plywood workers (Fransman et al., 2003;
9 Malaka and Kodama, 1990). The most recent of these was conducted at a plywood mill in New
10 Zealand by Fransman et al. (2003). Of an estimated 170 workers approached by site managers or
11 team leaders, 112 workers agreed to participate (66%). Personal samples of formaldehyde
12 exposure were taken for 22 workers and job titles were categorized into low and high exposure
13 groups using the median formaldehyde concentration. The geometric mean level of
14 formaldehyde was 0.08 mg/m3 (65 ppb) and the majority of samples were above the limit of
15 detection which was reported to be 0.03 mg/m3 (24 ppb). Formaldehyde exposure was
16 categorized into low and high groups for 38 and 11 workers who had the same job title as those
17 who had carried samplers. Compared with those with low levels of formaldehyde exposure, the
18 odds ratio for asthma defined as woken by shortness of breath in the last 12 months, asthma
19 attack in the last 12 months, or current asthma medication among workers with high levels of
20 exposure was 4.3 ([95% CI]: 0.7-27.7]). An association was not seen when examining
21 formaldehyde exposure and use of asthma medication. Asthma prevalence also was evaluated in
22 relation to terpene, inhalable dust, abietic acid and endotoxin, and a higher odds ratio was
23 suggested only for terpene categorized into low and high exposure (OR = 2.0 [95% CI]: 0.6-6.8).
24 The second study of plywood workers was completed in Indonesia (discussed in 4.1.1.2).
25 Background levels of formaldehyde ranged from 0.003 to 0.07 ppm. The highest concentration
26 of formaldehyde detected in an air sample was in the particleboard unit (range 1.16 to 3.48 ppm).
27 Asthma, which was defined as "have you ever had an attack of wheezing that made you feel
28 short of breath,?" was found to be positively associated with formaldehyde exposure (Malaka
29 and Kodama, 1990). The estimated odds ratio, controlling for age, smoking status, and dust for
30 the 93 exposed and 93 unexposed workers was 6.31 (p < 0.01).
31 Studies of exposure to formaldehyde at schools have been performed in China (Zhao et
32 al., 2008) and in Sweden (Smedje and Norback, 2001). In a cross-sectional study from China
33 (Zhao et al., 2008), the 7 day mean level of formaldehyde in 31 classrooms at 10 junior high
34 schools was reported to be 2.3 |ig/m3 (range 1.0-5.0 |ig/m3) indoors and 5.8 |ig/m3 (range
35 5.0-7.0 |ig/m3) outdoors. The prevalence of cumulative and doctor diagnosed asthma reported
This document is a draft for review purposes only and does not constitute Agency policy.
4-57 DRAFT—DO NOT CITE OR QUOTE
-------
1 by the 1993 children (90% of eligible) was 1.8% and 1.2%, respectively. In models controlling
2 for outdoor concentrations of formaldehyde, nocturnal attacks of breathlessness were
3 significantly associated with indoor formaldehyde (OR 2.72, 95% CI: 1.03-7.18). In models
4 controlling for indoor concentrations of formaldehyde, cumulative asthma was associated with
5 outdoor formaldehyde levels (OR 4.61, 95% CI 1.09-19.5). An increased risk for daytime
6 attacks of breathlessness also was observed (OR 1.29, 95% CI 0.99-1.68). The 3 level
7 hierarchical logistic models also adjusted for age, sex, parental asthma or allergy and home
8 environmental factors (environmental tobacco smoke etc), and indoor and outdoor
9 concentrations of sulfur dioxide, nitrogen dioxide, and ozone. Moreover, new furniture in the
10 home (a potential source of formaldehyde exposure) was associated with wheeze or whistling in
11 the chest (OR 1.76, 95% CI 1.10-2.81) and daytime attacks of breathlessness (OR 1.31, 95% CI
12 1.0-1.72). In Sweden (Smedje and Norback, 2001), the mean level of formaldehyde measured
13 indoors in school classrooms in 1993 and 1995 were higher than those measured by Zhao et al.
14 (2008) (mean 4, range <5.0-72 |ig/m3).
15 The Swedish investigators (Smedje and Norback, 2001) conducted a 4 year follow-up
16 study among 1732 students from 39 schools aged 7-13, who completed a mailed questionnaire in
17 1993. Both questionnaires were completed by 1347 students, 66% of those invited in 1993. In
18 1993, 6.6% reported having physician diagnosed asthma (N= 89) and 34% (N= 589) reported a
19 history of atopy defined by an affirmative answer regarding either childhood eczema, allergy to
20 pollens or pet dander. The 4 year incidence of physician diagnosed asthma was 4.5% (N= 56).
21 This study did not report an association between formaldehyde exposure and the incidence of
22 asthma (OR 1.2 [95% CI: 0.8-1.7]) among the whole study population. However, when the
23 investigators stratified on history of atopy, they reported that among 22 children without a
24 history of atopy, a new diagnosis of asthma was significantly more likely at higher
25 concentrations of formaldehyde (OR 1.7 per 10 |ig/m3 [95% CI: 1.1-2.6]) and at higher total
26 concentrations of mold (OR = 4.7 per 10-fold increased in total molds [95% CI: 1.2-18.4] in the
27 classroom air. The finding for adverse effects of formaldehyde and mold did not appear to
28 control for the other exposure and no information on the potential correlation between the two
29 exposures was provided.
30 In order to evaluate the potential for confounding of the reported formaldehyde
31 association by the reported mold association, the magnitude of effects must be compared on an
32 appropriate scale since the magnitude of an odds ratio depends on the magnitude of the change in
33 exposure level that is expected to produce increased risk. Standardizing the units to the reported
34 geometric mean standard deviation, the result for formaldehyde (GSM = 2.3 |ig/m3) is
This document is a draft for review purposes only and does not constitute Agency policy.
4-58 DRAFT—DO NOT CITE OR QUOTE
-------
1 OR2 = 1.13 per GSD and the result for mold is OR3 = 1.02 for a comparison of risks at the GSM
2 to 10*GSM and OR4 = 1.06 for a comparison of risks at the minimum value of total molds
3 (5* 103/m3) to 10*minimum. As it appears that the magnitude of the formaldehyde effect is
4 substantially stronger than that of the mold effect (following standardization of exposure
5 increment) it can be concluded that the reported formaldehyde effect could not have been the
6 spurious result of uncontrolled confounding by mold. Unfortunately the logistic regression
7 models did not account for the correlated formaldehyde concentrations for children by
8 classroom.
9 A recent meta-analysis of formaldehyde exposure and asthma in children (McGwin et al.,
10 2010) identified seven peer-reviewed studies providing quantitative results and summarized
11 those findings. Odds ratios and confidence intervals were abstracted and effect estimates were
12 standardized to odds ratios per 10 |ig/m3. Funnel plots were used to assess publication bias and
13 did not show such a bias. Fixed- and random-effects models were used to calculate pooled ORs
14 and 95% confidence intervals following a test of heterogeneity. A fixed-effect model assumes
15 that all the individual studies provided estimates of the same effect or slope while the random-
16 effect model allows for different effects or slopes in the source studies that may reflect difference
17 in baseline risk factors within in the study populations. The authors preferred the fixed-effect
18 model when heterogeneity was lower and the random-effect model was preferred when the data
19 were more heterogeneous. Both models were presented as the degree of heterogeneity, measured
20 by the Q test and I2 statistic, which indicated the presence of moderate heterogeneity. However,
21 the Q test value of 14.28 (p < 0.0001) and the I2 statistic of 51% met the authors definition of
22 sufficiently heterogeneous to prefer the random-effect model results.
23 Of the seven studies that were included in the meta-analysis, six reported increased risks
24 of asthma associated with exposure to formaldehyde. The results of the random-effect model
25 results showed an overall effect estimate of OR = 1.17 (95% CI: 1.01-1.036) (see Figure 4-2).
26 The three studies with the highest statistical weights based on the inverse of the variance of the
27 study ORs were for the studies by Rumchev et al. (2002), Garrett et al. (1999) and Krzyzanowski
28 et al. (1990). Higher weights are reflected by narrower confidence intervals in these studies
29 which implied that they were able to estimate effects with greater precision and so were assigned
30 greater weight in the meta-analysis. The authors (McGwin et al., 2010) noted that an influence
31 plot revealed that the study by Rumchev et al. (2002) may have had 'undue influence on the
32 study data' and recomputed the random effects model without that study. The authors suggest
33 that one difference is that this study is unique in focusing on very young children. Excluding
2 OR per GSD = exp[ln(OR per ng/m3)/10 ng/m3 * 2.3 ng/m3] = exp[ln(1.7)/10*2.3] = 1.13.
3 OR per GSD = exp[ln(OR per 10-fold increase)/(9*GSM)*1.6 ng/m3] = exp[ln(4.7)/162*1.6] = 1.02.
4 OR per GSD = exp[ln(OR per 10-fold increase)/(9*Minimum)*1.6 ng/m3]=exp[ln(4.7)/45*1.6] = 1.06.
This document is a draft for review purposes only and does not constitute Agency policy.
4-59 DRAFT—DO NOT CITE OR QUOTE
-------
1 Rumchev et al. (2002), the OR = 1.24 (95% CI: 1.07-1.45) was somewhat higher than the
2 OR = 1.17 for all the studies.
Study QR(95%Cli OR 195% CD Weight %
Krzyzanowskietal.1990 — | 1.070(0.810-1.420) 15.0
Smedjeetal.1997 1 2.590(1.090-6.140) 3.0
Garrettet 31.1999 —I— 1.270(1.040-1.550) 22.0
Smedje and Norback 2001 --i— 1.200(0.820-1.750) 11.0
Rumchev etal. 2002 I 1.030(1.020-1.040) 35.0
Mi eta 1.2006 —|— 1.300(0.960-1.750) 14.0
Zhao eta 1.2008 (school) —| 0.120(0.010-17.66) 0.0
Zhao eta 1.2008 (outdoor)
Overall (random)
561.6 (0.090-357 E4i 0.0
1.170(1.010-1.360)
0.1 1 10
Decreased risk Increased risk
Figure 2. Forest plot of the relative risk estimates and their 95% CIs from the studies included in the meta-
analysis of the association between formaldehyde exposure and asthma in children based o|n a random-
effects model.
4
5
6 Figure 4-2. McGwin Forest plot of relative risk estimates and 95% CIs from
7 studies included in a meta-analysis of formaldehyde exposure and asthma in
8 children based on the random effects models.
9
10 Source: McGwin et al. (2010).
11
12
13 Separate random-effects were fit for the six studies in which the ORs were for self-
14 reported asthma yielding an OR = 1.26 (95% CI: 0.97-1.64) and for the two studies that used
15 diagnosed asthma OR =1.12 (0.88-1.44). Meta-analytic results stratified by study design
16 yielded an OR = 1.25 (95% CI: 1.08-1.44) for the cross-sectional studies. This systematic
17 review of the literature on asthma and formaldehyde provide evidence of a concentration-
18 dependent increased risk of asthma (prevalence and incidence) associated with increased
19 concentrations of formaldehyde.
20 Garrett et al. (1999) also evaluated the prevalence and severity of allergic sensitization to
21 12 common allergens and reported increased prevalence with increasing formaldehyde
22 concentration in the home. A respiratory symptom score, developed using responses by parents
23 to a validated respiratory questionnaire during an interview, also was increased. The frequency
24 of each respiratory symptom reported during the past year was categorized into four groups
This document is a draft for review purposes only and does not constitute Agency policy.
4-60 DRAFT—DO NOT CITE OR QUOTE
-------
1 (none, 1-3 times, 4-12 times, and >12 times) with a score of 0-3, and scores for the eight
2 symptoms (cough, cough in the morning, shortness of breath, waking due to shortness of breath,
3 wheeze/whistling, asthma attacks, chest tightness, and chest tightness in the morning) were
4 combined to construct a total symptom score for each child. Mean respiratory symptom scores
5 increased with categories of the highest of four seasonal 4-day measurements of formaldehyde in
6 a multiple regression model after adjusting for parental asthma and allergy and interactions. For
7 the atopy endpoints, severity/prevalence of allergic sensitization to 12 common allergens was
8 increased in the medium (20-50 ug/m3) and high (>50 ug/m3) exposure groups relative to the
9 low (<20 ug/m3) exposure group, based on the highest of four seasonal 4-day formaldehyde
10 measurements in the home (p < 0.001). The proportion of atopic children by three categories of
11 average formaldehyde levels in bedrooms was 0.50, 0.59 and 0.74 for <10 ug/m3, 10-30 ug/m3,
12 and >30 ug/m3, respectively (test for trend, p = 0.06). The proportion by highest formaldehyde
13 level was 0.33, 0.64, and 0.75 for <20 ug/m3, 20-50 ug/m3, and >50 ug/m3, respectively (test for
14 trend, p < 0.001). In logistic regressions, the crude association for atopy with an increase in
15 bedroom formaldehyde concentration per 10 ug/m3 was OR = 1.34 which increased when
16 adjusted for parental asthma and gender to an odds ratio of 1.40 per 10 ug/m3 (95% CI:
17 0.98-2.00). Thus, parental asthma was not a confounder of the association between
18 formaldehyde and prevalence of atopy. The adjusted odds ratio for atopy with an increase in
19 highest recorded formaldehyde per 20 ug/m3 was 1.42 (95% CI 0.99-2.04). Passive smoking,
20 the presence of pets, indoor nitrogen dioxide concentrations, airborne fungal spores and house-
21 dust-mite allergens also did not influence the effect estimates and were unlikely to be
22 confounders. The association between formaldehyde concentrations and severity of allergic
23 sensitization was analyzed using two measures, number of positive skin prick tests and the ratio
24 of wheal diameters after skin pricks with allergens compared to the histamine wheal size.
25 Average levels of both measures of severity were higher in the two higher formaldehyde groups
26 compared to the lowest group (p < 0.05). Further, both measures were linearly related to
27 increasing formaldehyde categories in regression models controlling for parental asthma and
28 allergy and sex. The authors reported that 74% of the children had lived in the residences for at
29 least 5 years and 36% had lived there since birth. While the statistical analysis did not
30 specifically account for correlations in children living in same home (siblings), this
31 methodological limitation would not have induced a bias in the estimate of the effects, rather, the
32 likely impact would have been to underestimate the true variability in the effect estimate due to
33 the unlikely assumption that siblings can be treated as completely independent individuals.
34 Two controlled exposure studies in humans investigated whether exposure to a low level
35 of formaldehyde would enhance inhaled allergen responses among already atopic individuals
This document is a draft for review purposes only and does not constitute Agency policy.
4-61 DRAFT—DO NOT CITE OR QUOTE
-------
1 (Casset et al., 2006; Ezratty et al., 2007). These studies are also described in Section 4.1.1.2.2 on
2 pulmonary function. Casset et al. (2006) evaluated the effect of formaldehyde exposure on the
3 bronchial response to dust mite allergen in sensitized asthma patients. The study included
4 19 nonsmoking subjects (12 women and 7 men) ages 19-35 years with mild asthma. Subjects
5 had positive skin prick tests and specific IgE to Dermatophagoides pteronyssinus (dust mites).
6 Subjects had not had a respiratory tract infection in the two weeks priors to the testing.
7 Individuals sensitized to pollen were studied outside the relevant pollen season and people
8 sensitive to pet allergens did not have pets at home. The 30 minute crossover exposures to
9 100 |ig/m3 formaldehyde and air were randomly assigned with a three-week washout period in
10 between. Subjects underwent mite allergen challenge immediately after exposure to
11 formaldehyde or air to determine the dose of allergen that resulted in a 20% reduction in FEVi
12 (PD20). The dose which induced early-phase bronchial response was significantly lower in the
13 subjects exposed to formaldehyde (34.3 ng vs. 45.4 ng: p < 0.05). Late-phase bronchial response
14 was measured over the 6 hours following the dust mite challenge by comparing FEVi. The
15 maximum percentage FEVI reduction observed was significantly higher after exposure to
16 formaldehyde 15% vs. 11% (p < 0.05) (Casset et al., 2006).
17 While the study by Casset et al. (2006) clearly showed that acute formaldehyde exposure
18 (100 |ig/m3) enhanced both early-phase and late-phase bronchial responsiveness to mite allergen
19 in mite-allergen sensitized people with asthma, a subsequent study with a similar protocol did
20 not duplicate this finding. Ezratty et al. (2007) evaluated the response of asthmatics to inhaled
21 allergen after a 60 minute exposure to 500 ug/m3 (0.4 ppm) formaldehyde. The 12 subjects
22 (7 men and 5 women) were 18-44 years old, were nonsmokers, and were diagnosed with
23 intermittent asthma and allergy to pollen. No subjects had contracted an upper respiratory
24 infection for at least 4 weeks before the study. The crossover exposures (60 minutes) to filtered
25 air and 500 ug/m3 formaldehyde occurred 2 weeks apart in random order. Allergen inhalation
26 challenge, using an extract of 5 grass pollen allergens, was conducted immediately after each
27 exposure and the dose producing a 15% decrease in FEVI was determined (PDisFEVl).
28 Responsiveness to methacholine was determined 8 hours after the allergen inhalation challenge
29 ended. Lung function measurements were taken using a spirometer before, during and 8 hours
30 after the end of the allergen challenge. In addition, PEF and FEVI were measured with a
31 portable spirometer every 15 minutes during exposures to filtered air or formaldehyde and every
32 hour until the methacholine challenge. The authors reported that pulmonary function, expressed
33 as percent predicted, was not affected by the formaldehyde exposure, although the data were not
34 presented. The median PDi5FEVl for the allergen challenge was 0.80 (0.15-2.0) index of
35 reactivity (IR) after formaldehyde exposure and 0.25 (0.10-2.0) IR after the filtered air exposure
This document is a draft for review purposes only and does not constitute Agency policy.
4-62 DRAFT—DO NOT CITE OR QUOTE
-------
1 (p = 0.06). The ratio of response after formaldehyde exposure compared to filtered air exposure
2 was 1 in 7 subjects and higher in 5 subjects. The PD20 for methacholine challenge was 0.23
3 (0.01-3.6) and 0.17 (0.03-4) (p = 0.42). The PD20 for methacholine for formaldehyde compared
4 to filtered air was lower in 3 subjects, higher in 4 subjects, and not changed in 5 subjects.
5 There are multiple potential explanation for these seemingly conflicting findings in
6 Casset et al. (2006 and Ezratty et al. (2007) which include the small sample size of these studies,
7 the differencing in the particular allergen tested and difference in the protocols. It is possible
8 that the 12 subjects studied by Ezratty et al. (2007) may not have included individuals who were
9 especially susceptible to formaldehyde. In the Casset et al. (2006) study, the investigators mad
10 specific mention of the size of the particles used for the allergen challenge. They commented in
11 the discussion that dosimetric models of inhaled formaldehyde show that the flux is very large in
12 the first bronchial generations and then decreased rapidly. Casset et al. (2006) used an aerosol
13 with large particles (Mass Median Aerodynamic Diameter of 11.1 |im) specifically because these
14 large particles are deposited in the large airways where formaldehyde flux is higher. Ezratty et
15 al. (2007) did not report the size of the particle used in their challenge but both studies did report
16 the types of dosimeters jet nebulizers that were used and these were different. In a study
17 unrelated to formaldehyde, Praml et al. (2005) compared the physical and biologic doses of
18 methacholine for different nebulizers including the type used by Ezratty et al. (2007). Praml et
19 al. (2005) compared the airway responsiveness of 34 subjects using two types of nebulizers and
20 found that in 17 subjects, neither system caused a 20% decrease in FEV1, while among
21 8 subjects, both systems were able to provoke a 20% decrease. The remaining 9 subjects
22 responded to only one type of nebulizer. Using the same protocol and the same methacholine
23 agent produced different results based on the type of nebulizer. It may be that the results of
24 Casset et al. (2006) and Ezratty et al. (2007) which did have other differences can be explained
25 thusly. It may also be a difference in particle size plays a role if the pollen allergens used in
26 Ezratty et al. (2007) were smaller and penetrate the lung beyond the upper lung where
27 formaldehyde exposure are greater.
28 Multiple occupational cases reports have documented that formaldehyde can cause the
29 onset of asthma and an epidemiologic study of residential exposures has shown that the risk of
30 incident physician-diagnosed asthma is associated with formaldehyde in a concentration-
31 response relationship after controlling for multiple personal characteristics and other
32 coexposures. The regression slope was statistically significant and increased odds ratios were
33 reported at formaldehyde concentrations of 50-59 ug/m3 becoming statistically significant at
34 60 ug/m3 (OR = 1.39). The findings of Rumchev et al. (2002; 2004) are consistent with a causal
3 5 association of formaldehyde on the incidence of asthma.
This document is a draft for review purposes only and does not constitute Agency policy.
4-63 DRAFT—DO NOT CITE OR QUOTE
-------
1 Other studies of residential exposure to formaldehyde that evaluated the prevalence of
2 asthma diagnosed by a doctor among children support this finding (Krzyzanowski et al., 1990;
3 Garrett et al., 1999). Two studies reporting null results evaluated self-reports of asthma, a less
4 precise measure of asthma (Gee et al., 2005; Tavernier et al., 2006; Palczynski et al., 1999).
5 Tavernier et al. did not report the range of formaldehyde levels evaluated so the variation of
6 concentrations evaluated is not known. Among the 187 children aged 15 years or less studied by
7 Palczynski et al., only 9 were defined as having asthma. Studies using other designs also have
8 reported an association of asthma among children (self-reports of physician diagnosed asthma
9 and respiratory symptoms) and formaldehyde concentrations outdoors or in school classrooms
10 (Zhao et al., 2008; Smedje and Norback, 2001). While Smedje and Norback did not observe an
11 association of asthma incidence with formaldehyde exposure in a four year follow up of students
12 who had completed a mailed questionnaire, a subgroup with no history of atopy reported a new
13 diagnosis of asthma associated with formaldehyde. Mean formaldehyde concentrations in the
14 school classrooms were 8 ug/m3 (Geometric mean [GSD] 4 ug/m3 [2.3]).
15 Many of these studies were evaluated quantitatively in a meta-analysis by McGwin et al.
16 (2010). McGwin et al indicated a significant positive association between formaldehyde
17 exposure and childhood asthma both with the Rumchev study (2002) and without it. However,
18 the magnitude of the effect reported by Rumchev et al. (2002) may be smaller than that of other
19 studies in a linear model. As peak expiratory flow rate is a diagnostic criterion of physician-
20 diagnosed asthma incidence, the results of the residential epidemiology study by Krzyzanowski
21 et al. (1990) showing statistically significant decrements in PEFR associated with increased
22 formaldehyde concentration are strongly supportive. Formaldehyde exposure has also been
23 associated with an increased severity of asthma symptoms in children (Garrett et al. 1999;
24 Delfmo et al., 2003).
25 Studies of exposure among adults are suggestive of an association with asthma.
26 Residential exposures at average concentrations of 29 ug/m3 were associated with nocturnal
27 breathlessness (Norback et al., 1995) and occupational levels above 1 ppm were associated with
28 self-reported asthma (Malaka and Kodama, 1990). A null study in New Zealand included very
29 few participants resulting in wide confidence intervals (Fransman et al., 2003).
30 Garrett et al. observed an increase in the prevalence and severity of allergic sensitization
31 to common allergens associated with formaldehyde at concentrations of 20 ug/m3 and above.
32 The role of formaldehyde in exacerbating allergic responses to common allergens among atopic
33 individuals was demonstrated by Cassett et al. (2006) using formaldehyde concentrations of
34 100 ug/m3. Another study using different allergens, dosimeters, and study protocol did not
35 report an effect of formaldehyde on allergic responses (Ezratty et al., 2007). Finally, the report
This document is a draft for review purposes only and does not constitute Agency policy.
4-64 DRAFT—DO NOT CITE OR QUOTE
-------
1 by Smedje and Norback (2001) of increased incidence of asthma over a four year follow-up of a
2 cohort of school children among children without a history of atopy is consistent with a role of
3 formaldehyde in increasing sensitivity to allergens.
4
5 4.1.1.4. Respiratory Tract Pathology
6 Formaldehyde-induced respiratory tract pathology includes inflammation, rhinitis, goblet
7 cell hyperplasia, metaplastic changes, squamous cell hyperplasia, and impaired mucociliary
8 transport. Formaldehyde may bind to the trigeminal nerve and trigger the release of neurogenic
9 mediators of inflammation that result in tissue edema, lacrimation, mucus production and
10 leukocyte infiltration. How much inflammation, hyperplasia, and metaplastic change are due to
11 sensory irritation-induced inflammatory responses compared with formaldehyde-induced direct
12 cell damage cannot be distinguished. Increased mucus flow and metaplastic changes may
13 progress in relation to the concentration and duration of exposure to protect the underlying
14 tissue. When the exposure exceeds protective and defensive mechanisms, permanent damage
15 results (Swenberg et al., 1983). Nonetheless, these changes serve as a sensitive indicator of
16 formaldehyde exposure, since they occur before gross cellular damage and focal lesions
17 (Monticello et al., 1989), and potentially suggest a point at which the concentration and duration
18 of exposure exceed the protective nature of local responses (increased mucus flow, goblet cell
19 hyperplasia, squamous metaplasia, etc.) (Swenberg et al., 1983). A number of human studies
20 have reported nasal lesions associated with exposure to formaldehyde (Pazdrak et al., 1993;
21 Ballarin et al., 1992; Boysen et al., 1990; Holmstrom et al., 1989c; Edling et al., 1988), while
22 other studies have documented changes in mucociliary clearance and activity (Holmstrom and
23 Wilhelmsson, 1988; Andersen and Molhave, 1983). These studies are summarized below.
24
25 4.1.1.4.1. Nasal lesions.
26 Ballarin et al. (1992) did a case-control study of 15 workers from a plywood factory
27 where urea-formaldehyde glue is used. Mean levels of formaldehyde exposure (8-hour average)
28 were estimated to be 0.09, 0.1, and 0.39 mg/m3 in three regions of the facility (sawmill, shearing
29 press, and warehouse, respectively). Nasal respiratory samples were obtained. Stained cells
30 were scored for histopathology. Cytology examination revealed increased squamous metaplasia
31 cells in 10 of 15 (67%) factory workers (with an average severity score of 2.3) compared with
32 one of 15 (6%) controls (with an average histology severity score of 1.6). In addition, one
33 formaldehyde exposed worker (n = 15) exhibited mild dysplasia and had the highest severity
34 score (3.0). Authors suggest that these results may be due to chronic irritation of the nasal
35 respiratory mucosa. This small study reported only incidence of lesions and did not score based
This document is a draft for review purposes only and does not constitute Agency policy.
4-65 DRAFT—DO NOT CITE OR QUOTE
-------
1 on severity of lesions. The lesion incidence was not reported in relation to dose, so no dose-
2 response relationship could be determined, precluding the establishment of a point of departure
3 (POD).
4 Holmstrom et al. (1989c) collected nasal biopsy samples from 36 workers not exposed to
5 formaldehyde and 70 workers exposed to formaldehyde at a median concentration of 240 ppb.
6 Nasal biopsy samples were scored on a 0-8 range with normal respiratory epithelium as 0 and
7 carcinoma as 8. Observed histologic changes included loss of cilia, goblet cell hyperplasia, and
8 cuboidal and goblet cell metaplasia replacing normal columnar epithelium. The incidence
9 associated with each histologic change was not reported and cannot be compared between
10 formaldehyde-exposed and control individuals. Moreover, these biologically relevant changes
11 were not analyzed independently in the analysis. The mean scores were 1.56 (range, 0-4) for the
12 control group and 2.16 (range, 0-4) for the formaldehyde-exposed group. Although the range of
13 scores in the controls and formaldehyde-exposed groups were the same (0-4), the difference in
14 mean scores (1.56 versus 2.16) was statistically significant (p < 0.05); scores were worse in the
15 formaldehyde-exposed group. The authors reported no correlation between the duration of
16 exposure and histologic changes and no correlation between smoking habits and biopsy scores.
17 The loss of cilia, goblet cell hyperplasia, and cuboidal and squamous cell metaplasia replacing
18 the columnar epithelium were increased in the group exposed to formaldehyde and is a
19 biologically relevant change. This study provides a lowest-observed-adverse-effect level
20 (LOAEL) of 0.240 ppm for nasal histopathology.
21 Edling et al. (1988) collected nasal biopsy samples from workers (n = 75) exposed to
22 formaldehyde at three plants (workers in two of these plants were also exposed to wood dust)
23 compared with a referent group (n = 25). Concentrations ranged from 0.1 to 1.1 mg/m3 (TWA)
24 with peaks of 5 mg/m3. Nasal histology was scored from 0 to 8 by increasing severity, from
25 normal respiratory epithelium (0) to carcinoma (8). A normal respiratory epithelium was noted
26 in 3 of 75 workers. A loss of cilia and goblet cell hyperplasia (scores of 2) was reported in eight
27 workers. Mixed cuboid/squamous epithelium (metaplasia), stratified squamous epithelium, and
28 keratosis were reported in 58 of 75 workers (those with scores of 3, 4, and 5 were combined).
29 Dysplasia (score of 6) was reported in 6 of 75 formaldehyde-exposed workers. None of the
30 workers had lesions that warranted a histologic score higher than 6. Histologic scores did not
31 correlate with duration of exposure but could not be confirmed due to poor reporting. Data from
32 the referent group were not included. A POD could not be determined from this study.
33 Boysen et al. (1990) collected nasal biopsy samples from workers exposed to air (n = 37)
34 or to formaldehyde (n = 37) and sometimes wood dust. The exposed workers were classified
35 into two exposure groups, 0.5-2 ppm and >2 ppm. Nasal biopsy samples were assessed by using
This document is a draft for review purposes only and does not constitute Agency policy.
4-66 DRAFT—DO NOT CITE OR QUOTE
-------
1 a histopathology score range of 0-5, based on the pathology of pseudostratified columnar
2 epithelium (0) to dysplasia (5). Mean pathology scores for the control were decreased compared
3 with the formaldehyde-exposed group (1.4 and 1.9, respectively) but did not reach statistical
4 significance. Little quantitative pathology data were provided, although qualitative histology
5 revealed a range of observed effects from deciliated epithelial cells to mixed stratified cuboidal,
6 squamous epithelium to dysplasia. None of the control samples received histologic severity
7 scores of 4 or 5, indicating that keratinizing stratified squamous epithelium and dysplasia were
8 not observed in controls. A wider variety of histopathologic lesions were reported in exposed
9 workers compared with controls, and a greater number of exposed workers had histologic
10 changes compared with controls. Incidence data for each type of histopathology were not
11 reported, but the authors wrote that the degree of metaplastic alterations was more pronounced
12 among the exposed workers. An upper range for the high concentration group (>2 ppm) was not
13 reported, and median concentrations were not provided.
14 Pazdrak et al. (1993) exposed human subjects (six men, three women) to 0.4 ppm
15 formaldehyde in a chamber for 2 hours. Approximately half of the subjects suffered from skin
16 hypersensitivity to formaldehyde, while the other subjects were healthy. An evaluation of nasal
17 lavage pretest and following formaldehyde exposure revealed that the hypersensitive and healthy
18 groups had similarly elevated eosinophil counts at 0 hours after exposure (from
19 42 x 103 cells/mL to 72 x 103 cells/mL for healthy subjects \p < 0.05] and from
20 39 x 103 cells/mL to 69 x 103 cells/mL for hypersensitive subjects \p < 0.05]). Similar
21 eosinophil levels were also seen in both groups at 3 and 18 hours. Both groups had equivalent
22 increases in lavage albumin and total protein levels following exposure, but basophil counts were
23 unchanged. Based on evidence of formaldehyde-induced inflammation, these data provide a
24 LOAEL of 0.4 ppm for nasal histopathology.
25
26 4.1.1.4.2. Mucociliary clearance.
27 In addition to abnormal nasal histopathology, changes in mucociliary clearance were also
28 observed in some of these studies at similar exposure concentrations. The mucociliary apparatus
29 is an important barrier to infection and exogenous agents and, thus, is considered as a potential
30 adverse effect. These effects may be due to direct interaction of formaldehyde with the mucus
31 itself or to Si-induced inflammation in the nasal tissue that affects mucus production and creation
32 of an effective mucosal barrier.
33 Andersen and Molhave (1983) reviewed five controlled human studies, one of which
34 (Andersen and Lundqvist, 1974) examined mucus flow rate in 16 individuals acutely exposed to
35 0, 0.3, 0.5, 1, or 2 ppm formaldehyde for 4-5 hours in a chamber. Mucus flow rate was
This document is a draft for review purposes only and does not constitute Agency policy.
4-67 DRAFT—DO NOT CITE OR QUOTE
-------
1 decreased in the anterior and middle third of the ciliated mucosa at 0.3 ppm, but statistical
2 significance was not determined. This study included smokers and nonsmokers. The small
3 sample size, potential confounder effect from smoking, and lack of dose-response relationship
4 preclude the establishment of a POD.
5 Holmstrom and Wilhelmsson (1988) demonstrated reduced mucociliary clearance and
6 nasal mucosal swelling in 70 workers exposed to a mean formaldehyde concentration of
7 0.21 ppm, compared with a referent group of store clerks (n = 36) and was further averaged over
8 years of exposure. Mucosal swelling and mucociliary activity was measured in the nasal
9 turbinates. The authors also reported symptoms not only during the weekdays, but also over
10 weekends and vacation periods. Formaldehyde-exposed subjects self-reported significantly more
11 nasal discomfort, eye discomfort, deeper airway discomfort, and frequent headache than the
12 referent group. Groups exposed to formaldehyde had more pronounced mucosal swelling
13 (10.7 nasal resistance score) compared with the reference group (6.5 nasal resistance score).
14 This difference persisted when data were normalized for differential nasal congestion in the
15 subjects. Decreased mucociliary activity was seen in 3% of controls and 20% of formaldehyde-
16 exposed subjects and reached statistical significance (p < 0.05). It is not clear whether impaired
17 mucociliary clearance was a consequence of altered cell morphology or increased mucus
18 viscosity. These data provide a LOAEL of 0.21 ppm based on impaired mucociliary clearance.
19 Thus, mild nasal epithelial lesions observed in formaldehyde-exposed workers have been
20 observed consistently at levels of about 0.20 ppm to about 2 ppm (Boysen et al., 1990;
21 Holmstrom et al., 1989; Edling et al., 1988). Of these, Holmstrom et al. (1989) and Edling et al.
22 (1988) do not appear to be confounded by exposure to wood dust. Nasal biopsy pathology from
23 formaldehyde-exposed workers is consistent with irritant and reactive properties of
24 formaldehyde (Ballarin et al., 1992; Boysen et al., 1990; Holmstrom et al., 1989; Edling et al.,
25 1988; Berke, 1987). Moreover, these findings are supported by results from animal toxicity and
26 pharmacokinetic and anatomical airflow studies, indicating that, at concentrations less than
27 1 ppm, inhaled formaldehyde gas does not reach lower regions of the respiratory tract. Of the
28 available human studies that evaluated histopathology, Holmstrom and Wilhelmsson (1988)
29 appears to be the most robust and sensitive. The study was carefully designed and included a
30 large sample of formaldehyde-exposed subjects who were considered separately from workers
31 exposed to combinations of exposures (formaldehyde and wood dust). Study subjects had been
32 exposed to formaldehyde regularly for many years. The authors reported not only weekday
33 exposures but effects reported on weekends and on vacation. Total exposure was carefully
34 calculated and averaged. The data were controlled for potential confounders, such as smoking.
35 The endpoint of reduced mucociliary clearance has been substantiated by Andersen and Molhave
This document is a draft for review purposes only and does not constitute Agency policy.
4-68 DRAFT—DO NOT CITE OR QUOTE
-------
1 (1983) and Holmstrom et al. (1989). Animal studies have also reported formaldehyde-induced
2 changes on the nasal mucosa and are highlighted in Section 4.2.1.2.
O
4 4.1.1.5. Immunologic Effects
5 Numerous studies have examined the immunologic responses of individuals exposed to
6 formaldehyde. This section will discuss four specific areas related to immunotoxicity after
7 exposure to formaldehyde: increased upper respiratory tract (URT) infections, systemic immune
8 dysfunction, sensitization and atopy, and production of formaldehyde-protein complexes. Some
9 studies report increased incidence of URT infections after exposure to formaldehyde (Lyapina et
10 al., 2004; Krzyzanowski et al., 1990; Holness and Nethercott, 1989). This effect appears to
11 occur independently of systemic immune changes and may be due to damage to the mucosal
12 barrier, thus facilitating pathogen access. A number of studies have investigated the hypothesis
13 that formaldehyde may induce systemic immunomodulation (Ohtani et al., 2004a, b; Erdei et al.,
14 2003; Thrasher et al., 1990, 1987; Pross et al., 1987). Some studies have also evaluated immune
15 system effects by investigating the role of reactive oxygen species (ROS) from respiratory burst
16 associated with immune cells (Lyapina et al., 2004; Gorski et al., 1992) and by assessing
17 chromosomal damage in immune cells (Orsiere et al., 2006; Yu et al., 2005). In addition to the
18 effects of formaldehyde on asthmatics and the potential for formaldehyde exposure to exacerbate
19 asthmatic responses (Pourmahabadian et al., 2006; Herbert et al., 1994; Malaka and Kodama,
20 1990; Alexandersson and Hedenstierna, 1989; Alexandersson et al., 1982), reviewed in
21 Section 4.1.1.2, numerous studies have investigated whether formaldehyde may directly induce
22 sensitization and atopic responses by measuring immunoglobulin E (IgE) levels associated with
23 formaldehyde exposure (Ohmichi et al., 2006; Vandenplas et al., 2004; Doi et al., 2003; Baba et
24 al., 2000; Palczynski et al., 1999; Krakowiak et al., 1998; Wantke et al., 1996a, b; Liden et al.,
25 1993; Salkie, 1991; Grammer et al., 1990; Kramps et al., 1989). Findings are largely negative
26 and suggest that formaldehyde-induced IgE production is not likely. Lastly, studies have
27 investigated the production of formaldehyde-specific antibodies, formaldehyde-albumin
28 complexes, and formaldehyde-heme complexes (Kim et al., 2001; Carraro et al., 1997; Grammer
29 et al., 1993, 1990; Dykewicz et al., 1991; Thrasher et al., 1990). Heme complex formation is not
30 a strict immunologic endpoint but may trigger antibody formation and thus it will be discussed in
31 this section. This section will thus summarize the human studies that have specifically addressed
32 the increased incidence of URT infections, immunotoxic endpoints, atopy and sensitization, and
33 formation of formaldehyde-heme and formaldehyde-albumin complexes.
34
This document is a draft for review purposes only and does not constitute Agency policy.
4-69 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.1.1.5.1. Increased VRT infections.
2 Three studies have investigated the possibility that formaldehyde exposure leads to
3 increased URT infections (Lyapina et al., 2004; Krzyzanowski et al., 1990; Holness and
4 Nethercott, 1989). Lyapina et al. (2004) studied 29 workers who were occupationally exposed
5 occupationally to formaldehyde for an average of 12.7 years through contact with carbamide-
6 formaldehyde glue. The mean values of the average shift concentrations of formaldehyde in the
7 application of carbamide-formaldehyde glue was found to be 0.71 ppm TWA with a range of
8 0.32 to 1.57 ppm. The workers were divided into two subgroups, one (n = 12) that suffered from
9 either a long history (with clinical findings) of chronic mucous inflammation of the URT with
10 multiple relapses and a second group (n = 17) whose URT inflammations were short, acute, and
11 predominantly viral. Twenty-one healthy subjects served as controls. A statistically significant
12 association of self-reported chronic bronchitis and decreased resistance to URT infection was
13 reported in all the exposed workers compared with controls (p = 0.02). Of the workers, 41% had
14 a history of chronic respiratory infection and frequent long-lasting infectious inflammatory
15 relapses (group la). Another group (group Ib) consisted of 17 exposed workers, 12 of whom
16 had no history of recurrent viral infections of the URT. There was a statistically significant
17 association of frequency and duration of inflammatory relapses between groups la and Ib. No
18 dates were provided regarding when these measurements were made or over what period of time
19 they were calculated.
20 Krzyzanowski et al. (1990) measured formaldehyde levels in homes and recorded, by
21 way of a questionnaire, health histories from adult and child residents. Formaldehyde levels
22 were reported from samples taken for two 1-week periods in various rooms of the home (kitchen,
23 living room, subject's bedroom). The average formaldehyde level was 26 ppb in 202 homes, and
24 levels were stratified into homes with exposure levels below 40 ppb, between 40 and 60 ppb, and
25 above 60 ppb. Incidences of doctor-diagnosed chronic bronchitis were more prevalent in
26 children (under age 15) living in homes with higher formaldehyde (>60 ppb) readings in the
27 kitchen (p < 0.001). This effect was more pronounced (p < 0.001) in children simultaneously
28 exposed to environmental tobacco smoke. The prevalence of chronic cough was also increased
29 in adults living in homes with measurable levels of formaldehyde, but data were not shown.
30 Holness and Nethercott (1989) assessed chronic bronchitis in 87 funeral workers, where the
31 average formaldehyde exposure was reported at 0.38 ± 0.19 ppm. Chronic bronchitis was
32 observed in 20 funeral workers (n = 87) exposed to formaldehyde compared with 3 cases of
33 chronic bronchitis in nonexposed referent controls (n = 38).
34 These studies suggest that exposure to formaldehyde may be associated with increased
35 incidence of chronic bronchitis. The mechanism for this association has not been elucidated.
This document is a draft for review purposes only and does not constitute Agency policy.
4-70 DRAFT—DO NOT CITE OR QUOTE
-------
1 Pathogens may gain access to the URT via a compromised mucosal barrier, as has been shown in
2 histopathology studies (see Section 4.1.1.4).
O
4 4.1.1.5.2. Immune function.
5 A number of studies have evaluated the ability of formaldehyde to induce systemic
6 immunotoxic effects (Erdei et al., 2003; Thrasher et al., 1990, 1987; Pross et al., 1987). Some
7 studies have reported altered innate immune responses associated with formaldehyde exposure in
8 immunologically compromised children (Erdei et al., 2003), while others have noted adaptive
9 immune response suppression associated with formaldehyde exposure (Thrasher et al., 1990,
10 1987) and changes associated with alterations to a predominant T—lymphocyte helper 2 (Th2)
11 pattern (Ohtani et al., 2004a, b). In contrast, Pross et al. (1987) did not observe formaldehyde-
12 associated changes in systemic immune function.
13 Erdei et al. (2003) found that Haemophilus influenzae humoral biomarker (H.in.IgG),
14 Klebsiellapneumoniae biomarker (K.pn.IgG), and elevated monocyte concentrations were
15 significantly associated with high formaldehyde concentrations in asthmatic children, compared
16 with nonsensitive children. Briefly, Erdei et al. (2003) compared the immune system responses
17 in 9- to 11-year-old Hungarian school children whose respiratory systems were immunologically
18 compromised (chronic respiratory disease, asthma) and normal children who were exposed to
19 indoor air pollutants, including formaldehyde. In the homes of the children with the highest
20 levels of pollutants, 49.3% of formaldehyde measurements exceeded the Hungarian indoor
21 standard of 0.01 ppm, while 20% exceeded the World Health Organization's (WHO's) suggested
22 indoor level of 0.09 ppm. The authors excluded from consideration all measurements that
23 exceeded WHO's air quality guidelines in one unidentified city to prevent a "city-related bias,"
24 since these measurements occurred entirely in that city. The average formaldehyde
25 concentration in the 123 homes tested was 14 ppm with a range of 0.5 to 46 ppm. H.in.IgG and
26 K.pn.IgG were significantly associated with high formaldehyde concentrations (p < 0.013 and
27 p< 0.049, respectively) in sensitive children compared with nonsensitive children. These
28 markers were also correlated with high levels of nitrogen dioxide, the number of cigarettes
29 smoked, and exposure to paint, volatile organic compounds, and solvents. Additionally, indoor
30 formaldehyde exposure was significantly associated with increased monocyte concentrations
31 (p < 0.017) that are important to the innate immune response (inflammation) in diseased tissue.
32 The authors concluded that the elevation of immune biomarkers in sensitive children with
33 respiratory disease is likely the result of high concentrations of toxic indoor air pollutants,
34 including formaldehyde.
This document is a draft for review purposes only and does not constitute Agency policy.
4-71 DRAFT—DO NOT CITE OR QUOTE
-------
1 Thrasher et al. (1987) assessed the effects of formaldehyde exposure on cellular
2 immunity and antibody formation in eight exposed and eight unexposed individuals. The
3 exposed group consisted of three males and five females. Seven of the exposed individuals had
4 resided in mobile homes for periods ranging from 2 to 7 years; the eighth was a laboratory
5 worker who resided in a newly decorated, energy-efficient apartment. Air monitoring in four of
6 the homes revealed formaldehyde vapor concentrations ranging from 0.07 to 0.55 ppm. Venous
7 blood samples were collected from all subjects and T- and B-cells were counted and monitored
8 for blastogenesis. When IgG and IgE antibodies to formaldehyde were monitored in serum, no
9 IgE antibodies to formaldehyde were detected in exposed or control subjects. IgG antibody liters
10 in exposed subjects ranged from 1:8 to 1:256 but essentially were undetected (1:4) in seven of
11 the controls. T- and B-cell numbers were significantly lower (p < 0.05) in mobile home residents
12 (48 and 12.6%, respectively) compared with those of control subjects (65.9 and 14.75%,
13 respectively). As determined by incorporation of 3H-thymidine into 48-hour unaltered
14 lymphocytes, phytohemagglutinin-stimulated T- and B-cell blastogenesis was significantly
15 depressed (p < 0.01) in cells of mobile home residents compared with those of control subjects
16 (17,882 and 28,576 cpm, respectively). Thrasher et al. (1987) concluded that exposure to
17 formaldehyde decreases the proportion of peripheral T cells.
18 In a later study, Thrasher et al. (1990) evaluated five groups of subjects with varying
19 levels and durations of formaldehyde exposure. The groups consisted of (1) asymptomatic
20 chiropractic students exposed during anatomy classes (controls with only intermittent exposure
21 to formaldehyde), (2) mobile home residents, (3) office workers, (4) patients with multiple
22 symptoms who had been removed from the source of formaldehyde for at least a year, and
23 (5) occupationally exposed patients. All groups were assessed for immunologic function via
24 white cell, lymphocyte, and T-cell counts, T-helper/suppressor ratios and B-cell counts. When
25 compared with controls (chiropractic students), the patient groups had significant elevations in
26 formaldehyde antibody liters and B-cell liters.
27 Ohtani et al. (2004a, b) reported effects of exposure to formaldehyde and diesel exhaust
28 particles on cytokine production by human monocyte-derived dendritic cells (MoDCs) and
29 T cells in vitro. Dendritic cells were stimulated with CD40 ligand and interferon (IFN)-y, T cells
30 with anti-CD3/CD28 antibodies. Cytokine proteins and mRNA levels were measured in
31 supernatants by enzyme-linked immunosorbent assay (ELISA) and polymerase chain reaction
32 (PCR), respectively. Formaldehyde and diesel exhaust particles significantly increased tumor
33 necrosis factor (TNF)-a levels and suppressed interleukin (IL)-12p40 protein and mRNA levels
34 in MoDCs. The same treatment suppressed protein synthesis and mRNA expression of IFN-y
35 and IL-10 in T cells. The authors concluded that their findings support a role of formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-72 DRAFT—DO NOT CITE OR QUOTE
-------
1 and diesel exhaust particles in altering the immune response to a Th2-dominant pattern that
2 furthers allergic inflammation. Further details, such as exposure concentrations and
3 experimental protocols, are not available.
4 In contrast, Pross et al. (1987) concluded that formaldehyde does not induce altered
5 immune activity. The authors evaluated the immunologic response of asthmatic subjects
6 exposed to UFFI off-gas products. Subjects consisted of 23 individuals with a history of
7 asthmatic symptoms attributed to UFFI and 4 individuals with asthma unrelated to UFFI off-gas
8 products. All subjects were exposed in an environmental chamber according to the following
9 sequence: (1) room air (placebo) for 30 minutes; (2) 1 ppm formaldehyde gas for 3 hours;
10 (3) UFFI particles (4 um diameter, 0.5 particles/mL) for 3 hours, commencing 48 hours after
11 formaldehyde gas exposure; and (4) UFFI off-gas products for 3 hours, commencing 48 hours
12 after UFFI particle exposure. There was a significant increase in the percentage and absolute
13 number of eosinophils and basophils in the subjects who lived in UFFI homes but no differences
14 between exposure groups with respect to lymphocyte subpopulations either before or after UFFI
15 exposure. However, when T8 suppressor cells were counted, values in the UFFI-exposed group
16 pre-exposure and postexposure, a small but statistically significant (p < 0.01) increase in T8 cell
17 count was observed. The biological significance of this increase in T8 cell count in exposed
18 asthmatics is not known. Pross et al. (1987) concluded that short-term exposure to formaldehyde
19 was not immunosuppressive and did not result in systemic immune reactivity.
20 Respiratory burst from immune cells creates ROS that can incur further cellular damage.
21 Several studies have evaluated, either directly or indirectly, the potential role of ROS as potential
22 mediators of formaldehyde-associated effects, particularly those caused by immune cells. Gorski
23 et al. (1992) measured chemiluminescence resulting from the release of free radicals from
24 granulocytes of healthy and formaldehyde-sensitive (based on anamnesis and a positive patch
25 test) subjects. Thirteen subjects with contact dermatitis who were occupationally exposed to
26 formaldehyde and five healthy volunteers participated in the study. All underwent skin-prick
27 tests for common allergens as well as a histamine inhalation provocation test. Subjects were
28 exposed to 0.5 mg/m3 (0.41 ppm) formaldehyde for 2 hours, and the PEFR was measured
29 immediately before exposure, after 60 and 120 minutes of exposure, and 6 and 21 hours after
30 completion of exposure. Peripheral blood granulocyte chemiluminescence was measured in the
31 presence of luminol. Free radical production was increased significantly within 30 minutes of
32 beginning the exposure in subjects with allergic dermatitis and remained elevated for 24 hours
33 compared with baseline values. Gorski et al. (1992) concluded that granulocyte
34 chemiluminescence did not increase in healthy, formaldehyde-exposed patients but was
This document is a draft for review purposes only and does not constitute Agency policy.
4-73 DRAFT—DO NOT CITE OR QUOTE
-------
1 diagnostic for formaldehyde-sensitive patients. These results also suggest a putative role for
2 oxidative damage associated with formaldehyde exposure, particularly in sensitized individuals.
3 Lyapina et al. (2004) also reported effects of formaldehyde exposure on neutrophil
4 respiratory burst activity (NRBA), the capacity of polymorphonuclear leukocytes to produce
5 reactive oxygen radicals in response to chemical or microbial stimuli using flow cytometry.
6 Briefly, Lyapina et al. (2004) studied 29 workers who were occupationally exposed to
7 formaldehyde for an average of 12.7 years through contact with carbamide-formaldehyde glue
8 with a mean value of the average shift concentration of formaldehyde reported as 0.71 ppm
9 TWA with a range of 0.32 to 1.57 ppm. The workers were divided into two subgroups, one
10 (n = 12) that suffered from either a long history (with clinical findings) of chronic mucous
11 inflammation of the URT with multiple relapses, and a second group (n = 17) whose URT
12 inflammations were short, acute, and predominantly viral. Twenty-one healthy subjects served
13 as controls. A suite of hematological tests and flow cytometric analysis for respiratory burst
14 activity were performed. Although no significant difference was observed in the spontaneous
15 and stimulated NRBA (median percentage of oxidizing cells) between the 29 exposed workers
16 with URT inflammation and the healthy controls (0.83 versus 1.35, respectively), a separate
17 comparison of the NRBA of 12 workers with chronic, repeating URT infections and 17 workers
18 with short, infrequent episodes of URT inflammations was significant (0.45 versus 1.00,
19 p = 0.037). When the NRBA of the group with chronic URT infections (n = 12) was separately
20 compared with that of the healthy controls (n = 21), the results were also significant (0.45 versus
21 l.35,p = 0.012). Individuals with chronic URT infections have reduced NRBA that could be
22 due to formaldehyde exposure. Neutrophils respond to tissue damage or local invasion of
23 microorganisms and act to phagocytize foreign cells. If neutrophilic activity is hampered or
24 altered by formaldehyde exposure, then the ability to fight infection will be diminished, leading
25 to prolonged infection. However, no dose-response pattern of formaldehyde exposure could be
26 determined from this study.
27 Other investigators have reported chromosomal damage in immune cells due to
28 formaldehyde (Orsiere et al., 2006; Yu et al., 2005). Yu et al. (2005) evaluated chromosomal
29 damage in lymphocytes from 151 exposed and nonexposed workers from a plywood factory
30 detected by comet assay. The authors reported that chromosomal damage was statistically
31 elevated in lymphocytes from formaldehyde-exposed workers compared with controls.
32 However, no information on exposure duration or levels was provided. Orsiere et al. (2006)
33 studied DNA damage in lymphocytes from 59 hospital employees with formaldehyde exposures
34 from pathology and anatomy laboratories in five hospitals. Controls were 37 workers from the
35 same hospitals, matched on gender, age, and smoking habits, with no known exposure to
This document is a draft for review purposes only and does not constitute Agency policy.
4-74 DRAFT—DO NOT CITE OR QUOTE
-------
1 genotoxic agents. Study participations were excluded if workers had a history of radio- or
2 chemotherapy or had used therapeutic medications that were known to be mutagenic.
3 Occupational exposure was determined through 15-minute and 8-hour personal air sampling
4 during a typical workday. Mean formaldehyde concentrations were 2 ppm (range:
5 <0.1-20.4 ppm) for 15-minute sampling and 0.1 ppm (range: <0.1-0.7 ppm) for 8-hour
6 sampling. No change in DNA damage was found between the beginning and end of the workday
7 among exposed workers (3.9 ± 0.6 versus 3.6 ± 0.5 relative light units/ng DNA). However,
8 exposed workers had significant elevations in the binucleated micronucleated cell rate (BMCR)
9 per 1,000 cells compared with controls (16.9 ± 9.3 versus 11.1 ± 6.0%; p < 0.001) suggesting a
10 clastogenic response, but BMCR did not appear to be correlated with formaldehyde
11 concentration. Linear regression analysis showed that the effect for exposure remained after
12 adjusting for gender, age, smoking, and drinking habits. For 18 exposed and 18 control workers
13 who underwent cytokinesis-blocked micronucleus assay (CBMA) combined with fluorescent in
14 situ hybridization (FISH) with pan-centromeric DNA probe, results showed that the frequency of
15 micronuclei (MN) containing only one centromere (Cl+MN) was elevated among the exposed
16 compared with unexposed workers (11.0 ± 6.2% versus 3.1 ± 2.4%;p < 0.001). The effect of
17 exposure remained significant after controlling for gender, age, smoking, and drinking habits.
18 Results from Yu et al. (2005) and Orsiere et al. (2006) suggest that formaldehyde exposure may
19 promote chromosomal damage leading to micronucleated lymphocytes.
20 Compromised lymphocyte function may significantly contribute to altered immune
21 status. The mechanism underlying this effect has not been elucidated.
22
23 4.1.1.5.3. Sensitization and atopy.
24 Numerous studies have documented formaldehyde-induced exacerbation of asthmatic
25 responses (Garrett et al., 1999; Kriebel et al., 1993; Delfmo et al., 2003; Norback et al., 195;
26 Cassett et al., 2006; also see Section 4.1.1.2). The mechanism of this effect has not been
27 clarified and has led investigators to assess the potential for formaldehyde to directly induce
28 formation of formaldehyde-specific antibodies, leading to allergic responsiveness. One case
29 report showed systemic allergic reactions (e.g., anaphylaxis) to formaldehyde in a patient
30 undergoing hemodialysis (Maurice et al. [1986] referenced in Thrasher et al. [1990]). Some
31 studies have evaluated the potential association of formaldehyde-specific IgE in already -
32 sensitized individuals (Baba et al., 2000; Palczynski et al., 1999). Other studies have
33 investigated whether formaldehyde can directly induce IgE in nonsensitized individuals. Most of
34 the studies have not identified presence of formaldehyde-specific IgE (Ohmichi et al., 2006;
35 Krakowiak et al., 1998; Grammer et al., 1993, 1990; Kramps et al., 1989; Thrasher et al., 1987)
This document is a draft for review purposes only and does not constitute Agency policy.
4-75 DRAFT—DO NOT CITE OR QUOTE
-------
1 and are summarized below. A few studies (Vandenplas et al., 2004; Doi et al., 2003; Liden et
2 al., 1993) reported positive IgE against formaldehyde, associated with exposure, but the IgE
3 liters were either transient (Vandenplas et al., 2004) or were positive in a small subset of
4 previously sensitized subjects (2 of 15) (Liden et al., 1993). Doi et al. (2003) detected IgE
5 against formaldehyde in two asthmatic children (out of 122 asthmatic children), but the response
6 severity did not correlate with exposure level.
7 Palczynski et al. (1999) evaluated whether exposure to formaldehyde might facilitate
8 specific sensitization to common allergens. The study population was comprised of residents of
9 apartments built in 1989-1990. Only households with children from 5-15 years were eligible
10 for the study. A random sample of 120 apartments was selected in which lived a total of
11 465 persons aged 5-65 years. Individual demographic characteristics and medical histories were
12 determined by questionnaire. Residents were tested, using the skin-prick method, for allergen
13 response to a variety of materials, such as household dust, pollens, and feathers. Total serum IgE
14 levels were measured, and the presence of formaldehyde-specific IgE antibodies was determined.
15 Measured mean levels of formaldehyde were 21.05 ± 8.94 ppb. No significant relationship
16 between respiratory allergy prevalence and indoor exposure to formaldehyde was detected.
17 Significant increases in serum IgE levels were found in children exposed to both environmental
18 tobacco smoke and formaldehyde.
19 Baba et al. (2000) investigated whether production of formaldehyde-specific IgE could be
20 detected in adult asthmatics. Formaldehyde exposure levels were not documented.
21 Formaldehyde-IgE was detected in two asthmatic patients (n = 80), one male and one female, but
22 the titer of IgE did not parallel the severity of the asthmatic responses and could not be linked to
23 formaldehyde exposure. Thus, formaldehyde-specific IgE-mediated allergy was rare in adult
24 chronic asthmatics.
25 Several studies have examined serum for formaldehyde-specific IgE antibodies in groups
26 of formaldehyde-exposed humans (Ohmichi et al., 2006; Krakowiak et al., 1998; Wantke et al.,
27 1996a, b; Salkie, 1991; Grammer et al., 1990; Kramps et al., 1989). While formaldehyde-
28 specific IgE was reported in one study (Wantke et al., 1996a), results from most other studies
29 failed to find a consistently strong association between formaldehyde-specific IgE or IgG
30 antibodies in groups of formaldehyde-exposed humans.
31 Wantke et al. (1996a) detected elevated levels of formaldehyde-specific IgE as
32 determined by Radio Allergo Sorbent Test or RAST, which detects allergen-specific IgE) in 24
33 of 62 8-year-old children who were students in three particleboard-paneled classrooms in which
34 the estimated formaldehyde air concentrations were 0.075, 0.069, and 0.043 ppm. In a health
35 survey, the children reported headaches (29/62), fatigue (21/62), dry nasal mucosa (9/62), rhinitis
This document is a draft for review purposes only and does not constitute Agency policy.
4-76 DRAFT—DO NOT CITE OR QUOTE
-------
1 (23/62), cough (15/62), and nosebleeds (14/62). The number of children with symptoms in each
2 classroom decreased with decreasing formaldehyde concentration (49, 47, and 24, respectively,
3 for the 0.075, 0.069, and 0.043 ppm classrooms). However, the investigators reported that
4 elevated levels of specific IgE did not correlate with the number and severity of symptoms.
5 When the children were evaluated after 3 months in a new school that did not have particleboard
6 paneling and had lower ambient formaldehyde concentrations (0.029, 0.023, and 0.026 ppm), the
7 number of children reporting symptoms decreased significantly from earlier figures, and, when
8 measured in 20 of the children, the mean serum levels of formaldehyde-specific IgE declined
9 significantly compared with premoving mean levels.
10 In contrast, a study by Krakowiak et al. (1998) measured serum IgE levels in asthmatic
11 and healthy subjects as part of a larger study to characterize the mechanism of
12 formaldehyde-induced nasal and bronchial response in asthmatic subjects with suspected
13 formaldehyde allergy. Ten subjects reported to have formaldehyde rhinitis and asthma and
14 10 healthy subjects underwent a 2-hour inhalation challenge in an exposure chamber with
15 formaldehyde at a concentration of 0.5 mg/m3 (0.41 ppm). Formaldehyde-specific serum IgE
16 antibodies were measured, and cellular, biochemical, and mediator changes were assessed in
17 nasal lavage before, immediately after, and at 4 and 24 hours after challenge. Challenges with
18 formaldehyde caused only transient symptoms of rhinitis in both groups. Furthermore, none of
19 the subjects thought to have occupational asthma developed clinical symptoms of bronchial
20 irritation. No specific IgE antibodies to formaldehyde were detected in persons with
21 occupational exposure to formaldehyde. No differences in the nasal response to formaldehyde
22 were found between subjects reported to have occupational allergic respiratory diseases and
23 healthy subjects (p > 0.05). The study showed that inhaled formaldehyde at a level as low as
24 0.5 mg/m3 did not induce a specific allergic response either in the upper or in the lower part of
25 the respiratory tract. In addition, it demonstrated that there was no difference in nasal response
26 to formaldehyde between asthmatic subjects occupationally exposed to formaldehyde and
27 healthy subjects.
28 Similarly, formaldehyde-specific IgE antibodies were detected in only 1 serum sample
29 (out of 86) from four groups of formaldehyde-exposed subjects (Kramps et al., 1989). The
30 subject with detected formaldehyde-specific IgE displayed allergic symptoms. The groups
31 included (1) 28 subjects living or working in places with formaldehyde-containing construction
32 materials (e.g., chipboard) and estimated formaldehyde concentrations ranging from 0.08 to
33 0.37 ppm, (2) 18 occupationally exposed subjects from an anatomy laboratory and in other
34 unspecified industries where air concentrations were not measured, (3) 12 hospital attendants
35 who worked with formaldehyde-sterilized hemodialysis equipment, and (4) 28 hemodialysis
This document is a draft for review purposes only and does not constitute Agency policy.
4-77 DRAFT—DO NOT CITE OR QUOTE
-------
1 patients coming into contact with equipment sterilized with formaldehyde. Other subjective
2 symptoms, such as headache, eye irritation, and respiratory complaints, were reported by
3 24/28 subjects in the construction material group and confirm that formaldehyde is an irritant
4 (reviewed in Section 4.1.1.1). Durations of exposure or length of employment were not reported
5 for the subjects in this study.
6 Grammer et al. (1990) studied the immunologic nature of formaldehyde sensitivity in
7 37 workers who were examined by a group of physicians in response to complaints of
8 formaldehyde-related illness. Air sampling of formaldehyde ranged from 0.003 to 0.078 ppm,
9 but specific levels were not tied to specific workplace areas. Blood samples were collected and
10 assayed for IgE and IgG activity against formaldehyde-human serum albumin (formaldehyde-
11 HSA) and HSA alone. None of the workers had IgG activity against formaldehyde-HS A. Five
12 workers had comparable IgE activity against both formaldehyde-HS A and HAS that was more
13 than twice the normal control serum levels. No IgE antibodies were detected in the other
14 32 workers. The authors concluded that there was no evidence of an immunologically mediated
15 response to formaldehyde in this group of workers.
16 Formaldehyde-specific IgE was not detected in any of a group of 45 medical students
17 before or after the students attended a 4-week anatomy dissecting course (Wantke et al., 1996b).
18 Estimates of ambient air concentrations of formaldehyde ranged from 0.059 to 0.219 ppm
19 (0.124 ± 0.05 ppm; mean ± SD). However, the survey revealed frequencies of irritation
20 symptoms that were consistent with other studies (e.g., itching of the skin in 33/45 students,
21 headache in 15/45, and burning eyes in 13/45).
22 Similarly, Ohmichi et al. (2006) were unable to correlate formaldehyde exposure with
23 specific IgE production among eight students attending a gross anatomy laboratory.
24 Formaldehyde exposure was estimated to range from 0.33 to 1.47 ppm during the laboratory
25 sessions. The sample size was small, and IgE levels varied substantially (ranging from <19 to
26 >5,000 international units/mL). Compared with IgE levels taken 90 minutes prior to the start of
27 the first session, IgE levels measured shortly after the last session and up to 23 days following
28 the last session showed no association with exposure.
29 Salkie (1991) investigated the prevalence of formaldehyde-specific IgE in practicing
30 pathologists who complained of formaldehyde sensitivity. Exposure levels were not reported.
31 Serum samples were assayed for total IgE and formaldehyde-specific IgE. Of the 46 subjects,
32 29 self-reported atopy that was confirmed in 12 subjects by positive IgE. Moreover, 29 subjects
33 complained of formaldehyde-specific sensitivity. However, zero subjects had formaldehyde-
34 specific IgE, and there was no evidence that atopic individuals were more sensitive to
This document is a draft for review purposes only and does not constitute Agency policy.
4-78 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde than nonatopic individuals. The authors noted that atopic individuals may have
2 selectively reduced their exposure to formaldehyde.
3 Vandenplas et al. (2004) evaluated a case study of a 31-year-old male who was
4 accidentally exposed to formaldehyde for 2 hours. The exposure level was not provided. The
5 subject had smoked a pack of cigarettes a day for 13 years and was admitted to the emergency
6 room for asthmatic symptoms. Eight days following exposure, increased levels of
7 formaldehyde-specific IgE antibodies were detected but could not be detected in subsequent
8 assessments.
9 A clinical study by Liden et al. (1993) evaluated IgE-specific antibodies against
10 formaldehyde in 23 subjects who had previously tested positive for skin sensitization by skin
11 prick test. Subjects were exposed to formaldehyde by skin patch (1% formaldehyde in water).
12 Ten of the subjects were classified as atopic Though 15 of 23 of the sensitized subjects were
13 also sensitive to formaldehyde applied by skin patch, formaldehyde-IgE was positive in 2 of
14 15 individuals who were not classified as atopic. No dose-response relationship could be
15 determined from the study design of this study.
16 Doi et al. (2003) conducted a clinical study in 155 children of which 122 were
17 asthmatics. No specific exposure to formaldehyde was documented. IgE against formaldehyde
18 was determined in blood. Formaldehyde-specific IgE was found in two asthmatic children.
19 Thus, while several studies have documented formaldehyde-specific IgE, the occurrence is rare
20 and may be transient. Asthmatic children may be more predisposed to form formaldehyde-
21 specific IgE than nonatopic individuals or adults. The formation of formaldehyde-specific IgE is
22 quite rare.
23
24 4.1.1.5.4. Formaldehyde-albumin and formaldehyde-hemoglobin complexes.
25 Numerous studies have shown that formaldehyde can bind to blood proteins such as
26 hemoglobin (Hb) and human serum albumin (HSA) forming formaldehyde-Hb (Bono et al.
27 2006) and formaldehyde-HSA complexes (Carraro et al., 1997; Grammer et al., 1993, 1990;
28 Dykewicz et al., 1991; Thrasher et al., 1990). Kim et al. (2001) failed to identify IgE against
29 formaldehyde-HSA complexes in one case-control subject following industrial occupational
30 exposure to formaldehyde. These complexes may serve to traffic formaldehyde throughout the
31 bloodstream and throughout the body. While formaldehyde may be too small to engender an
32 immune response, these complexes may be able to trigger formaldehyde-protein-specific
33 antibodies, leading to an immune response, including sensitization.
34 Thrasher et al. (1990) evaluated five groups of subjects as follows with varying levels
35 and durations of formaldehyde exposure: asymptomatic chiropractic students exposed during
This document is a draft for review purposes only and does not constitute Agency policy.
4-79 DRAFT—DO NOT CITE OR QUOTE
-------
1 anatomy classes (controls with only intermittent exposure to formaldehyde), mobile home
2 residents, office workers, patients with multiple symptoms who had been removed from the
3 source of formaldehyde for at least a year, and occupationally exposed patients. All groups were
4 assessed for production of IgG, IgM and IgE class of antibodies against formaldehyde-HS A.
5 The level of all classes of autoantibodies was significantly elevated in patients exposed long-
6 term to formaldehyde. From the data, Thrasher et al. (1990) concluded that exposure to
7 formaldehyde stimulates IgG antibody production to formaldehyde-HSA.
8 Grammer et al. (1990) studied the immunologic nature of formaldehyde sensitivity in
9 37 workers who were examined by a group of physicians in response to complaints of
10 formaldehyde-related illness. Air sampling of formaldehyde ranged from 0.003 to 0.078 ppm,
11 but specific levels were not tied to specific workplace areas. Blood samples were collected and
12 assayed for IgE and IgG reactivity against HSA alone and formaldehyde-HSA. None of the
13 workers had IgG activity against formaldehyde. Five workers had IgE against both HSA alone
14 and against formaldehyde-HSA complexes. No IgE antibodies were detected in the other
15 32 workers. The authors concluded that there was no evidence of an immunologically mediated
16 response to formaldehyde in this group of workers.
17 Grammer et al. (1993) described the evaluation of a worker with bronchospasm
18 symptoms caused by formaldehyde exposure. The worker was evaluated by means of ELISA,
19 cutaneous tests, and methacholine and formaldehyde inhalation challenges. The ELISA showed
20 that the worker had positive IgE and IgG liters to formaldehyde-HSA. The worker also had a
21 positive cutaneous test for formaldehyde-HSA but a negative methacholine challenge at
22 25 mg/mL and negative formaldehyde inhalation challenges at exposure concentrations of 0.3, 1,
23 3, and 5 ppm for 20 minutes. The worker might have developed a positive response if a higher
24 concentration of formaldehyde had been used for the challenge, but it is more probable that the
25 worker's symptoms were not caused by immunologically mediated asthma.
26 Dykewicz et al. (1991) evaluated whether IgE or IgG antibodies to formaldehyde were
27 related to formaldehyde exposure or to respiratory symptoms arising from such an exposure.
28 The authors studied 55 potentially exposed subjects (hospital histology technicians, internal
29 medicine residents, pathology residents, current smokers, and persons with known workplace
30 exposure to formaldehyde) and compared them to controls with no history of formaldehyde
31 exposure. Reported workplace formaldehyde concentrations were 0.2-0.64 ppm for pathology
32 residents, 0.64 ppm for histology technicians, and 0.6-11 ppm for miscellaneous formaldehyde
33 exposure scenarios. Workplace air concentrations were not measured for the other occupations.
34 Occupational exposure to formaldehyde averaged 12.45 years for histology technicians,
35 0.38 years for medical residents, 3.21 years for pathology residents, and 18.34 years for five
This document is a draft for review purposes only and does not constitute Agency policy.
4-80 DRAFT—DO NOT CITE OR QUOTE
-------
1 subjects exposed to formaldehyde in miscellaneous workplaces. Blood samples were analyzed
2 for IgE and IgG reactivity with formaldehyde-HSA complexes. Three subjects had IgE against
3 formaldehyde-HSA; these three and two others had low levels of anti-formaldehyde-HSA IgG.
4 The presence of IgG and IgE antibodies to formaldehyde was not clearly related to formaldehyde
5 exposure or pack-years of smoking. One subject had both IgE and IgG antibodies and also
6 suffered from eye and respiratory symptoms when exposed to formaldehyde at his workplace.
7 However, the authors concluded that they could not establish a relationship between IgE and IgG
8 levels and formaldehyde exposure. This study has several limitations. First, the volunteers
9 (hospital workers) may not be representative of exposed workers in the general population. One
10 of the exposure groups comprised cigarette smokers. Although the study focused on
11 formaldehyde-HSA antibodies, which would be unaffected by the other chemicals, respiratory
12 symptoms among smokers would reflect exposures to the constituents of cigarette smoke.
13 Dykewicz et al. (1991) concluded that immunologically mediated asthma caused by
14 formaldehyde is extremely rare and may not exist at all.
15 Carraro et al. (1997) reported development of a reliable assay to effectively measure
16 formaldehyde-HSA complexes in smokers, ex-smokers, and nonsmokers. A correlation between
17 formaldehyde-HSA antibodies and smoking status was detected with highest percentage of
18 individuals for polyclonal antibodies detected in smokers and lowest percentage detected in
19 nonsmokers . This study did not correlate formaldehyde exposure and formaldehyde-HSA
20 antibodies.
21 Given that formaldehyde is a sensory irritant that is particularly bothersome to
22 individuals with compromised lung function or asthma, numerous studies have assessed the
23 ability of formaldehyde to induce immunotoxic effects (Wieslander et al . 1997; Norback et al .
24 1995; Grammer et al . 1993; Pross et al . 1987). Some studies have documented increased rates
25 of chronic bronchitis and upper respiratory tract infections associated with exposure to
26 formaldehyde, which suggests a possible immunomodulatory effect (Gorski and Krakowiak
27 1991; Krzyzanowski et al . 1990; Malaka and Kodama 1990; Holness and Nethercott 1989;
28 Tuthill 1984). However, of the numerous articles that have investigated systemic
29 immunomodulatory effects due to formaldehyde (Lyapina et al., 2004; Gorski et al., 1992;
30 Thrasher et al., 1990, 1988, 1987; Pross et al., 1987), few have reported significant immune
31 modulation related to formaldehyde exposure (Thrasher et al., 1990, 1988, 1987; Pross et al.,
32 1987). Significant decreases in specific adaptive immune cell populations do not appear
33 correlated to formaldehyde exposure (Erdei et al., 2003; Gorski et al., 1992; Thrasher et al.,
34 1990, 1987; Pross et al., 1987). Thus, the tendency for increased infection rates associated with
35 formaldehyde may not be related to altered immune function. Perhaps altered mucociliary
This document is a draft for review purposes only and does not constitute Agency policy.
4-81 DRAFT—DO NOT CITE OR QUOTE
-------
1 clearance and disturbed mucosal barrier may provide greater access for pathogens and result in
2 greater infection rates. Moreover, formaldehyde has been associated with exacerbation of
3 asthmatic or atopic responses, particularly in sensitized individuals. However, this effect does
4 not appear to occur by increased IgE or formaldehyde-specific IgE levels (Ohmichi et al., 2006;
5 Palczynski et al., 1999; Krakowiak et al., 1998; Wantke et al., 1996b; Salkie, 1991; Grammer et
6 al., 1990; Kramps et al., 1989). Thus, formaldehyde-associated enhanced allergic responses does
7 not appear to be due to direct induction of sensitization and may not occur via an immunologic
8 mechanism. Lastly, the formation of formaldehyde-heme and formaldehyde-HSA complexes
9 has been well documented (Grammer et al., 1993, 1990; Dykewicz et al., 1991; Thrasher et al.,
10 1990) and may serve as a biomarker of exposure (Carraro et al., 1997). Moreover, these
11 complexes may provide a means by which formaldehyde travels throughout the bloodstream and
12 may drive antibody formation that may lead to immune activation.
13
14 4.1.1.6. Neurological/Behavioral
15 There is some suggestion of neurological impairment in humans following occupational
16 exposure to formaldehyde; the data are limited and the results from several studies are potentially
17 confounded by exposure to other solvents. Two studies of histology technicians with
18 occupational exposure to formaldehyde and other solvents found neurological deficits and poorer
19 performance on neurocognitive tests associated with formaldehyde exposure (Kilburn et al.,
20 1987, 1985). In another study, Kilburn and Warshaw (1992) found no change from initial
21 performance, for as long as 4 years, in follow-up evaluations of histology technicians with
22 continuing exposure to formaldehyde. In a preliminary report from a prospective study,
23 Weiskopf et al. (2009) found a strong association between duration of formaldehyde exposure
24 and death from amyotrophic lateral sclerosis (ALS). In a controlled exposure study, Bach et al.
25 (1990) found that, when workers with chronic formaldehyde exposure were challenged with an
26 acute formaldehyde exposure, they exhibited poorer performance on some neurocognitive tests
27 compared with workers without chronic exposure undergoing the same acute challenge
28 conditions. In another controlled exposure study, Lang et al. (2008) found equivocal changes in
29 reaction time following an acute exposure.
30
31 4.1.1.6.1. Epidemiological studies.
32 Kilburn et al. (1985) reported that a group of 76 female histology technicians displayed
33 statistically significantly greater frequencies of neurobehavioral deficits (lack of concentration
34 and loss of memory, disturbed sleep, impaired balance, variations in mood, and irritability), than
35 did a referent group of 56 unexposed female clerical workers. The technicians had been
This document is a draft for review purposes only and does not constitute Agency policy.
4-82 DRAFT—DO NOT CITE OR QUOTE
-------
1 employed from 2 to 37 years (mean 12.8 years). Analysis of workplace air samples indicated the
2 presence of several solvents, ranging from 0.2 to 1.9 ppm for formaldehyde, 3.2 to 102 ppm for
3 xylene, 2 to 19.1 ppm for chloroform, and 8.9 to 12.6 ppm for toluene. Subsequently, Kilburn et
4 al. (1987) administered a battery of 10 tests to 305 female histology technicians to assess various
5 aspects of cognitive and motor function. The researchers analyzed the results by regression
6 analysis with age, years of smoking, and hours per day of exposure to formaldehyde and other
7 solvents as explanatory variables. Increased daily hours of exposure to formaldehyde were
8 significantly correlated with decreased performance in several tests (including several types of
9 memory, dexterity, and balance), whereas hours of daily exposure to other solvents were only
10 correlated with decreased performance in a single memory test. In a later prospective study of
11 performance, 318-494 histology technicians were tested in a battery of neurobehavioral tests,
12 and testing for a subset of subjects was repeated yearly for up to 4 years. No statistically
13 significant decrement in performance was found when initial test results were compared with
14 retest results to evaluate effects of continuing occupational exposure to formaldehyde (or other
15 solvents) or possible effects of aging (Kilburn and Warshaw, 1992). Kilburn (1994) later
16 reported that three anatomists and one railroad worker, occupationally exposed to airborne
17 formaldehyde for 14-30 years, each showed impaired performance on several neurobehavioral
18 tests (e.g., choice reaction time, abnormal balance, digit symbol, and perceptual motor speed).
19 Weisskopf et al. (2009) evaluated the association between chemical exposure and death
20 from ALS, using the cohort of 987,229 people from the prospective Cancer Prevention Study II
21 of the American Cancer Society. From 1989-2004, 1,156 deaths from ALS were identified from
22 mortality records from the National Death Index. Exposure assessment occurred prior to follow-
23 up and was based on a questionnaire; participants were asked about current exposure to 12 types
24 of chemicals and whether they had been regularly exposed in the past. Exposure was evaluated
25 by duration (in years), as information regarding exposure levels was not available. After
26 controlling for a number of potentially confounding factors (including age, sex, smoking status,
27 military service, education, alcohol intake, occupation, vitamin use, and exposure to other
28 chemicals), it was found that exposure to formaldehyde for a known duration was statistically
29 significantly associated with increased risk of death from ALS (p < 0.0001) with a relative risk
30 (RR) of 2.47 (95% CI: 1.58-3.86) based on 22 deaths. Weisskopf et al. (2009) reported that the
31 association had a strongly significant dose-response relationship, with increased duration of
32 exposure associated with increased RR of ALS mortality with a reported/* value for continuous
33 trend of 0.0004. Multivariate adjusted rate ratios were 1.5 for known formaldehyde exposures
34 less than 4 years, 2.1 for 4-10 years, and 4.1 for >10 years. Although the authors indicated that
This document is a draft for review purposes only and does not constitute Agency policy.
4-83 DRAFT—DO NOT CITE OR QUOTE
-------
1 these results need independent verification, the results of this study of the nearly one million
2 people followed for 15 years is unlikely to be biased due to its longitudinal design.
O
4 4.1.1.6.2. Controlled exposure studies.
5 Bach et al. (1990) examined whether cognitive and motor performance of humans
6 responded acutely to formaldehyde exposure and whether previous chronic exposure to
7 formaldehyde affected the responses observed following acute exposure. Thirty-two men with at
8 least 5 years of occupational exposure to formaldehyde and 29 matched controls were exposed to
9 formaldehyde at concentrations of 0.04, 0.21, 0.48, or 1.10 mg/m3 (32, 170, 390, or 890 ppb) for
10 5.5 hours. During the exposure period, symptoms were assessed by using a standardized
11 questionnaire, and subjects were evaluated in four tests designed to estimate several aspects of
12 cognitive function. Testing was performed once prior to exposure and twice during the exposure
13 period. The authors noted that the typical dose-related symptoms of respiratory irritation were
14 not seen in this study. Previously unexposed subjects reported more headaches, "heavy head,"
15 and physical tiredness than the exposed workers (p < 0.025). In both occupationally exposed and
16 unexposed subjects, decreased performance in an addition test was significantly correlated with
17 increasing concentration of formaldehyde (decreased number of additions, p< 0.025; increased
18 number of errors,/* < 0.05). Compared with previously unexposed subjects, occupationally
19 exposed subjects showed significantly decreased performance in three other tests (digit symbol
20 test [pooled exposure groups,/? < 0.025]; digit span [total digit sum],/? < 0.025; graphic
21 continuous line test, p < 0.05), although the effect was not dose related. The study did not adjust
22 for several potential confounders, including prior exposure to other chemical agents, and the age
23 and health status of the cases and controls. Authors concluded that their data indicated that acute
24 exposure to formaldehyde might cause acute effects on CNS functions at exposures of 0.40 and
25 1.2 mg/m3 (equivalent measured doses of 390 or 890 ppb, with a NOAEL of 170 ppb), but that
26 more investigation was needed to verify their results.
27 In a study evaluating chemosensory irritation, Lang et al. (2008) assessed possible
28 changes in reaction time during an acute (4-hour) exposure to formaldehyde concentrations
29 between 0-0.5 ppm (some exposure sessions also included short concentration peaks of up to
30 1 ppm) with or without a masking agent (ethyl acetate). Twenty-one healthy volunteers were
31 exposed once per day to each of 10 different exposure combinations in random order (for a total
32 of 10 sessions per subject). Reaction time was tested before and after each exposure session.
33 Significant increases in reaction time were seen at 0.3 ppm formaldehyde, with or without
34 masking agent, but not at 0.5 ppm. The significance of these findings is unclear.
This document is a draft for review purposes only and does not constitute Agency policy.
4-84 DRAFT—DO NOT CITE OR QUOTE
-------
1 Performance of 16 healthy volunteers on addition, multiplication, and card punching
2 tasks was measured by Andersen and Molhave (1983) before and during a 5-hour exposure to
3 formaldehyde at concentrations up to 2 mg/m3. The authors reported that formaldehyde
4 exposure had no effect on performance, but results were not presented.
5
6 4.1.1.6.3. Summary.
1 The limited information currently available from human studies does not permit a
8 definitive conclusion regarding an association between formaldehyde exposure and human
9 neurotoxicity. There is, however, sufficient information to raise a serious concern for this type
10 of effect, and additional studies are needed.
11
12 4.1.1.7. Developmental and Reproductive Toxicity
13 Epidemiologic studies suggest a convincing relationship between occupational exposure
14 to formaldehyde and adverse reproductive outcomes in women. Several of these studies deal
15 with spontaneous abortion following maternal occupational formaldehyde exposure (Taskinen et
16 al., 1999, 1994; John et al., 1994; Seitz and Baron, 1990; Hemminki et al., 1985, 1982; Axelsson
17 et al., 1984), but not all reported a significant association between exposure and spontaneous
18 abortion. A study of fecundability found an increase in time to pregnancy among female
19 workers exposed to formaldehyde (Taskinen et al., 1999). Three studies that examined the effect
20 of occupational exposures on the incidence of congenital malformation produced mixed results
21 (Dulskiene and Grazuleviciene, 2005; Taskinen et al., 1994; Hemminki et al., 1985). A
22 population-based, semiecologic study found an association between atmospheric formaldehyde
23 exposure and low birth weight (Grazuleviciene et al., 1998).
24
25 4.1.1.7.1. Spontaneous abortion.
26 Several epidemiologic studies report a relationship between occupational exposure to
27 formaldehyde and increases in risk of spontaneous abortion following maternal occupational
28 formaldehyde exposure (Taskinen et al., 1999, 1994; John et al., 1994; Seitz and Baron, 1990;
29 Axelsson et al., 1984). Increased RRs were in the range of 1.7 to more than 3.0. However, other
30 studies (Hemminki et al., 1985, 1982) found no association between occupational formaldehyde
31 exposure and spontaneous abortion. Paternal occupational exposure to formaldehyde was not
32 related to spontaneous abortion (Lindbohm et al., 1991).
33 The earliest report of an association between spontaneous abortion and formaldehyde
34 exposure comes from a Swedish cohort study of female laboratory workers (Axelsson et al.,
35 1984). Subjects were women born in 1935 or later and worked in a university laboratory during
36 1968-1979. There were 745 women who responded to a mailed questionnaire (response
This document is a draft for review purposes only and does not constitute Agency policy.
4-85 DRAFT—DO NOT CITE OR QUOTE
-------
1 rate = 95%), 556 of whom reported on 1,180 pregnancies that resulted in 997 births. Exposure to
2 formaldehyde was estimated based on answers to the questionnaires. Formaldehyde exposure
3 was reported only in connection with 10 pregnancies, of which 5 went to term, 3 were reported
4 as miscarriages, and 2 were terminated by induced abortion. Excluding the latter, the
5 spontaneous abortion rate among women exposed to formaldehyde in the first trimester was 3/8
6 (37.5%) compared with 14/148 (9.5%) in the population of laboratory workers not exposed to
7 any solvent in the first trimester.
8 While not computed by the authors, the OR can be calculated as 5.7 (95% CI: 1.2-26.6).
9 The exposure assessment on which this result is based was methodologically weak but unlikely
10 to be a source of bias. Given the exploratory nature of this study, potential confounders were not
11 controlled for, but no other coexposure was more strongly related to the increased risk of
12 miscarriage, so this result is not likely to be explained by confounding. Selection bias is also an
13 unlikely explanation given the high participation rate. However, although this association of an
14 increased risk of pregnancy loss with formaldehyde exposure is statistically significant, the CI is
15 wide and chance may be a possible explanation for this finding.
16 In a 1988 Health Hazard Evaluation, the National Institute for Occupational Safety and
17 Health (NIOSH) investigated complaints of adverse reproductive outcomes at a plant where
18 work pants were cut and sewn with a fabric that was treated with a resin that releases
19 formaldehyde (Seitz and Baron, 1990). In a NIOSH laboratory, the fabric released 163 to
20 1,430 ug of formaldehyde/gram of cloth. TWA personal breathing space formaldehyde levels
21 ranged from trace to 0.46 ppm, while workstation values ranged from 0.32 to 0.70 ppm. The
22 investigators studied the outcomes by using a mailed questionnaire. The response rate for
23 current employees was 98%. There were 296 pregnancies among a cohort of 188 women. The
24 investigators found increased rates of spontaneous abortion, premature birth, and congenital
25 malformations. The crude rate of spontaneous abortion was 21% among women working at the
26 plant while pregnant (4 of 19 pregnancies), 15% among women employed elsewhere while
27 pregnant (11 of 71 pregnancies), and 5% among women at home while pregnant (10 of 206
28 pregnancies). The investigators did not explain how workers employed elsewhere or at home
29 during pregnancy were categorized compared with current workers, nor did they calculate RRs.
30 As calculated from data presented in Table 5 of the monograph, the crude OR (not corrected for
31 multiple observations per woman) for those pregnant while currently working at the plant
32 compared with all others was 3.2 (95% CI: 0.8-12). There were also excess congenital
33 malformations (13 versus 2%) and premature births (13 versus 4%) among the live births (both
34 based on two births each in the exposed group) from the women who were pregnant while
This document is a draft for review purposes only and does not constitute Agency policy.
4-86 DRAFT—DO NOT CITE OR QUOTE
-------
1 employed at the textile plant compared with women who stayed at home. After the NIOSH
2 investigation, changes were made in the plant to improve ventilation.
3 Because the report provides insufficient details of the methodology and the fact that there
4 was no personal exposure classification in this study, it is difficult to validate the findings in this
5 report. The results did not take into account other potential risk factors for spontaneous abortion
6 or correct for multiple pregnancies per woman. Furthermore, the marked differences between
7 the "home" and "work" pregnancies were difficult to interpret.
8 A cohort study of effects of paternal occupational exposures in Finland by Lindbohm et
9 al. (1991) found that exposure to formaldehyde had little effect on the rate of spontaneous
10 abortions among 99,186 pregnancies listed in the national hospital discharge register. The
11 analysis was limited to births/spontaneous abortions in 1976 and from May 1980 to April 1982.
12 Spontaneous abortion incidence came from the hospital discharge register and data from
13 outpatient clinics. There were 808 pregnant wives among potentially formaldehyde-exposed
14 fathers. Exposure to formaldehyde was based on employment information listed in the Finnish
15 1980 census. Compared with pregnancies among wives of unexposed spouses, the age and
16 socioeconomic level-adjusted ORs were 1.1 for low paternal exposure to formaldehyde and 1.0
17 for moderate to high paternal exposure. Paternal occupational exposures to ethylene oxide,
18 gasoline/benzene, and rubber industry chemicals were associated with spontaneous abortion.
19 The authors hypothesized that the mode of action (MOA) for spontaneous abortion following
20 male exposure to chemicals is genetic damage to germ cells.
21 The indirect exposure assessment was a substantial limitation of this study. Some
22 confounders in a study of this type could not be controlled for (smoking history, previous
23 spontaneous abortions, alcohol use), and census data could not provide completely accurate
24 information, potentially masking associations between paternal formaldehyde exposure and
25 spontaneous abortion.
26 A case-control study by Taskinen et al. (1994) of effects of maternal occupational
27 exposure to chemicals used in laboratories in Finland indicated that exposure to formalin, which
28 is a 37% aqueous solution of formaldehyde, was related to an increased risk of spontaneous
29 abortion. The investigators studied subjects from payrolls of Finnish state-employed laboratory
30 workers, the laboratory workers' union, and a register of workers occupationally exposed to
31 carcinogens. These records were cross-referenced with the hospital discharge register. The
32 investigators selected women who had a single spontaneous abortion during the period
33 1973-1986 and two controls who had delivered a baby without malformations. The final sample
34 size was 208 cases and 329 controls after refusals and other exclusions. The response rate was
35 82.4%.
This document is a draft for review purposes only and does not constitute Agency policy.
4-87 DRAFT—DO NOT CITE OR QUOTE
-------
1 Information on occupational exposure, health status, medication, contraception, and
2 pregnancy history came from mailed questionnaires. Industrial hygienists' construction of an
3 exposure index was based on the subjects' descriptions of their work assignments, use of
4 solvents including estimates of quantity used, and use of a fume hood. ORs were adjusted for
5 employment, smoking, alcohol consumption, parity, previous miscarriage, birth control failure,
6 febrile disease during pregnancy, and other organic solvents found in laboratory work.
7 Spontaneous abortion was associated with 3-5 days per week of formalin exposure (OR 3.5
8 [95% CI: 1.1-11.2]). A contemporaneous study of formaldehyde concentrations in similar
9 Finnish workplaces (pathology and/or histology laboratories) reported workroom air to range
10 from 0.01 to 7 ppm with a mean of 0.45 ppm formaldehyde (Heikkila et al., 1991 [as cited in
11 Taskinen et al., 1994]) and that the highest exposures occurred during emptying of sample
12 containers, dish washing, and preparation of formaldehyde solution.
13 Although the results of this study indicate an increased risk between spontaneous
14 abortion and exposure to formaldehyde/formalin, the women were also exposed to several
15 chemicals concurrently, of which toluene (OR 4.7 [95% CI: 1.4-15.9]) and xylene (OR 3.1
16 [95% CI: 1.3-7.5]) were also significantly associated with the incidence of spontaneous
17 abortion. However, the investigators reported that the women were more likely to be coexposed
18 to formalin and xylene, which would make confounding by toluene less likely, and, since xylene
19 was not as strongly associated with the outcome as was formaldehyde, it too is unlikely to fully
20 explain the reported relationship between formaldehyde and increased risk of spontaneous
21 abortion. While it is possible that exposure misclassification may have occurred because of the
22 indirect assessment of workplace chemical exposure, an overall conclusion is that, since the
23 exposure assessment was conducted by industrial hygienists, it is unlikely that this form of bias
24 will have impacted the results of the study to any great extent.
25 In a U.S. study (John et al., 1994), the results of a case-control study of cosmetologists
26 also supported an association between spontaneous abortion and the use of formaldehyde-based
27 disinfectants. The study population came from the 1988 North Carolina cosmetology license
28 registry. Women on this list who were 22-36 years of age were screened to find those who were
29 recently pregnant. The cases were full-time cosmetologists who experienced a spontaneous
30 abortion before gestational week 20 during 1983-1988. The most recent spontaneous abortion
31 was used as the reference case. Controls were full-time cosmetologists who delivered a live
32 infant during the same time period.
33 Information was based on mailed questionnaires. Women were not told the purpose of
34 the study in order to avoid selection and recall bias. Of 8,356 women who received the
35 screening questionnaires, 72.5% responded. Of those, 1,696 qualified for the study and 73.6%
This document is a draft for review purposes only and does not constitute Agency policy.
4-88 DRAFT—DO NOT CITE OR QUOTE
-------
1 completed a more detailed questionnaire. Among them, 96 women were "absolutely sure" they
2 had a spontaneous abortion and qualified as cases. There were 1,058 live births that qualified as
3 controls. Exposure assessment included identification of disinfectants used as well as types of
4 chemicals used on hair, use of gloves, hours worked, number of procedures involving chemicals,
5 and use of manicure products. Presence of formaldehyde in the cosmetology profession in
6 general was confirmed in two NIOSH hazard reports (Almaguer and Klein, 1991; Almaguer and
7 Blade, 1990). ORs were adjusted for age, smoking, pregnancy characteristics, other jobs, hours
8 worked, education (cosmetology school), hours standing per week, number of chemical
9 procedures per week, hair dyes per week, bleachings per week, permanents per week, use of
10 gloves, beauty salon characteristics, and use of alcohol or formaldehyde disinfectants.
11 Among women who worked full time in cosmetology (61 cases and 315 controls), the
12 crude odds ratio for use of formaldehyde-based disinfectants was 2.0 (95% CI: 1.1-3.8). In
13 models adjusted for maternal characteristics and other workplace exposures, the odds ratio
14 remained elevated (OR = 2.1 [95% CI: 1.0-4.3]) indicating a lack of confounding by other
15 covariates. Other chemical exposures were also associated with spontaneous abortion, including
16 number of chemical services per week, hair dyes, bleaches, and alcohol-based disinfectants.
17 Strengths of this study include adjustment for important confounding risk factors for spontaneous
18 abortion, detailed collection of interview-based information, a favorable response rate, and the
19 fact that the index population had a high likelihood of formaldehyde exposure. These data
20 provide overall support for an association between formaldehyde exposure and spontaneous
21 abortion.
22 In a retrospective cohort study by Taskinen et al. (1999) of female woodworkers in
23 Finland, exposure to formaldehyde was associated with delayed conception and spontaneous
24 abortion. The subjects, recruited from a woodworkers' union and other businesses involving
25 wood processing, were linked to a national register of births. Women were included if they were
26 born between 1946 and 1975, had a live birth at age 20-40 years during 1985-1995, had worked
27 in the wood processing industry for at least 1 month, and had first employment in the wood
28 processing industry beginning at least 6 months before the index pregnancy. The first pregnancy
29 that fulfilled the above criteria was the index pregnancy. There were 1,094 women with these
30 criteria. Information about personal characteristics, pregnancies, and exposures were collected
31 from mailed questionnaires for which the response rate was 64%. After other exclusions
32 (primarily infertility history, unknown time to pregnancy, and contraceptive failure), the final
33 sample included 602 women.
34 Estimates of mean daily exposure to formaldehyde were based on measurements taken at
35 the women's factories of employment during the early 1990s. Where measurements were
This document is a draft for review purposes only and does not constitute Agency policy.
4-89 DRAFT—DO NOT CITE OR QUOTE
-------
1 unavailable, measurements from equivalent industries were used. An exposure index
2 representing a TWA exposure was established for every person in the study based on the
3 concentration of workplace formaldehyde multiplied by the proportion of the workday exposed
4 to formaldehyde. The investigators categorized TWA formaldehyde exposure into categories of
5 low (mean of 18 ppb), medium (mean of 76 ppb), and high (mean of 219 ppb) exposure.
6 Time-to-pregnancy data were analyzed by a discrete proportional hazards regression
7 procedure with, as the outcome, a fecundability density ratio (FDR), in which a ratio of average
8 incidence densities of pregnancies for exposed women was compared with that of the employed,
9 unexposed women. As explained by Taskinen et al. (1999), an FDR significantly below unity
10 suggests that conception was delayed. The age-, employment-, smoking-, alcohol consumption-,
11 parity-, and menstrual irregularity-adjusted FDR was 0.64 (95% CI: 0.43-0.92) for women
12 exposed to high formaldehyde levels compared with the unexposed controls, indicating that there
13 was a substantial delay in time to conception in this group of women. Among a subset of women
14 with high exposure who did not use gloves, the FDR was even lower (0.51 [95% CI:
15 0.28-0.92]), suggesting that these results might be explained in part through dermal contact with
16 formaldehyde or might indicate an individual's failure to follow appropriate precautions, which
17 might have increased inhalation exposures in other ways. Exposure to solvents, wood dust and
18 other dusts, and phenols was not associated with decreased fecundability.
19 The investigators further conducted an analysis of the risk of spontaneous abortion after
20 carefully including only women who had the same workplace during the year of the spontaneous
21 abortion as they had during the beginning of the time-to-pregnancy period. Spontaneous
22 abortion was associated with formaldehyde exposure in the low exposure group (OR = 2.4 [95%
23 CI: 1.2-4.8]), in the medium exposure group (OR = 1.8 [95% CI: 0.8-4.0]), and in the high
24 exposure group (OR = 3.2 [95% CI: 1.2-8.3]). Endometriosis was also associated with the
25 highest formaldehyde level (OR = 4.5 [95% CI: 1.0-20.0]).
26 This study by Taskinen et al. (1999) was a well-designed population-based case-control
27 study that appears to have been well executed and appropriately analyzed. The study population
28 of Finnish women was well defined and adequately selected so as to allow for meaningful
29 comparisons of health effects between individuals with different levels of exposure to
30 formaldehyde. The participation rate was 64%, which is low enough to raise a concern about the
31 potential for selection bias. However, the authors noted that selection bias has not influenced the
32 results of other reproductive epidemiology studies reporting results on smoking, irregular
33 menstruation, and earlier miscarriages, which are known to lengthen the time to pregnancy
34 (Bolumar et al., 1996; Sallmen et al., 1995; Baird and Wilcox, 1985). Furthermore, there is no
35 evidence to support conjecture that an individual's decision to participate in this study would be
This document is a draft for review purposes only and does not constitute Agency policy.
4-90 DRAFT—DO NOT CITE OR QUOTE
-------
1 differential with respect to their workplace formaldehyde exposures while being nondifferential
2 with respect to the other exposures of interest, including organic solvents, wood dust, and
3 phenols. Since the women who chose to participate in this study were not likely to be aware of
4 the specific hypotheses under investigation, nor could they have known the formaldehyde
5 exposures that were independently estimated by an industrial hygienist, selection bias is not a
6 likely explanation for the findings of adversity.
7 Data on pregnancy history, including spontaneous abortions, were collected by
8 questionnaire. Spontaneous abortion is the most common adverse outcome of pregnancy (Klein
9 et al., 1989), and retrospective self-report of spontaneous abortion has been found to match well
10 with prospectively collected reproductive histories (Wilcox and Horney, 1984). Many
11 spontaneous abortions, however, are missed with self-reporting with the magnitude likely
12 exceeding 25%, but only rarely do women self-report false positive events (Wilcox and Horney,
13 1984). The effect of such an undercount is to cause a bias towards the null when the likelihood
14 of undercounting is unrelated to formaldehyde exposure. The implication is that the observed
15 association of increased risk of spontaneous abortion associated with occupational exposure to
16 formaldehyde may be an underestimation of the true risk.
17 Two studies (Hemminki et al., 1985, 1982) specifically assessed the effects of
18 formaldehyde exposure and reported no significant increase in the risk of spontaneous abortion.
19 Hemminki and colleagues (1982) conducted a retrospective cohort study of nurses who were
20 potentially exposed to chemical sterilizing agents, including formaldehyde, ethylene oxide, and
21 glutaraldehyde. The risk of having a spontaneous abortion among the women on the sterilizing
22 staff was compared with that among the control population of nursing auxiliaries whom the
23 supervisory nurses thought to be unexposed to the chemical sterilizing agents during the previous
24 three decades. However, no measurements of the chemical sterilizing agents were taken.
25 Information about exposure to chemical sterilizing agents was obtained from the supervising
26 nurses. When the women were conducting sterilizing procedures during their pregnancies, the
27 frequency of spontaneous abortion was 15.1% compared with 4.6% for the nonexposed
28 pregnancies among the sterilizing staff. The increased frequency of spontaneous abortion
29 correlated with exposure to ethylene oxide but not with exposure to glutaraldehyde or
30 formaldehyde. The investigators reported that ethylene oxide concentrations have been
31 measured in many sterilizing units in Finnish hospitals; 8-hour weighted mean concentrations
32 have ranged from 0.1 to 0.5 ppm with peak concentrations up to 250 ppm (measurements by the
33 Finnish Institute of Occupational Health) (Hemminki et al, 1982). No measurements of
34 glutaraldehyde concentrations were available. Hemminki et al. (1982) reported that exposure to
35 formaldehyde in the sterilization units may be minimal, particularly when gas chambers are used.
This document is a draft for review purposes only and does not constitute Agency policy.
4-91 DRAFT—DO NOT CITE OR QUOTE
-------
1 The range of formaldehyde concentrations measured in sterilizing units has been reported as
2 0.03-3.5 ppm.
3 It is not clear that the unexposed women who served as controls were an appropriate
4 comparison group to the sterilizing staff. The investigators reported that, among the sterilizing
5 staff, those women who were unexposed during pregnancy experienced a rate of spontaneous
6 abortion of 4.6% but that, among the comparison population of nursing auxiliaries who were
7 presumed to be unexposed, the rate of spontaneous abortion was 10.5%. Had the nursing
8 auxiliaries been an appropriate comparison group, it would be expected that their rate of
9 spontaneous abortion would be similar to the unexposed sterilizing staff. Given this anomaly in
10 study design and the unknown concentrations of formaldehyde exposure that were assessed as
11 positive or negative by supervisory nurses regarding occupational exposures in the previous
12 30 years, it is concluded that this report of no association between formaldehyde exposure and
13 the risk of spontaneous abortion does not temper the conclusion that formaldehyde exposure has
14 been shown to increase the risk of spontaneous abortion.
15 A second study by the same lead author (Hemminki et al., 1985) used a different study
16 design to reassess the hypothesis that chemical exposures common in the field of nursing could
17 be risk factors for spontaneous abortion. This case-control study found no increase in the risk of
18 spontaneous abortion associated with exposure to formaldehyde. The head nurses at each
19 hospital were asked by the investigators whether each case or control had been exposed to
20 formaldehyde during a given 3-month period corresponding to the first trimester of a study
21 participant's pregnancy during 1973-1979. Formaldehyde exposure was assessed as positive or
22 negative for either use as a sterilizing agent or use of sterilized instruments. The reported crude
23 OR for formaldehyde exposure was 0.6; no CIs were provided. From the data reported in
24 Table 2 in Hemminki et al. (1985), the unadjusted OR and its CI can be computed post hoc as
25 OR (0.70 [95% CI: 0.28-1.73]). The authors acknowledged that the study failed to distinguish
26 between sterilizing work and the use of sterilized instruments, where only very small exposures
27 could be expected. Given the likelihood of extreme exposure misclassification and the
28 presentation of only crude results without control of potential confounding for formaldehyde,
29 these results do not appear to be exculpatory of a true causal association between formaldehyde
30 exposure and the risk of spontaneous abortion.
31 A meta-analysis of formaldehyde exposure and spontaneous abortion was conducted by
32 Collins et al. (2001b). However, the published results should be interpreted with caution. This
33 meta-analysis included one very large null study of paternal formaldehyde exposure along with
34 seven studies of maternal exposure. The two null studies by Hemminki et al. (1985, 1982) were
35 also included without consideration of the potentially extreme exposure misclassification that
This document is a draft for review purposes only and does not constitute Agency policy.
4-92 DRAFT—DO NOT CITE OR QUOTE
-------
1 may have attenuated any true adverse effect. Nevertheless, the overall reported meta-analytic
2 RR for parental formaldehyde exposure based on eight maternal and paternal exposure studies
3 was 1.6 (95% CI: 0.9-2.7). For case-control studies the RR was 1.8 (95% CI: 0.7-4.8), and for
4 cohort studies the RR was 1.7 (95% CI: 1.2-2.3). Collins et al. (2001b) argued that the method
5 of exposure evaluation may have influenced the observed results; they stated that several of the
6 studies whose exposures were based on the investigator's judgment were likely misclassified,
7 which may have obscured the true relationship, while other studies that assessed exposure based
8 on self-reporting could have suffered from recall bias. They report that RRs were higher for
9 studies based on self-reported exposures (RR =1.9 [95% CI: 1.3-2.6]) than those based on
10 objective exposure assessments (RR =1.5 [95% CI: 0.6-3.7]) and suggested that this difference
11 might reflect recall bias in the exposure assessment. However, for recall bias to have been
12 operable in these studies, the women who provided self-reported data on pregnancy history and
13 occupational exposure would have had to appreciate that the hypothesis of interest was the
14 specific effect of formaldehyde on the risk of spontaneous abortion. In the specific case of the
15 study by Taskinen and colleagues (1999), the investigator also looked at the effects of other
16 exposures, such as organic solvents, dust, and phenols, and did not report adverse effects. It is
17 therefore unlikely that the women providing exposure data were doing so in a manner indicative
18 of recall bias. If the supposition of nondifferential misclassification error in exposure is indeed
19 correct, the observed results of the meta-analysis would likely have been biased towards the null.
20 Therefore, the true RR for maternal formaldehyde could be higher than Collins et al. (2001b)
21 reported and would likely be statistically significant. Had the study of paternal exposure been set
22 aside, the meta-analysis almost surely would have shown a statistically significant increase in the
23 risk of spontaneous abortion associated with maternal formaldehyde exposure. This single study
24 reported a null finding based on exposure assessment from census records of employment, and,
25 as the largest of the studies in the meta-analysis, it contributed the greatest weight.
26 Lastly, Collins and coworkers (200Ib) suggested that there were potential confounding
27 factors in each of the workplaces that might have produced the observed findings of increased
28 risk of spontaneous abortion associated with formaldehyde. While each of these occupational
29 studies focused on women who were coexposed to formaldehyde and other chemicals, the
30 occupational groups were quite different and had different sets of coexposures. The
31 woodworkers in the Taskinen et al. (1999) study were potentially coexposed to organic solvents
32 related to painting and lacquering, dusts, and phenols, none of which was shown to be an
33 independent predictor of adverse risk. The cosmetologists studied by John et al. (1994) were
34 coexposed to hair dyes, bleach, alcohol-based disinfectants, and chemicals specific to services,
35 such as fingernail sculpturing, but, in analyses that were specifically adjusted for other work
This document is a draft for review purposes only and does not constitute Agency policy.
4-93 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposures and their potentially confounding effects, the investigators reported an OR of 2.1
2 (95% CI: 1.0-4.3) for the use of formaldehyde-based disinfectants. The laboratory workers
3 studied by Axelsson et al. (1984) were potentially coexposed to a wide range of solvents, but the
4 miscarriage rate was highest among those exposed to formaldehyde, and, for a potential
5 confounder to entirely explain an observed effect of another exposure, it must be more strongly
6 associated with the adverse outcome.
7 It does not appear that the collective results of formaldehyde exposures associated with
8 increased risk of spontaneous abortion—often in spite of exposures being crudely measured—
9 can be explained by information bias or confounding.
10 The findings by Taskinen et al. (1999) of reduced fertility and increased risk of
11 spontaneous abortion are internally consistent and coherent with other reports of increased risk
12 of pregnancy loss associated with exposure to formaldehyde (John et al., 1994; Taskinen et al.,
13 1994; Seitz and Baron, 1990; Axelsson et al., 1984). Absent evidence of alternative explanation
14 for these findings, it is concluded that exposure to formaldehyde is associated with pregnancy
15 loss and diminished fertility.
16
17 4.1.1.7.2. Congenital malformations.
18 Only three studies have reported on the epidemiologic evidence of an association
19 between formaldehyde exposure and the risks of births having congenital malformations. In the
20 earliest study by Hemminki et al. (1985), the investigators presented an analysis of 34 congenital
21 malformations from the Finnish Register of Congenital Malformations and compared them with
22 a group of 95 controls from those used in the larger study. An association was found between
23 formaldehyde exposure and malformations based on three exposed cases (OR = 1.8).
24 The case-control study by Taskinen et al. (1994) of effects of occupational exposure to
25 chemicals used in laboratories in Finland examined the potential effects of exposure to formalin
26 on both spontaneous abortions and congenital malformation. The investigators reported on a
27 study of 36 laboratory workers with a child registered in the Finnish Register of Congenital
28 Malformations and 105 controls. There was no association between formalin and congenital
29 malformations.
30 A Lithuanian study (Dulskiene and Grazuleviciene, 2005) for which only a brief
31 summary is available in English investigated the risk of congenital heart malformations as a
32 result of exposure to 43 different agents. The number of births included in the study was not
33 given in the English abstract. Exposure to residential ambient formaldehyde concentrations of
34 >2.42 |ig/m3 (0.002 ppm) was associated with a 24% increase in the risk of congenital heart
35 malformations (OR = 1.24 [95% CI: 0.81-2.07]). The details of this study are unavailable in
This document is a draft for review purposes only and does not constitute Agency policy.
4-94 DRAFT—DO NOT CITE OR QUOTE
-------
1 English translation, making it impossible to critically analyze details, such as coexposure and
2 other possible confounders.
O
4 4.1.1.7.3. Low birth weight.
5 A case-control study by Grazuleviciene et al. (1998) examined the association of low
6 birth weight (<2,500 grams) and air pollutants, including formaldehyde, particulates, sulfur
7 dioxide, lead, ozone, and nitrogen dioxide, measured in 12 areas in the city of Kaunas, Lithuania.
8 This city has conducted environmental pollutant measurements since 1993, and the investigators
9 classified formaldehyde exposure based on the area of residence of the study subjects.
10 Formaldehyde levels in the 12 districts of Kaunas in 1994 ranged from 1.36 to 5.28 |ig/m3
11 (0.0011-0.0043 ppm), with a citywide average of 3.14 |ig/m3 (0.0026 ppm). Information on
12 infants came from a birth registry. There were 244 cases of low birth weight and 4,089 normal
13 controls born in 1994. Personal data came from record-based prenatal interviews, and pregnancy
14 data came from hospital records.
15 The crude RR of low birth weight among women exposed to the highest airborne
16 formaldehyde level was 1.68 (95% CI: 1.24-2.27). After adjustment for age, occupation,
17 hazardous work, education, marital status, smoking, hypertension, and other air pollutants, the
18 OR was still elevated but no longer statistically significant (OR 1.37 [95% CI: 0.90-2.09]).
19 Although formaldehyde exposure was the only single air pollutant associated with low birth
20 weight, factors such as smoking, marital status, and pregnancy-related factors had more of an
21 impact on birth weight. Total suspended particulates (OR 2.58 [95% CI: 1.34-4.99]) and
22 hazardous work (OR 2.62 [95% CI: 1.12-6.10]), which was not defined by the authors, were also
23 related to low birth weight.
24 Aside from studies of birth weight deficits from tobacco smoke and occupational
25 exposure, the literature on exposure to ambient air pollutants to support the investigators'
26 hypothesis is limited. The strength of the association between total suspended particulates and
27 low birth weight supports the idea that incidence of birth weight <2,500 grams may be related to
28 atmospheric pollution, although this finding may not be specific to formaldehyde. Because of
29 the large number of variables evaluated in the analysis, large fluctuations in the atmospheric
30 formaldehyde measurements, coexposure to other pollutants, and geographic variability of low
31 birth weight, it is difficult to estimate the impact of formaldehyde alone on low birth weight.
32
33 4.1.1.7.4. Summary.
34 Although all studies on potential developmental toxicity of formaldehyde have
35 limitations and do not uniformly report positive results, the associations between spontaneous
This document is a draft for review purposes only and does not constitute Agency policy.
4-95 DRAFT—DO NOT CITE OR QUOTE
-------
1 abortion, delayed conception, or reproductive outcomes and formaldehyde exposure in multiple
2 studies cannot be dismissed, because several studies report concordant findings across several
3 populations and study methodologies. The results of most of the studies with positive findings
4 were adjusted for many potentially confounding factors that may be related to spontaneous
5 abortion and infertility, including smoking and alcohol use, pregnancy and reproductive history,
6 and other chemical exposures.
7 The association between fertility and formaldehyde (Taskinen et al., 1999) stands out
8 because of its strong quantitative statistical analysis, adequate sample size, and rigorous exposure
9 assessment. This study was designed to specifically assess the effect of formaldehyde on
10 reproductive outcomes. Furthermore, it was the only study with an exposure assessment based
11 on quantitative measurements from the subject's workplace. Moreover, the investigators
12 conducted a multivariable survival analysis that approximates a longitudinal life table or person-
13 year analysis while simultaneously adjusting for important confounders. The findings were
14 strengthened by statistically significant associations between formaldehyde and spontaneous
15 abortion and endometriosis. The fact that the use of gloves may reduce the reproductive effect of
16 formaldehyde supports the dose-response relationship in this study, and the lack of an association
17 between time to pregnancy and any other workplace exposures strengthens the specificity of
18 formaldehyde effects. The results also support associations reported between formaldehyde and
19 increased risk of spontaneous abortion because subfertility and spontaneous abortion are
20 biologically linked (subclinical pregnancy losses are increased among women with fertility
21 problems) (Gray and Wu, 2000; Hakim et al., 1995), and both subfertility and spontaneous
22 abortion may be related to sensitivity to environmental agents (Correa et al., 1996).
23
24 4.1.1.8. Oral Exposure Effects on the Gastrointestinal Tract
25 No human epidemiology studies exist to determine an association between oral exposure
26 of formaldehyde and adverse health effects in the gastrointestinal (GI) tract.
27
28 4.1.1.9. Summary: Noncarcinogenic Hazard in Humans
29 Formaldehyde has clearly and consistently been shown to be a potent sensory irritant
30 with a variety of adverse health effects. Eye, nose, and throat irritation as a result of
31 formaldehyde exposure has been documented in a wide range of epidemiologic studies (Ritchie
32 and Lehnen, 1987; Hanrahan et al., 1984; Liu et al., 1991; Kriebel et al., 1993; Horvath et al.,
33 1988; Holmstrom and Wilhelmsson, 1988; Akbar-Khanzadeh et al., 1994). Workers chronically
34 exposed to formaldehyde have exhibited signs of reduced lung function consistent with BC,
35 inflammation, or chronic obstructive lung disease (Malaka and Kodama, 1990; Herbert et al.,
This document is a draft for review purposes only and does not constitute Agency policy.
4-96 DRAFT—DO NOT CITE OR QUOTE
-------
1 1994; Alexandersson and Hedenstierna, 1989). A well-conducted residential epidemiology study
2 has convincingly shown a concentration response for decreased pulmonary function among
3 children with increased formaldehyde exposures (Krzyzanowski et al., 1990). Several cross-
4 sectional studies have described associations between increased concentrations of formaldehyde
5 and increased prevalence of asthma (Garrett et al., 1999; Krzyzanowski et al., 1990; Norback et
6 al., 1995; Zhao et al., 2008). In addition, a case-control study that focused on risk factors for the
7 initial physician diagnosis of asthma, which is indicative of atopic switching, has shown
8 concentration-dependent adverse effects associated with formaldehyde exposure (Rumchev et
9 al., 2002).
10 Results of research on the effects of formaldehyde on tissue histology suggest that
11 formaldehyde is also responsible for reduced mucociliary clearance and the induction of
12 histopathologic lesions in the nose Pazdrak et al., 1993; Holmstrom et al., 1989; Holmstrom and
13 Wilhelmsson, 1988; Boysen et al., 1990). In addition, there is evidence of neurological
14 impairment in several studies of formaldehyde-exposed histology technicians, but confounding
15 exposures to other neurotoxic solvents and inconsistent results prevent drawing definitive
16 conclusions concerning the neurotoxicity of formaldehyde from these studies (Kilburn et
17 al.,1985, 1987, 1994; Kilburn and Warshaw, 1992).
18 Finally, there is epidemiologic evidence that formaldehyde is associated with adverse
19 reproductive outcomes. Four of six occupational studies found an increased risk of spontaneous
20 abortion among formaldehyde-exposed women (Taskinen et al., 1999, 1994; John et al., 1994;
21 Seitz and Baron, 1990; Axelsson et al., 1984). Results of other studies suggested associations
22 among formaldehyde and congenital malformations, low birth weight, and endometriosis
23 (Hemminki et al., 1985; Grazuleviciene et al., 1998; Taskinen et al, 1999). The strongest
24 evidence of an association between formaldehyde and an adverse reproductive outcome came
25 from a well-conducted study of infertility in women employed in the wood processing industry
26 (Taskinen et al., 1999). This study found a greater than threefold increased risk of spontaneous
27 abortion, a nearly 50% decrease in a measure of delayed conception indicating reduced fertility,
28 and increased time to pregnancy associated with average daily formaldehyde exposures of
29 0.15-lppm.
30
31 4.1.2. Cancer Health Effects
32 The potential for an association between formaldehyde exposure and human cancer has
33 been studied by examining mortality statistics for occupationally exposed individuals as well as
34 in case-control studies of specific cancers. Studies which provide evidence for various
35 respiratory tract cancers, lymphohematopoietic cancers, brain cancer, pancreatic cancer, and
This document is a draft for review purposes only and does not constitute Agency policy.
4-97 DRAFT—DO NOT CITE OR QUOTE
-------
1 other cancers have been published for a range of exposure environments. The following
2 discussion examines the evidence for each cancer type and site. An evaluation is provided
3 regarding the strength of the available epidemiologic evidence for the association between
4 formaldehyde exposure and each cancer.
5
6 4.1.2.1. Respiratory Tract Cancer
1 4.1.2.1.1. Nasopharyngeal Cancer (NPC).
8 Nasopharyngeal cancer is a very rare form of cancer. In the United States, the incidence
9 rate has been estimated at 0.7 cases per 100,000 person-years (Lee and Ko, 2005). In contrast,
10 incidence rates for lung cancer are approximately 100 times higher (60-65 per 100,000 person-
11 years). The most common form of nasopharyngeal cancer arises from the epithelial cells lining
12 the nasopharynx. This presentation constitutes between 75 and 100% of all nasopharyngeal
13 cancers. There are two types, squamous cell carcinoma, and nonkeratinizing carcinoma. In the
14 U.S., the 5-year survival rate for nasopharyngeal cancer is about 25% (Burt et al., 1992). Certain
15 exposures have been implicated in its etiology, including Epstein-Barr virus (EBV), wood dust
16 and particles and substances (including formaldehyde) applied to wood as a preservative or
17 insecticide, exhaust fumes, nickel dust from smelting and refining operations, and nitrosamines.
18 The major epidemiologic studies of formaldehyde exposure in relation to nasopharyngeal cancer
19 are summarized in Table 4-2. This table includes cohort studies that reported data pertaining to
20 nasopharyngeal cancer with risk estimates based on more than two observed cases. Table 4-3
21 summarizes the case-control studies with detailed job history data allowing for characterization
22 of likelihood of formaldehyde exposure
23
24 4.1.2.1.1.1. Cohort studies of nasopharyngeal cancer.
25 Several studies examined exposure to formaldehyde in relation to solid tumor-related
26 mortality risk at 10 production facilities included in the cohort study conducted by the National
27 Cancer Institute (NCI) (Hauptmann et al. 2004; Blair et al., 1987, 1986). This cohort consisted
28 of 25,619 workers from 10 manufacturers of formaldehyde, formaldehyde resins, molding
29 compounds, plastic products, film or plywood who were first employed prior to 1966.
30 Occupational histories from company records and exposure to formaldehyde was estimated for
31 each individual job category from work histories, with calendar-time and plant-specific estimates
32 based on assessments of job titles and tasks associated with those jobs using plant visits by
33 industrial hygienists and monitoring data (Blair et al. 1986; Stewart et al. 1986; Blair and Stewart
34 1990). Exposures were categorized by highest peak exposure, average intensity of exposure,
35 cumulative exposure and duration of exposure. The highest peak exposure categories were
36 defined as nonexposed, low (>0- <2.0 ppm), medium (2.0- <4.0 ppm), and high (> 4.0 ppm).
This document is a draft for review purposes only and does not constitute Agency policy.
4-98 DRAFT—DO NOT CITE OR QUOTE
-------
ts
3 ,
Table 4-2. Major cohort studies of formaldehyde exposure and nasopharyngeal cancer (with 2 or more cases)
Reference
Hauptmann et al. (2004)
[Extension of NCI study
by Blair etal., 1987,
1986)], United States
Study design
Retrospective cohort
mortality study of 25,619
workers employed at 10
formaldehyde plants in the
U.S. followed from either
plant start-up or first
employment through 1994.
The 10 plants produced
formaldehyde (3 plants),
molding compounds
(3 plants), photographic
film (2 plants), plywood
(1 plant), and formaldehyde
resins (6 plants).
Exposure assessment
Time-dependent exposure
estimates3 based on job
titles, tasks, visits to plants
by study industrial
hygienists, and monitoring
data measurements. Peak
exposure = short-term
excursions >8-hour TWA
formaldehyde intensity
and knowledge of job
tasks. Workers
contributed pre-exposure
person time to nonexposed
category. RRs were from
Poisson regression
models, using a 15 -year
lag to account for tumor
latency.
Results (number of observed deaths)
Overall
Nonexposed
Exposed
SMR
SMR
1.56
2.10
Peak exposure (ppm)
0
>0 to <2.0
2.0 to <4.0
4.0 or greater
RRb
1.00
TV/A
MA
1.83
(95% CI: 0.39-23)
(95% CI: 1.05-21)
(referent)
Not provided
Trend p < 0. 001 (Trend on categorical data)
Average intensity of exposure (ppm)
0
<0.5
0.5to<1.0
1.0 or greater
RRb
1.00
MA
0.38
1.67
(referent)
Not provided
Not provided
(2)
(8)
(2)
(0)
vu/
(0)
(7)
(2)
(0)
(1)
(6)
Trend p = 0. 066 (Trend on continuous data among exposed only)
Cumulative exposure (ppm-years)
0
>0to<1.5
1.5to<5.5
5.5 or mo re
RRb
2.40
1.00
1.19
4.14
Not provided
(referent)
Not provided
Not provided
(2)
(3)
(1)
(3)
Trend p = 0. 025 (Trend on continuous data among exposed only)
Duration (years)
0
>0 to <5
5 to <15
15 or mo re
RRb
1.77
1.00
0.83
4.18
Not provided
(referent)
Not provided
Not provided
(2)
(4)
(1)
(2)
Trend p = 0.147 (Trend on continuous data among exposed only)
VO
VO
fe
H
o
o
-------
Table 4-2. Major cohort studies of formaldehyde exposure and nasopharyngeal cancer (with 2 or more cases)
Reference
Marsh et al. (2002),
Connecticut, United
States
Study design
Retrospective cohort
mortality study of 7,328
workers hired up to 1984
and followed until 1998 in
one plant from Hauptmann
et al. (2004). Mortality was
compared with death rates
in two Connecticut counties
and the U.S.
Exposure assessment
Worker-specific exposure3
from job exposure matrix
based on available
sporadic sampling data
from 1965-1987, job
descriptions, and verbal
job descriptions by plant
personnel and industrial
hygienists. Exposures
ranked on a 7-point scale
with exposure range
assigned to each rank.
17% of jobs validated with
company monitoring data;
remaining 83% based on
professional judgment.
Assumed pre- 1965
exposure levels same as
post-1965 levels.
Results (number of observed deaths)
Overall
U.S. referent
County referent
Short-term worker (<1
year)
Long-term worker (<1
year)
Formaldehyde exposure
SMR
SMR
4.94
5.00
5.35
4.59
6.03
(95% CI: 1.99-10)
(95% CI: 2.01-10)
(95% CI: 1.46-14)
(95% CI 0.95-13)
(95% CI: 2.42-12.42)
Duration of formaldehyde exposure (years)
Oto0.2 ppm (years)
Unexposed
Oto
-------
g | ^ Table 4-2. Major cohort studies of formaldehyde exposure and nasopharyngeal cancer (with 2 or more cases)
-* TO' 5^
5 ^ ^
§^ §•
& « °
t"* I
sJ F> s
Sa. Co •-»
<§ § «>
$*z
^ § £
O S^ H5
ll^
Reference
Marsh et al. (2002),
Connecticut, United
States (continued)
Hayes etal. (1990),
Study design
Proportionate mortality
Exposure assessment
Exposure presumed.
Results (number of observed deaths)
Duration of exposure to >0.7 ppm (years)
Unexposed
-------
Table 4-3. Case-control studies of formaldehyde exposure and nasopharyngeal cancer
Reference, study area
Olsenetal. (1984)
Denmark
Vaughanetal. (1986a),
Washington
Vaughanetal. (1986b),
Washington
Study design
Population-based, n = 314
cases from Danish Cancer
Registry during 1970-1982.
Three controls/case sampled
with cancer of the colon,
rectum, breast, and prostate
by age, sex, and year of
diagnosis of cases
Population-based, n = 21
incident cases (1980-1983)
from a 13 -county area
(Washington State Cancer
Surveillance System) and
552 matched controls from
random digit dialing in
same area, for occupational
exposures. Adjusted for
cigarette smoking, alcohol
consumption, gender, and
age.
Same cases and controls as
Vaughanetal. (1986a).
Adjusted for ethnic origin
and cigarette smoking.
Exposure assessment
Employment histories after 1964
from files maintained by Danish
Cancer Registry evaluated by
industrial hygienists.
Interview-based information on
lifetime occupational exposure to
formaldehyde with cases, next of
kin, and controls. Exposure from
available hygiene data, NIOSH
and other data, and NCI job
exposure linkage system.
Exposure score based on sum of
no. years spent per job weighted
by estimated formaldehyde level.
Same as Vaughan et al. (1986a).
Also included residential history in
past 50 years, and use of
particleboard or plywood.
Results
Men
Women
OR
OR
0.7
2.6
(95% CI: 0.3-1.7)
(95% CI: 0.3-22)
Intensity
Low
Medium/high
OR
1.2
1.4
(95% CI: 0.5-3.3)
(95% CI: 0.4-4.7)
No. years exposed
1-9
10 or more
OR
1.2
1.6
(95% CI: 0.5-3.1)
(95% CI: 0.4-5.8)
Exposure score (no lag)
5-19
20 or more
OR
0.9
2.1
(95% CI: 0.2-3. 2)
(95% CI: 0.6-7.8)
Exposure score (15 year lag)
5-19
20 or more
OR
1.7
2.1
(95% CI: 0.5-5.7)
(95% CI: 0.4-10)
Years of residence in mobile home
1-9
10 or more
OR
2.1
5.5
(95% CI: 0.7-6.6)
(95% CI: 1.6-19)
Years of exposure to particleboard
1-9
10 or more
OR
1.4
0.6
(95% CI: 0.5-3.4)
(95% CI: 0.2-2.3)
& s
II
5'
s ,
o
to
O
O
-------
Table 4-3. Case-control studies of formaldehyde exposure and nasopharyngeal cancer (continued)
Reference, study area
Vaughanetal. (1986b),
Washington (continued)
Roushetal. (1987),
Connecticut
West etal. (1993),
Phillipines
Study design
Population-based, w = 173
male cases from the
Connecticut Tumor
Registry who died of any
cause from 1935-1975.
605 male controls randomly
selected from state death
certificates during same
time period, results adjusted
for age at death, year at
death, and availability of
occupational information
(Roushetal., 1987).
Hospital-based, w = 104
non-Chinese incident cases
from the Philippine General
Hospital, matched with 104
hospital and 101
community controls.
Adjusted for years since
first exposure to dust and
exhaust fumes.
Exposure assessment
Four exposure categories based on
probability and duration were
classified by an industrial
hygienist according to job title,
industry, specific employment and
year of employment. Exposure
categories: I, probably exposed
most of working life; II, probably
exposed most of working life and
probably exposed 20+ years before
death; III, probably exposed most
of working life and probably to
high level in some year; IV,
probably exposed most of working
life and probably exposed to high
level 20+ years before death.
Personal interview, including job
history. Industrial hygienists
blinded to case-control status
reviewed and rated jobs as likely
or unlikely to be exposed.
Analysis by length of exposure,
length of exposure lagged 10
years, time since first exposure,
and age at first exposure, based on
date of interview or death.
Results
Exposure source
Occupation only
Mobile home
Both
OR
1.7
2.8
6.7
(95% CI: 0.5-5.7)
(95% CI: 1.0-7.9)
(95% CI: 1.2-39)
Exposure category
I
II
III
IV
OR
1.0
1.3
1.4
2.3
(95% CI: 0.6-1. 7)
(95% CI: 0.7-2.4)
(95% CI: 0.6-3.1)
(95% CI: 0.9-6.0)
Length of exposure (years)
<15
15 or more
RR
2.7
1.2
(95% CI: 1.1-6.6)
(95% CI: 0.5-3.2)
Length of exposure lagged 10 years (years)
<15
15 or more
RR
1.6
2.1
(95% CI: 0.7-3. 8)
(95% CI: 0.7-6.2)
J^.
o
fe
H
O
o
-------
a,
C5
§ 5-
!•»
a ,2
Table 4-3. Case-control studies of formaldehyde exposure and nasopharyngeal cancer (continued)
Reference, study area
Study design
Exposure assessment
Results
West etal. (1993),
Phillipines (continued)
Years since first exposure
<25 RR
25 or more
Age at first exposure (years)
<25
25 or older
RR
1.3
2.9
2.1
1.2
(95% CI: 0.6-3.2)
(95% CI: 1.1-7.6)
(95% CI: 1.1-6.6)
(95% CI: 0.5-3.3)
Armstrong et al. (2000),
Malaysia
Hospital-based, n = 282
Chinese cases, individually
matched to 282 controls by
age and sex recruited
through multistage area
sampling. Adjusted for diet
and smoking variables.
Personal interview, including
detailed job history (description,
time, machines, tools, substances
used, exposures); exposure
classification blinded to case-
control status
Ever exposed
OR
0.71 (95% CI: 0.34-1.43)
o
j^.
fe
H
O
O
Vaughan et al. (2000),
United States
Population-based, n = 196
incident epithelial cases
from 5 U.S. cancer
registries (1987-1993)
matched with 244 controls
from random digit dialing in
the same geographic
regions. Adjusted for age,
sex, race, SEER site,
cigarette usage, proxy
status, and education.
Interviewed for lifetime
occupational and chemical
exposure. Exposure estimates by
industrial hygienist without
knowledge of case-controls status.
Probability of exposure: definitely
not or unlikely (<10%); possible
£10% and <50%); probable
(>50% and <90%); and definite
>90%). Jobs with potential
exposure assigned estimated
concentration levels based on
TWA: low (<10 ppm), moderate
£10 and <50 ppm), and high
£50 ppm).
Possible, probable or definite exposure (61 cases, 76 controls)
Ever
Duration (years)
1-5
6-17
18 or more
OR
OR
1.6
0.9
1.9
2.7
(95% CI: 1.0-2.8)
(95% CI: 0.4-2.1)
(95% CI: 0.9-4.4)
(95% CI: 1.2-6.0)
Trend p = 0.014
Cumulative exposure (ppm-years)
0.05-0.40
0.41-1.10
OR
0.9
3.0
(95% CI: 0.4-2.0)
(95% CI: 0.8-4.1)
(95% CI: 1.3-6.6)
Trend p = 0.033
-------
ts
Table 4-3. Case-control studies of formaldehyde exposure and nasopharyngeal cancer (continued)
Reference, study area
Vaughan et al. (2000),
United States (continued)
Hildesheim et al. (2001)
Study design
Population-based, n = 375
incident cases from two
Taiwanese hospitals
between 7/1 5/91 and
12/31/94. 325 controls
came from a random sample
of households from a
national household
registration system and
were age, sex, and area-of-
residence matched. Tumors
were histologically
confirmed. All subjects
were tested for the EBV;
subset analysis based on
Exposure assessment
In-person interviews collected
information on risk factors and job
history for jobs held >1 year,
including length of time job held,
type of industry, and tasks, tools,
and materials used on the job.
Industrial hygienist assigned
Standard Industry Classification/
Standard Occupational
Classification codes to jobs,
assigning each a probability and
intensity of exposure on a 0-9
scale. Exposure metrics were
duration, average intensity
(intensity scale), average
Results
Probable or definite exposure (27 cases, 30 controls)
Ever
Duration (years)
1-5
6-17
18 or more
OR
OR
2.1
2.0
3.3
1.6
(95% CI: 1.1-4.2)
(95% CI: 0.8-5.0)
(95% CI: 0.9-12)
(95% CI: 0.5-5.6)
Trend p = 0.069
Cumulative exposure (ppm-years)
0.05-0.40
0.41-1.10
>1.10
OR
1.9
2.6
2.2
(95% CI: 0.7-4.9)
(95% CI: 0.7-9.5)
(95% CI: 0.7-7.0)
Definite exposure (10 cases, 2 controls)
Ever
OR
13.3
(95% CI: 2.5-70)
All cases and controls
Ever
OR
1.4
(95% CI: 0.93-2.2)
Duration
>0-< 10
>10
OR
1.3
1.6
(95% CI: 0.69-2.3)
(95% CI: 0.91-2.9)
Cumulative exposure (ppm-yrs)
> 0- <25
>25
OR
1.3
1.5
(95% CI: 0.70-5.8)
(95% CI: 0.88-2.7)
s
"3
o
5'
s ,
J^.
o
fe
H
O
O
-------
it
Table 4-3. Case-control studies of formaldehyde exposure and nasopharyngeal cancer (continued)
Reference, study area
Hildesheim et al. (2001)
(continued)
Study design
EBV positivity (360 cases
and 94 controls).3 Adjusted
for age, sex, education, and
ethnicity.
Exposure assessment
probability (probability scale),
cumulative (average intensity),
years since 1st exposure, and age at
1st exposure. Analysis of
nonkeratinizing or undifferentiated
tumors yielded similar results as
overall analysis
Results
El
E^
Di
BV positive cases a
/er
.ration
>0-< 10
nd contr
OR
OR
ols
2.7
2.8
2.6
(95% CI: 0.1.2-6.2)
(95% CI: 0.83 -9.7)
(95% CI: 0.87-7.7)
Cumulative exposure (ppm-yrs)
> 0- <25
>25
OR
4.0
2.2
(95% CI: 0.92-17)
(95% CI: 0.80-5.8)
& s
II
5'
s ,
J^.
o
aEBV seropositives defined as positive for one of the following anti-EBV antibodies known to be associated with nasopharyngeal cancer: viral capsid antigen IgA,
EBV nuclear antigen 1 IgA, early antigen IgA, DNA binding protein IgG, and anti-DNase IgG.
O
O
-------
1 Average intensity categories of exposure were defined as nonexposed, low (>0- <0.5
2 ppm), medium (0.5- <1.0 ppm), and high (>1.0 ppm). Cumulative exposure was defined as
3 nonexposed, low (>0- <1.5 ppm-years), medium (1.5- <5.5 ppm-years), and high (>5.5 ppm-
4 years). Duration of exposure was defined as nonexposed, low (>0 years- <5 years), medium
5 (5- <15 years), and high (>15 years). The presence of formaldehyde-containing particulates and
6 other potential chemical coexposures in the plants were indentified as suspected carcinogens
7 including antioxidants, asbestos, carbon black, dyes and pigments, hexamethylenetetramine,
8 melamine, phenol, plasticizers, urea, wood dust and benzene. Standardized mortality ratios
9 (SMRs) were calculated using sex-, race, age-, and calendar-year-specific U.S. mortality rates.
10 Relative risks based on internal comparisons within the cohort were estimated using Poisson
11 regression, controlling for calendar year, age (5-year categories), sex, race, and pay category.
12 Subjects were initially followed to January 1, 1980, accruing approximately 600,000
13 person-years of follow-up (Blair et al., 1987). Information on potential formaldehyde exposures
14 after 1980 was unavailable. Hauptmann et al. (2004) updated the cohort through December 31,
15 1994 reporting a total accrual of person-time of 865,708 person-years with a median duration of
16 follow-up of 35 years. Cause of deaths was taken from death certificates. A total of 10 deaths
17 from nasopharyngeal cancer were identified. Among them 8 were classified as ever exposed and
18 2 never exposed.
19 Hauptmann et al. (2004) reported an increased risk of nasopharyngeal cancer in exposed
20 workers based on U.S. population death rates (standardized mortality ratio [SMR] = 2.1 [95%
21 CI: 1.05-4.21]). This association was based on a total of 8 nasopharyngeal exposed cancer
22 cases; however, 1 of these cases was reclassified to oropharyngeal cancer based on secondary
23 sources of data and thus 7 nasopharyngeal cases were included in additional analyses using
24 Poisson regression modeling of relative risks (RRs) with an internal referent group.
25 In addition to the SMR analysis, Hauptman et al. (2004) provided both RRs based on
26 internal referent groups (unexposed employees, and the low exposed workers) and regression
27 analysis of all exposed workers. Use of an internal referent group reduces potential selection
28 bias due to the healthy worker effect and potential confounding from area- and employment-
29 related factors. Both of these analyses were conducted for four different exposure metrics
30 Exposure measures used in the Poisson regression were based on a standardized exposure
31 assessment protocol (summarized in Table 4-2), and included measures of highest peak
32 exposure, average exposure intensity, cumulative exposure, and duration of exposure to
33 formaldehyde (Hauptmann et al., 2004). RRs for nasopharyngeal cancer increased with various
34 exposure metrics with trend/?-values (based on continuous exposure measures) < 0.001 and
35 0.025 for the peak exposure and cumulative exposure metrics, respectively (see Table 4-2).
This document is a draft for review purposes only and does not constitute Agency policy.
4-107 DRAFT—DO NOT CITE OR QUOTE
-------
1 These analyses also adjusted for potential confounders, including calendar year, age, sex, race,
2 and pay category. The investigator evaluated exposures to 11 suspected carcinogens and other
3 widely used chemicals in the plants including antioxidants, asbestos, carbon black, dyes and
4 pigments, hexamethylenetetramine, melamine, phenol, plasticizers, urea, wood dust, and
5 benzene. Hauptmann et al. (2004) reported that the relative risks for most cancers did not
6 meaningfully change when adjusted for these exposures except for nasopharyngeal cancer and
7 melamine exposure. While the RR for the highest exposure categories of peak and average
8 intensity of formaldehyde exposure declined with adjusted for melamine exposure, Hauptmann
9 et al. (2004) reported that the relative risks for cumulative exposure and duration of exposure
10 were still elevated. The tests for trend in exposure-response remained highly statistically
11 significant for peak (p < 0.001), average (p = 0.021), and cumulative (p = 0.006) exposure.
12 There was no evidence of any differential measurement error that could have produced a
13 spurious association; nondifferential measurement error would likely have led to an observed
14 effect of formaldehyde that was less than that which would otherwise have been observed in the
15 ab sence of measurement error.
16 Overall this study provides strong causal evidence of an association between
17 formaldehyde exposure and the risk of nasopharyngeal cancer mortality. Although this is a rare
18 cancer the cohort was large with sufficient follow-up to accrue 7 exposed cases of
19 nasopharyngeal cancer compared to 2 unexposed cases making this cohort study the most
20 statistically powerful of the cohort studies of nasopharyngeal cancer. The objective exposure
21 assessment was of high quality which also increased the power of this study to detect an effect—
22 even with a small number of cases. Potential sources of confounding or bias were also well
23 controlled. Finally, evidence of a stronger association, with the high exposure groups for each of
24 the 4 different exposure metrics, indicates the consistency of exposure-response relationship.
25 Following these reports of increased risk of nasopharyngeal cancer associated with
26 formaldehyde exposure, a series of post hoc analyses of a subset of the NCI cohort were
27 undertaken by Marsh and coworkers (Marsh et al., 2007a, b, 2002, 1996; Marsh and Youk,
28 2005). Briefly, these studies focused on the specific findings from a single plant in the NCI
29 cohort (Wallingford, Connecticut) that generated the majority of the nasopharyngeal cancer
30 cases. The Marsh et al. (2002) study had both a cohort analysis and a nested case-control
31 analysis. The cohort study followed 7328 workers employed in the plant from 1941 to 1984,
32 extending the follow-up period to 1998 and independently evaluated the exposure assessment
33 (Marsh et al., 1996 present the earlier analysis with follow-up through 1984). Seven deaths from
34 nasopharyngeal cancer were seen in this cohort, for SMRs of 4.94 (95% CI 1.99-10.19) using the
35 United States population as the referent group and 5.00 (95% CI 2.01-10.30) using the local
This document is a draft for review purposes only and does not constitute Agency policy.
4-108 DRAFT—DO NOT CITE OR QUOTE
-------
1 county referent group. Nearly identical results using the national and local populations as the
2 referent group shows that those background rates were very similar. Strong associations were
3 seen with measures of ever exposure to formaldehyde (SMR = 6.03; 95% CI: 2.42-12.42), as
4 well as duration of exposure (Greater than 10 years, SMR = 12.46; 95% CI: 1.51-45.02) and
5 cumulative exposure (Greater than 0.22 ppm, SMR =7.51; 95% CI: 1.55-21.93) (see Table 4-2
6 for additional detail). This study used a different exposure assessment and even computed an
7 SMR based on local rates. Additionally the exposure response seen with both cum and duration
8 of exposure adds strength to the observed associations. The nested case-control study
9 ascertained data on smoking and date of hire, however, these data were not able to be modeled
10 due to the limited number of nasopharyngeal cases (n = 7). This analysis supports the findings
11 of the NCI study, although only a single plant was included in the study (Marsh et al., 2002).
12 However, Marsh et al. (2002) interpret these data as providing evidence against the involvement
13 of formaldehyde in nasopharyngeal cancer because elevated risks were seen in short-term (as
14 well as in long-term workers) and in the lower exposure categories (as well as in the higher
15 exposure categories) of the various formaldehyde measures.
16 Marsh and Youk (2005) further suggest that this interpretation of the Marsh et al. (2002)
17 data weakens the support seen in the larger NCI study for a causal link between formaldehyde
18 and nasopharyngeal cancer because the Hauptman et al. (2004) results are primarily driven by
19 the results from the Wallingford, Connecticut plant. Marsh and Youk (2005) re-analyzed of
20 nasopharyngeal cancer data from the Hauptmann et al. (2004) study. They compared the results
21 from the Wallingford, Connecticut plant (Plant 1), which contributed five of the nine
22 nasopharyngeal cancer deaths in the NCI study to the other nine plants. Marsh and Youk (2005)
23 reported that when the SMR for nasopharyngeal cancers in Plants 2 to 10 combined in was not
24 elevated (SMR = 0.65, 95% CI = 0.8 to 2.3, 4 deaths), in comparison with that of Plant 1 alone
25 (SMR = 10.3, 95% CI = 3.8 to 22.5, 6 deaths). However, from Marsh and Youk (2005), the two
26 plants with the highest exposures were Plants 1 and 2. The SMR for Plant 2 was elevated at 5.35
27 based on a single case (95% CI: 0.13-29.83). While this effect estimate is extremely unstable
28 statistically, it is may be a reflection of the relatively higher formaldehyde exposures than at
29 other plants.
30 In the most recent reports, Marsh et al. (2007a) provides additional data from a nested
31 case-control study with the 7 nasopharyngeal cancer cases and 16 other pharyngeal cancer cases
32 identified from the Wallingford, Connecticut cohort. Cases were matched to 4 controls in the
33 cohort based on age, race, sex, and year of birth. Smoking history and other work history data
34 (jobs held other than at the Wallingford plant) were obtained from interviews with 5 of the 7
35 cases and with 68 of 88 controls. The source of the interview differed between cases and
This document is a draft for review purposes only and does not constitute Agency policy.
4-109 DRAFT—DO NOT CITE OR QUOTE
-------
1 controls, with family member interviews used for most of the cases, and self-interviews used for
2 most of the controls. Details of the content and structure of the interview were not provided, but
3 the authors say that little useful information was obtained from these interviews (Marsh et al.
4 2007a, 2002). Additional sources of work history included data from the plant job application,
5 city directories and genealogy-based search strategies. An association was seen between
6 nasopharyngeal cancer risk and work in silversmithing occupations (including brass plating or
7 other jobs relating to silver or brass, OR 14.4, 95% CI 1.3, 758; 4 exposed cases); the odds ratio
8 for other types of metal work (steel work and welding) was 3.61 (95% CI 0.50-22.7; 3 exposed
9 cases) and for the combined category of silversmithing and other metal work the OR was 7.31
10 (9% CI 1.08-82.1; 5 exposed cases). There are no prior citations of an association between
11 silversmithing exposures and nasopharyngeal cancer in the medical literature, but Marsh et al.
12 review the literature pertaining to related exposures (sulfuric acid mists, metal dusts) and
13 respiratory and laryngeal cancer to support this association. However, the results for these
14 exposures and laryngeal cancer are inconsistent, and data pertaining to these exposures and
15 nasopharyngeal cancer are quite limited. Despite these limitations, Marsh et al. (2007) suggest
16 that the observed associations between nasopharyngeal cancer and formaldehyde exposure in the
17 Wallingford plant are due to these other occupational exposures. Marsh et al. (2007) do note that
18 history of silversmithing and other metal work was not associated with formaldehyde exposure,
19 and so was not a confounder of the formaldehyde results as reported for the Wallingford Plant.
20 In addition, there was no evidence of confounding due to smoking history. Regardless, Marsh et
21 al. (2007a) suggest that these other occupational exposures, rather than formaldehyde, could
22 explain the risk of nasopharyngeal cancer seen in the Wallingford cohort.
23 Two cohort studies of professional groups, such as anatomists, pathologists, embalmers,
24 and funeral directors, examined the risk of nasopharyngeal cancer and formaldehyde exposure
25 (see Table 4-2). In general, measurements of formaldehyde concentrations were not available in
26 studies of these groups but are generally below 1 ppm (IARC, 1995; Korczynski, 1994; Stewart
27 et al., 1992; Moore and Ogrodnik, 1986). Hayes et al. (1990) conducted a study of 3,649
28 deceased white and 397 deceased nonwhite U.S. male embalmers and funeral directors who had
29 died between 1975 and 1985, using records from local licensing boards, state funeral directors'
30 associations in 32 states and the District of Columbia, the National Funeral Directors'
31 Association, and state offices of vital statistics (n = 894). Expected deaths by cause were derived
32 from 5-year age- and calendar-year-specific proportions of deaths among appropriate race groups
33 from the U.S. population. No measured exposure data were available. Hayes et al. (1990)
34 reported an excess risk of nasopharyngeal cancer among male professional embalmers and
35 funeral directors, based on 4 deaths with 1.85 expected (PMR 2.16, 95% CI: 0.59-5.54). The
This document is a draft for review purposes only and does not constitute Agency policy.
4-110 DRAFT—DO NOT CITE OR QUOTE
-------
1 authors note that although these results are based on small numbers of deaths, the specific
2 excesses of nasopharyngeal cancer that they observed were consistent with those reported by
3 Vaughan et al. (1986a,b; and Blair et al., 1987) which showed increased risks of nasopharyngeal
4 cancer associated with exposure to formaldehyde.
5 Hansen and Olsen (1995) studies workers in 265 Danish industries that produced more
6 than one kg of formaldehyde per employee per year, where 2,041 of 91,182 cancer patients had
7 at least 10 years of continuous formaldehyde-related work experience before diagnosis. The risk
8 of cancer incidence was estimated from standardized proportionate incidence ratios as
9 denominators were not available. The standardized proportionate incidence ratio (SPIR) is the
10 proportion of cases of a particular cancer among exposed persons compared to the proportion of
11 cases of a particular cancer among the general population. Hansen and Olsen (1995) reported an
12 SPIR, based on 4 observed and 3.2 expected cases, of 1.3 (95% CI 0.03, 3.2).
13 Other studies also examined nasopharyngeal cancer risk in occupational cohorts, but are
14 limited by the small number of expected cases. Pinkerton et al. (2004) updated of a cohort study
15 of 11,030 workers (82% female) followed from 1955 or the beginning date of exposure through
16 1982 in three garment factories conducted by Stayner et al. (1988). Formaldehyde resins were
17 used to treat fabrics in these factories beginning in 1955 and 1959. Although formaldehyde
18 levels were available on a subset of the employees from monitoring data available from surveys
19 completed in 1981 and 1984, they were not used in this analysis. The geometric mean 8-hour
20 time weighted average exposure across all departments and plants was 0.09-0.20 ppm. The
21 geometric mean formaldehyde concentration across all departments was estimated at 0.15 ppm.
22 No cases of nasopharyngeal cancer were observed compared to an expected number of one case.
23 Gardner et al. (1993) studied a cohort of 14,017 workers exposed to formaldehyde in the
24 British chemical industry and followed until the end of 1989. Exposure to formaldehyde was
25 classified according to job title and assigned to categories including nil/background (<0.1 ppm),
26 low (0.1-0.5 ppm), moderate (0.6-2.0 ppm) and high (over 2.0 ppm). Results were stratified by
27 date of first employment and among men first employed before 1965, the SMR for cancer of the
28 pharynx was 1.47 (95% CI: 0.59-3.03). No cases were reported among the men first employed
29 after 1965. Although the cause-specific cancer mortality was for cancer of pharynx (ICD-9
30 146-9) and would include nasopharyngeal cancer, the authors stated separately that there were
31 1.3 cases of nasopharyngeal cancer that were expected but no deaths recorded for
32 nasopharyngeal cancer. Coggon et al. (2003) updated this cohort study with additional follow-up
33 through the end of 2000 and reported the effects of formaldehyde exposure on mortality from
34 cancer among 14,014 workers. Coggon et al. (2003) reported only one case compared to an
This document is a draft for review purposes only and does not constitute Agency policy.
4-111 DRAFT—DO NOT CITE OR QUOTE
-------
1 expected number of two cases (SMR 0.5, 95% CI 0.07-3.55 as calculated in a meta-analysis by
2 Bossetti et al., 2007).
3 Other cohort studies examined occupational formaldehyde exposure in relation to total
4 cancer or lung cancer risk, but did not provide specific data for nasopharyngeal cancer risk
5 (Andjelkovich et al., 1995; Bertazzi et al. 1989, 1986; Edling et al., 1987;). These studies are not
6 discussed further in this section. A summary of studies of oral cavity cancers (including
7 oropharynx) and laryngeal cancers is presented in Section 4.1.2.1.4.
8
9 4.1.2.1.1.2. Case-control studies of nasopharyngeal cancer.
10 Eight case-control studies (Hildesheim et al., 2001; Vaughan et al., 2000; Armstrong et
11 al., 2000; West et al., 1993; Roush et al., 1987; Vaughan et al., 1986a, b; Olsen et al., 1984)
12 provided evidence of excess risks of nasopharyngeal cancer due to formaldehyde exposure (see
13 Table 4-3). Olsen et al. (1994) conducted a population-based case-control study of
14 nasopharyngeal cancer in Denmark. Cases of nasopharyngeal cancer (n = 266) were ascertained
15 from the Danish Cancer Registry during 1970-1982. Controls were diagnosed with cancer of the
16 colon, rectum, breast, and prostate and three controls were matched to each case by age, sex, and
17 year of diagnosis. Employment histories after 1964 were evaluated by industrial hygienists who
18 classified subjects as exposed or unexposed and wee blinded to case-control status. Among men
19 the odds ration for occupational exposure to formaldehyde was OR = 0.7 (95% CI: 0.3-1.7);
20 among women the OR = 2.6 (95% CI: 0.3-21.9).
21 Vaughan et al. (1986a) conducted a population-based case-control study of the pharynx,
22 sinus and nasal cavity. Incident cases of nasopharyngeal cancer (n = 27) were matched to
23 controls selected by random digit dialing in the same areas. Medical, occupational, and
24 residential histories as well as information of smoking and alcohol consumption were obtained
25 by telephone interview for cases and controls. When cases were deceased, next-of-kin were
26 interviewed. Logistic regression was used to control for potential confounders. Exposure was
27 evaluated from available hygiene data, NIOSH and other data, and NCI job exposure linkage
28 system. Exposure scores were based on sum of number of years spent per job weighted by
29 estimated formaldehyde level. Occupational formaldehyde exposures were evaluated by
30 maximum exposure level, number of years exposed, an exposure score across all years and an
31 exposure score that excluded the previous 15 years to account for the induction periods between
32 possible etiologically relevant exposures and the detection of incident cancers. Exposure scores
33 were based on a weighted sum of the years in a job with weights depending upon the estimated
34 exposure level. In all but one of eight comparisons, the odds ratios were elevated; however,
35 none to the odds ratios was statistically significant. Effect estimates were most elevated for
This document is a draft for review purposes only and does not constitute Agency policy.
4-112 DRAFT—DO NOT CITE OR QUOTE
-------
1 nasopharyngeal cancer among those with the highest exposures score. Ignoring the potential
2 induction period, the OR = 2.1 (95% CI; 0.6-7.8). With the 15-year induction period the
3 OR = 2.1 (95% CI: 0.4-10).
4 Vaughan et al. (1986b) evaluated the residential exposures to formaldehyde among the
5 same nasopharyngeal cases as reported in Vaughan et al. (1986a). Logistic regression analyses
6 controlled for smoking and ethnic origin (White, Black, Asian, other). Potential residential
7 exposure to formaldehyde was estimated by utilizing residence in a mobile home with or without
8 the presence of urea-formaldehyde foam insulation (UFFI) or particleboard or plywood as a
9 surrogate for exposure. The authors found statistically significant OR of 5.5 (95% CI: 1.6-19.4)
10 for subjects reporting residence of 10 or more years in a mobile home with UFFI before
11 diagnosis. An elevated odds ratio was reported for less than 10 years of mobile home residence
12 (OR = 2.1; 95% CI: 0.7-6.6). The authors also evaluated whether the risk associated with living
13 in a mobile home was affected by also working in a mobile home. Occupational exposure to
14 mobile homes without residential exposure had an OR =1.7 (95% CI: 0.5-5.7). Residential
15 exposure to mobile homes without occupational exposure had an OR = 2.8 (95% CI: 1.0-7.9).
16 Both having residential and occupational exposure to mobile homes was statistically
17 significantly associated with incidence of nasopharyngeal cancer (OR = 6.7, 95% CI: 1.2-38.9).
18 Roush et al. identified 173 cases of nasopharyngeal cancer in men from the Connecticut
19 Tumor Registry who dies of any cause between 1935 and 1975. Male controls (n = 605) were
20 selected from state death certificates during the same time period. Four exposure categories
21 were created based on the probability and duration of formaldehyde exposure based on industrial
22 hygienist evaluations of subject's job titles, industry, specific employment and duration of
23 employment. Logistic regression adjusted for potential confounders including age at death, year
24 of death. For the lowest exposure category defined as 'probably exposed most of working life,'
25 the OR =1.0 (95% CI: 0.6-1.7); for the second exposure category defined as 'probably exposed
26 most of working life and probably exposed 20+ years before death,' the OR =1.3 (95% CI: 0.7-
27 2.4); for the third category of exposure defined as 'probably exposed most of working life and
28 probably to high level in some year,' the OR =1.4 (95% CI: 0.6-3.1); and for the fourth exposure
29 category defined as 'probably exposed most of working life and probably exposed to high level
30 20+ years before death,' the OR = 2.3 (95% CI: 0.9-6.0).
31 West et al. (1993) investigated risk factors for nasopharyngeal cancer in the Philippines
32 using a matched case-control study of 104 cases and 205 hospital and community controls.
33 Personal interviews were conducted by a trained nurse who ascertained information on
34 demographic factors including ethnicity and education, dietary consumption, occupational
35 history, smoking, use of herbal medicine, betel nut, and anti-mosquito coils. An industrial
This document is a draft for review purposes only and does not constitute Agency policy.
4-113 DRAFT—DO NOT CITE OR QUOTE
-------
1 hygienist who was blinded to case-control status evaluated subject's occupational histories to
2 estimate likely exposures to formaldehyde, solvents, dusts, exhausts, and pesticides. Conditional
3 logistic regression was used to control for coexposures. Controlling for education, dust,
4 processed meat and fresh fish consumption, mosquito coil usage, and use of herbal medicine, the
5 RR of occupational formaldehyde exposure more than 25 years earlier was 4.0 (95% CI:L 1.3-
6 12.3). The independent effect of dust was also statistically significant for exposure more than 35
7 years prior (OR = 4.4; 95% CI: 1.1-17.5) as was the independent effect of daily use of mosquito
8 coils (OR = 5.9; 95% CI: 1.7-20.1). While the authors mentioned only that the compounds in the
9 anti-mosquito coil smoke might be of further interest, they did identify those compounds.
10 However, independent testing of 6 brands of East Asian mosquito coils evaluated the emission
11 rates of carbonyl compounds in the mosquito smoke and reported that formaldehyde and
12 acetaldehyde had the highest emission rates (Liu et al., 2003). Among the three experiments on
13 each of the six brands, the range of formaldehyde concentrations was from 0.87 |ig/m3 (1 ppb) to
14 25|ig/m3(31ppb).
15 Armstrong et al. (2000) studied 282 ethnically Chinese residents of Malaysia with
16 recently diagnosed nasopharyngeal cancer (diagnosis from 1987-1992, with case identification
17 occurring between 1990 and 1992). Cases were individually matched to 282 controls by age and
18 sex; structured interviews collected residential and work history data. No association was seen
19 between formaldehyde exposure and nasopharyngeal cancer (adjusted OR = 0.71 [95% CI:
20 0.34-1.43]), controlling for wood dust and industrial heat.
21 In a population-based case-control study of incident nasopharyngeal cancer cases
22 identified through the U.S. SEER cancer registry, Vaughn et al. (2000) used work history data
23 collected from interviews (see Table 4-3) to estimate each individual worker's formaldehyde
24 exposure. Workers with more than 1.10 ppm-years of cumulative exposure were found to be at
25 significantly higher risk of nasopharyngeal cancer, with an odds ratio (OR) of 3.0 (95% CI
26 1.3-6.6). Both duration of exposure and cumulative exposure were positively associated with
27 increased risk of nasopharyngeal cancer (trend/? = 0.014 and 0.033, respectively). The OR
28 increased in magnitude as the probability of "Ever" having occupational exposure increased,
29 from OR =1.6 (95% CI 1.0-2.8) among the 61 cases whose exposure was judged to be
30 "Possible, probable or definite", to 2.1 (5% CI 1.1, 4.2) among the 27 cases with "probable or
31 definite" exposure, to OR = 13.3 (95% CI 2.5-70) among the 10 cases with "definite" exposure
32 (p-trend < 0.001) (Vaughn et al., 2000).
33 The study design of the Hildesheim et al. (2001) study was a matched case-control study
34 with controls individually matched to nasopharyngeal cancer cases on age, sex, and
35 district/township of residence. Hildesheim et al. (2001) reported that exposure to formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-114 DRAFT—DO NOT CITE OR QUOTE
-------
1 produced modest risk elevations: for duration of exposure, OR = 1.6 for 10 years or less and 1.2
2 for over 10 years of exposure; for cumulative exposure, OR =1.3 for <25 years of exposure and
3 1.5 for 25+ years of exposure. Among those with EBV, the OR was 2.7 (95% CI: 1.2-6.2) for
4 ever-exposed persons. The risk was higher among exposed persons whose work history was
5 within the last 10 years (OR = 4.7 [95% CI: 1.1-20.0]) and for those followed 20+ years after
6 exposure (OR = 2.8 [95% CI: 1.1-7.6]). The analysis of these data, however, was conducted
7 using unconditional logistic regression. The authors stated that conditional logistic regression
8 was not used "to avoid the loss of information from cases and controls without a matched pair."
9 This loss of information was particularly acute among the EBV+ strata in which 360/375 cases
10 were EBV+ while only 94/325 controls were EBV+. While the investigators did include age, sex
11 and other potential confounders in the analytic models, this analytic methodology may not be
12 sufficient to effectively control for these factors given the individual matching used in the study.
13 Failure to condition on the matching factors can create bias that is unpredictable in direction or
14 magnitude (Cox and Hinkley, 1974, p. 292 and Breslow and Day, 1980, p. 249). Had the study
15 design specified frequency matching rather than individual matching, then the logistic regression
16 analysis with the inclusion of the matching factors would have been appropriate; however, this
17 apparent weakness in all the analyses by Hildesheim et al. (2001) raises considerable uncertainty
18 regarding the results and less weight should be assigned to this study in the overall evaluation.
19
20 4.1.2.1.1.3. Summary of nasopharynseal cancers.
21 Because of the low incidence rate of nasopharyngeal cancer, cohort studies that rely on
22 nonspecific exposure assessment techniques may not be expected to provide sufficiently precise
23 data needed for evaluation of the contribution of formaldehyde to the etiology of this particular
24 cancer type due to a lack of statistical power. Even the latest cohort study of the British
25 industrial worker which evaluated 14,014 men (Coggon et al., 2003) only expected to observe
26 two cases while the latest cohort study of the garment workers which evaluated 11,030 men and
27 women (Pinkerton et al., 2004) only expected to observe one case. Cases counts depend upon
28 both the number of people followed and the length of follow-up time as well as the baseline
29 incidence rate of nasopharyngeal cancer. However, it is important to understand that the
30 statistical power of these cohort studies largely depends upon the count of the number of
31 observed and expected cases and not the number of people who were followed. The variance of
32 an observed relative risk is a function of the inverse of the observed and expected case counts.
33 Small case counts produce large statistical variances. The large variances in effect estimates
34 results in wide confidence intervals that may be unstable and offer little insight into any
35 exposure-response relationship other than to statistically rule out very strong adverse effects.
This document is a draft for review purposes only and does not constitute Agency policy.
4-115 DRAFT—DO NOT CITE OR QUOTE
-------
1 Thus EPA's evaluation of the epidemiologic data on formaldehyde and nasopharyngeal
2 cancer focuses on the large NCI cohort study which evaluated the mortality among 25, 619
3 workers (Hauptmann et al., 2004) and the case-control studies with extensive work history data
4 (e.g., lifetime job histories) and industrial hygiene assessment of exposure potential.
5 Hauptmann et al. (2004) reported that for mortality from all types of cancer combined,
6 the workers experienced fewer deaths than would be expected in the general population
7 (SMR = 0.90, 95% CI: 0.86-0.95). This typically reflects that fact that workers are healthier than
8 the general population which includes nonworkers. In spite of this 'healthy worker effect,' the
9 Hauptmann et al. (2004) analysis reported a doubling of nasopharyngeal cancer mortality risk in
10 workers from 10 formaldehyde producing or using factories (SMR = 2.1, 95% CI: 1.05-4.21).
11 Additional analyses demonstrated that relative risks increased with average exposure intensity
12 (p-trend = 0.066), cumulative exposure (p-trend = 0.025), highest peak exposure (p-trend <
13 0.001) and duration of exposure to formaldehyde (p-trend = 0.147).
14 In addition to the evidence from the NCI cohort studies, modest additional evidence is
15 found in the professional cohort studies of Hayes et al. (1990) and Hansen and Olsen (1995)
16 however, the elevated risks overlapped with unity and the individual study results were not
17 statistically significant. The rarity of the disease and difficulties in obtaining valid and reliable
18 historical exposure estimates are substantial limitations of these cohort studies.
19 The case-control studies also provide a robust collection of data supporting a similar
20 overall magnitude of association between formaldehyde exposure and nasopharyngeal cancer
21 (i.e., relative risk estimates in the range of 1.5-3.0) (see Table 4-3). These studies were
22 conducted in several different countries and work settings. Only one study (Armstrong et al.,
23 2000) did not observe a risk consistent with these estimates.
24 Of particular note were three sets of effect estimates of unusual magnitude, especially
25 when the definitions of exposure to formaldehyde were more specifically defined. Vaughan et
26 al. (1986b) reported that the odds ratio for living in a mobile home for 1-9 years was 2.1 (95%
27 CI: 0.7-6.6) compared to not living in a mobile home, while living in a mobile home for more
28 than 10 years had an OR = 5.5 (95% CI: 1.6-19.4). Vaughan et al. (1986b) also reported results
29 that could be considered as consistent with an exposure-response relationship. Compared to
30 having no exposure to mobile homes, they reported OR =1.7 (95% CI: 0.5-5.7) for working in a
31 mobile home, OR = 2.8 (95% CI: 1.0-7.9) for living in a mobile home and OR = 6.7 (95% CI:
32 1.2-38.9) for both living in and working in a mobile home. While exposure to mobile homes is a
33 crude proxy for formaldehyde exposure, the gradient in response and the magnitude of effect at
34 the highest exposure level is striking.
This document is a draft for review purposes only and does not constitute Agency policy.
4-116 DRAFT—DO NOT CITE OR QUOTE
-------
1 Vaughan et al. (2000) also demonstrated a version on an exposure-response gradient.
2 They reported a statistically significant trend of increasing risk of nasopharyngeal cancer with
3 the probability of being exposed to formaldehyde. "Possible, probable or definite" probability of
4 being exposed to formaldehyde having an OR =1.6 (95% CI 1.0-2.8), with "probable or
5 definite" exposure having an OR = 2.1 (5% CI 1.1, 4.2) and with "definite" exposure having an
6 OR = 13.3 (95% CI 2.5-70) (p-trend < 0.001).
7 The study by West et al. (1993) was also remarkable for the exposure-response shown for
8 formaldehyde and for anti-mosquito coils which have been shown to release high concentrations
9 of formaldehyde. In multivariate analyses, compared to never having had occupational exposure
10 to formaldehyde, having occupation exposure within the last 25 years had an OR =1.2 (95% CI:
11 0.41-3.6) while having occupational exposure more than 25 years earlier had an OR = 4.0 (95%
12 CI: 1.1-17.5). Compared to never using anti-mosquito coils, the OR for using them less than
13 daily was 1.4 (95% CI: 0.64-2.8) while daily use of anti-mosquito coils had an OR = 5.9 (95%
14 CI: 1.7-20.1).
15 Establishment of a concentration-response relationship is important evidence of causality.
16 An attenuation of risk is often seen at the highest dose level in epidemiologic studies, however, it
17 may possibly be due to exposure misclassification and of the use of inappropriate exposure
18 models (Stayner 1985, 1988). Identification of concentration-response relationships in spite of
19 potentially substantial effect measure attenuation provides even stronger evidence of causality.
20 Findings from the large NCI cohort studies of nasopharyngeal cancer risk (Hauptmann et al.,
21 2004) show a pattern of increased risk with increased formaldehyde exposures across multiple
22 exposure metrics.
23 Several reanalyses of the NCI cohort have challenged the interpretation of these findings,
24 although the reported excess in nasopharyngeal cancer mortality is also seen in these new
25 analyses (Marsh et al., 2007a, b, 2002, 1996; Marsh and Youk, 2005). The major questions that
26 have been raised by Marsh and coworkers highlight the observation that the nasopharyngeal
27 cancer findings appear to depend on the results of 1 of the 10 plants that made up the NCI cohort.
28 While it is possible for coexposures at that plant or among those workers to act as potential
29 confounders or modifiers of the observed effect of formaldehyde on increased risk of
30 nasopharyngeal cancer, there is no evidence of such a relationship as Hauptmann et al. (2004)
31 reported that controlling for each of 11 coexposures did not meaningfully alter the formaldehyde
32 association. While seven out of nine members of the cohort members at the Wallingford,
33 Connecticut, plant were also exposed to particulates, the NCI investigators did observe an
34 exposure-response relationship with formaldehyde among individuals with particulate exposures,
35 thereby strengthening the causal interpretation of the formaldehyde relationship with an
This document is a draft for review purposes only and does not constitute Agency policy.
4-117 DRAFT—DO NOT CITE OR QUOTE
-------
1 increased risk of nasopharyngeal cancer (Hauptmann et al., 2004). The described association of
2 a potential occupational relationship with silversmithing (brass plating, silver and other brass
3 work) and nasopharyngeal cancer has little support in the medical literature. In addition, if
4 silversmithing exposures are indeed independent risk factors for nasopharyngeal cancer, it would
5 be expected that the rates of nasopharyngeal cancer in the surrounding counties with historical
6 silver-and other metal-related exposures would be elevated. However they are not increased, as
7 evidenced by the comparability of the increased rates of nasopharyngeal cancer among the plant
8 workers based on the national and based on local county rates (Marsh et al., 2007a). The
9 comparable rates indicate the counties' rates of nasopharyngeal cancer were very similar to the
10 national rates. Given the limitations and the lack of comparability in the source of the work
11 history data in the nested case-control study of Marsh et al. (2007a) (next of kin interviews for
12 cases and self-interviews for controls) and the many post hoc re-examinations of alternative
13 hypotheses to explain the original NCI findings, EPA believes the association seen with
14 silversmithing does not represent an unmeasured confounder of the results of Hauptman et al.
15 (2004). A more likely explanation for the increased risk seen in short-term workers at the
16 Wallingford, CT plant is the specific work conditions of those jobs at that plant. It is also
17 plausible that the observed association at the Wallingford plant reflects higher formaldehyde
18 exposures than at other plants. Marsh and Youk (2005) reported that the exposure levels at Plant
19 2 were even higher than at the Wallingford plant; the fivefold increase in risk of nasopharyngeal
20 cancer seen at this plant, although based on only a single observed case, supports the results seen
21 in Wallingford, CT.
22 Two large cohort studies did not report associations between formaldehyde exposure and
23 increased risk of nasopharyngeal cancer (Coggon et al. 2003; Pinkerton et al., 2004). These
24 studies were extremely limited in terms of the observed and expected numbers of cases, with just
25 three cases expected between them during their follow-up periods. Thus these studies did not
26 have the statistical power to rule out effect sizes of the magnitudes reported from the studies of
27 the NCI cohort. As such, their results provide little contribution to the weight of evidence
28 regarding formaldehyde and nasopharyngeal cancer.
29 This evidence of a concentration-response relationship in the largest cohort study
30 (Hauptman et al., 2004) is further supported by a meta-analysis of the relative risks for
31 nasopharyngeal cancer by level or duration of exposure to formaldehyde which reported a
32 concentration-response relationship (p < 0.05) with an effect estimate of RR = 2.1 for the highest
33 exposure category based on five studies which reported sufficient information for an exposure-
34 response evaluation (Blair et al., 1990). A subsequent meta-analysis by Partanen (1993)
35 reabstracted the source data from the studies in the Blair et al. (1990) meta-analysis in an effort
This document is a draft for review purposes only and does not constitute Agency policy.
4-118 DRAFT—DO NOT CITE OR QUOTE
-------
1 to maximize the amount of relevant data from the source studies. The Partanen (1993) meta-
2 analysis confirmed the finding of Blair et al. (1990) that the relative risk of nasopharyngeal
3 cancer was increased among people exposed to 'substantial' levels or durations of formaldehyde
4 exposure. Partanen (1993) computed confidence intervals for the Blair et al. (1990) results for
5 the highest exposure category as RR = 2.06 (95% CI: 1.10-3.52); two other methods of
6 computing effect estimates and confidence intervals yielding effect estimates of RR = 2.74 (95%
7 CI: 1.36-5.55) and RR = 2.59 (95% CI: 1.29-5.36). More recent case-control studies also
8 provide evidence of increasing risks with longer duration or other measures of exposure
9 (Vaughan et al., 2000; West et al., 1993).
10
11 4.1.2.1.2. Nasal andparanasal (sinonasal) cancer.
12 4.1.2.1.2.1. Cohort studies of nasal and paranasal cancers.
13 IARC (1995) reported the results of several cohort studies of professional and industrial
14 workers exposed to formaldehyde for the induction of nasal and paranasal cancers (Andjelkovich
15 et al., 1995; Gardner et al., 1993; Hall et al., 1991; Hayes et al., 1990; Bertazzi et al., 1989;
16 Edling et al., 1987; Blair et al., 1986; Stroup et al., 1986; Harrington and Oakes, 1984; Levine et
17 al., 1984; Walrath and Fraumeni, 1984, 1983; Friedman and Ury, 1983). The likelihood of
18 finding this rare tumor type in even long-term cohort study is low, and only a few studies
19 reported any cases of nasal and paranasal cancer. No cases of this type of cancer were reported
20 in any of the studies of professional workers examined by the IARC. Only two cases (2.2
21 expected) were reported in the NCI cohort of more than 25,000 workers by Blair et al. (1986)
22 while the update by Hauptmann et al. (2004) identified three cases (2.5 expected). Marsh et al.
23 examined a subset of the NCI cohort from the Wallingford, Connecticut plant and identified two
24 cases compared to 0.5 in the 1984 follow-up (Marsh et al., 1996) and three cases against one
25 expected in the 1998 follow-up (Marsh et al., 2002). The cohort study of 14,017 British
26 chemical industry workers exposed to formaldehyde by Gardner et al. (1993) identified only one
27 case (1.7 expected). Coggon et al. (2003) updated this cohort study with additional follow-up
28 and reported only two cases compared to an expected number of 2.3 cases. The other large
29 cohort studies of more than 11,000 garment workers by Stayner et al. (1988) and Pinkerton et al.
30 (2004) did not identify any cases of sinonasal cancer. In contrast, a large population-based study
31 using nationwide Danish Product Register data found an increased risk of sinonasal cancer
32 (standardized proportional incidence ratio [SPIR]= 2.3 [95% CI: 1.3-4.0] based on 13 observed
33 cases; 2,041 of 91,182 cancer patients had at least 10 years of continuous formaldehyde-related
34 work experience before diagnosis (Hansen and Olsen, 1995).
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-119 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.1.2.1.2.2. Case-control studies of nasal and paranasal cancers.
2 Eight case-control studies were evaluated in the 1995 IARC monograph regarding the
3 risk of cancers of nasal cavity and accessory sinuses from exposure to formaldehyde (Luce et al.,
4 1993; Roush et al., 1987; Hayes et al., 1986; Olsen and Asnaes, 1986; Vaughan et al., 1986a, b;
5 Brinton et al., 1984; Olsen et al., 1984). Study details of the epidemiologic studies of nasal and
6 paranasal cancer are summarized in Table 4-4. Five studies were not limited to a specific cell
7 type and of these, only Roush et al. (1987) and Olsen et al. (1984) found positive results. The
8 remaining studies (Vaughan et al., 1986a, b; Brinton et al., 1984) did not find associations
9 between exposure and sinonasal cancer.
10 Three other studies conducted analyses limited to a specific cell type (Luce et al., 1993;
11 Hayes et al., 1986; Olsen and Asnaes, 1986) (see Table 4-4). Hayes et al (1986) conducted a
12 case-control study of occupational exposure to formaldehyde and the risk of nasal and paranasal
13 cancer in the Netherlands based on 116 cases and 259 controls. Analyses were done by
14 histological type of tumor and controlled for history of tobacco use and occupational exposure to
15 wood dust. Controlling for potential confounding by wood dust was important as wood dust was
16 reported to be strongly associated with sinonasal cancer (Hayes et al., 1986). Two independent
17 assessments of formaldehyde exposure were conducted with RRs of 3.0 (90% CI 1.3-6.4) and 1.9
18 (90% CI 1.0-3.6) seen for squamous cell carcinoma based on assessments A and B, respectively.
19 Olsen and Asnaes (1986) conducted a case-control of 287 histologically verified cancers
20 of the nasal cavity and 179 cases of paranasal cancer in Denmark. Controlling for wood dust
21 exposures, the investigators reported adjusted RR for formaldehyde exposure of 2.2 (95% CI:
22 0.9-5.8) for squamous cell carcinoma and RR = 2.3 (95% CI: 0.7-7.2) for adenocarcinoma. The
23 results for squamous cell carcinoma did not appear to be influenced by coexposure to wood dust;
24 however, for adenocarcinoma, among those who were never exposed to wood dust, the RR of
25 formaldehyde exposure was 7.0 (95% CI: 1.1-43.9) and among those were ever exposed to wood
26 dust the RR of formaldehyde exposure was 39.5 (95% CI: 22.0-70.8). Given the differences
27 between the stratum-specific results, which indicate effect modification, and between the
28 stratum-specific results and the adjusted results, it is the stratum-specific results that provide the
29 most valid measures of effect. The analyses by Olsen and Asnaes (1986) did not control for
30 smoking as that data was unavailable but tobacco use was not shown to be a confounder in the
31 study by Hayes et al (1986) or the study by Luce et al. (2002).
32 The third study that identified cancers by cell type (Luce et al., 1993) did not find an
33 association between formaldehyde exposure and squamous cell carcinoma after controlling for
34 age, wood dust, glues, and adhesives. For adenocarcinoma, the investigators attempted to
35 control for wood dust but found that almost all of the cases of were coexposure to formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-120 DRAFT—DO NOT CITE OR QUOTE
-------
1 and wood dust. Among workers with lower exposures to wood dust, the OR for any exposure to
2 formaldehyde was 8.1 (95% CI: 0.9-72.9). The OR for wood dust among those not exposed to
3 formaldehyde was 130 (95% CI: 14.2-1191). Since the large but not statistically significant
4 result for formaldehyde was not among those never exposed to wood dust but may have included
5 workers 'probably' exposed or 'definitely exposed to low lifetime levels,' residual confounding
6 by wood dust may have yielded the reported finding.
7 More recently, Luce et al. (2002) pooled data from 12 case-control studies. Pooled
8 studies are different from meta-analyses in that they compile the original raw data at the level of
9 the individual into a single pooled dataset while meta-analyses examine the summary results
10 across studies. Luce et al. (2002) assessed the associations between sinonasal cancer and
11 occupational exposures to formaldehyde after developing a common job exposure matrix
12 specifically for this study which standardized the different classifications of exposures and
13 potential confounders. Disease coding was also standardized to a single system to increase
14 compatibility. Combined, these studies contributed 195 adenocarcinomas and 432 squamous cell
15 carcinomas of the sinonasal passages compared with 3,136 controls. Analyses controlled for age
16 and study and were conducted separately by gender. Potential confounders were evaluated and
17 controlled for. Among men, the investigators were able to control for wood dust and leather dust
18 when evaluating the effects of formaldehyde on adenocarcinoma. The authors reported a
19 statistically significant increase in the risk of sinonasal adenocarcinoma in men (adjusted
20 OR = 3.0 [95% CI: 1.5-5.7]) and in women (adjusted OR = 6.2 [95% CI: 2.0-19.7]) with a high
21 probability of exposure to formaldehyde. Luce et al. (2002) checked for residual confounding by
22 wood dust by regrouping wood dust exposure into 5 categories, using the cumulative level of
23 wood dust as a continuous variable and using the highest lifetime level of exposure and reported
24 that the results were 'not markedly changed' indicating no evidence of residual confounding.
25 For squamous cell carcinomas, the ORs were more modest: OR = 1.2 in men and OR = 1.5 in
26 women for a high probability of exposure to formaldehyde. In an analysis of 11 formaldehyde-
27 exposed cases of sinonasal adenocarcinomas who were not exposed to wood dust, there was an
28 elevated risk in men (OR = 1.9; 3 cases) and a significantly increased risk in women (OR =11.1
29 [95% CI: 3.2-38.0]; 5 cases) with a high probability of exposure to formaldehyde. Limitations
30 of these studies were the lack of information about the actual levels or intensity of exposure to
31 formaldehyde, exposure to multiple occupational carcinogens, and the small number of cases in
32 some subgroups. In spite of those limitations, which generally obfuscate the observation of a
33 true underlying effect, the studies evaluated by Luce et al. (2002) identified effects of
34 formaldehyde that were statistically significant predictors of sinonasal cancers and this
35 association was not likely to be attributable to bias, chance or confounding.
This document is a draft for review purposes only and does not constitute Agency policy.
4-121 DRAFT—DO NOT CITE OR QUOTE
-------
ts
Table 4-4. Case-control studies of formaldehyde exposure and nasal and paranasal (sinonasal) cancers
^ 8
1!
f^ S
s ,
Reference
Study design
Exposure assessment
Results
Studies without specification by cell types
Brintonetal. (1984)
Olsenetal. (1984)
Vaughan et al.
(1986a),
Washington
Case-control study of 160
patients with cancer of the nasal
cavity and paranasal sinuses
from four North Carolina and
Virginia hospitals matched with
290 hospital controls with other
conditions. Odds ratios
adjusted for cigarette smoking,
alcohol consumption, gender,
and age.
Case-control study of 488 cases
of nasal cancer linked to the
Danish Cancer Registry during
1970-1982. Controls were
individuals with cancer of the
colon, rectum, breast, and
prostate. Three controls per
case were selected for the same
distributions of age, sex, and
year of diagnosis as cases.
Population-based, n = 53
incident cases (1980-1983)
from a 13 -county area
(Washington State Cancer
Surveillance System) and 552
matched controls from random
digit dialing in same area, for
occupational exposures.
Adjusted for cigarette smoking,
Interview data on job
history. Estimation of
exposure based on
industry type. Only two
cases employed in
industry associated with
formaldehyde. There
were no deaths in the
high exposure category.
Employment histories
after 1964 from files
maintained by Danish
Cancer Registry
estimated by industrial
hygienists.
Interview-based
information on lifetime
occupational exposure to
formaldehyde with cases,
next of kin, and controls.
Exposure from available
hygiene data, NIOSH
and other data, and NCI
job exposure linkage
system. Exposure score
Overall male and female
RR
0.35
(95% CI: 0.1-1.8)
Residence in mobile home
OR
0.6
(95% CI: 0.2-1. 7)
Years of exposure to particleboard
Ito9
10 or more
Men
Formaldehyde only
Ever exposed
Exposure to wood dust and
formaldehyde
Ever exposed
1st exposure >10 years or
more before diagnosis
OR
RR
RR
RR
RR
1.8
1.5
2.8
3.1
3.5
4.1
(95% CI: 09-3. 8)
(95% CI: 0.7-3. 2)
(95% CI: 1.8-4.3)
(95% CI: 1.8-5.3)
(95% CI: 2.2-5.6)
(95% CI: 0.2.3-7.3)
Intensity
Low
Medium/high
OR
0.8
0.3
(95% CI: 0.4-1. 7)
(95% CI: 0.0-1.3)
No. years exposed
1-9
10 or more
OR
0.7
0.4
(95% CI: 0.3-1.4)
(95% CI: 0.1-1.9)
to
to
fe
H
o
o
-------
§
'
Table 4-4. Case-control studies of formaldehyde exposure and nasal and paranasal (sinonasal) cancers (continued)
& s S
I"! 1
(^ ^ 2
<£• o ~-
111
^3 S ^
O w^ ^t>
£ | 3
[\
to
OJ
o
^
fe
Reference
Vaughan et al.
(1986a),
Washington
(continued)
Vaughan et al.
(1986b),
Washington
Roushetal. (1987),
Connecticut
Study design
alcohol consumption, gender,
and age.
Same cases and controls as
Vaughan etal. (1986a).
Adjusted for ethnic origin and
cigarette smoking.
Population-based case-control
study of 198 male cases of
sinonasal cancer from the
Connecticut Tumor Registry
who died of any cause in
1935-1975. Controls were 605
males dying in Connecticut
during the same time period,
randomly selected from state
death certificates. Adjusted for
age at death, year at death, and
availability of occupational
information.
Exposure assessment
based on sum of no.
years spent per job
weighted by estimated
formaldehyde level.
Same as Vaughan et al.
(1986a). Also included
residential history in past
50 years, and use of
particleboard or
plywood.
Four exposure categories
based on probability and
duration: I, probably
exposed most of working
life; II, probably exposed
most of working life and
probably exposed 20+
years before death; III,
probably exposed most
of working life and
probably to high level in
some year; IV, probably
exposed most of working
life and probably
exposed to high level
20+ years before death.
Results
Exposure score (no lag)
5-19
20 or more
OR
0.5
0.3
(95% CI: 0.1-1.6)
(95% CI: 0.0-2.3)
Exposure score (15 year lag)
5-19
20 or more
Residence in mobile home
OR
OR
1.0
0.0
0.6
(95% CI: 03-2.9)
(95% CI: -)
(95% CI: 0.2-1. 7)
Years of exposure to particleboard
1-9
10 or more
OR
1.8
1.5
(95% CI: 0.9-3.8)
(95% CI: 0.7-3. 2)
Sinonasal cancer
Exposure level
I
II
III
IV
ORa
0.8
1.0
1.0
1.5
(95% CI: 0.5-1.3)
(95% CI: 0.5-1.8)
(95% CI: 0.5-2.2)
(95% CI: 0.6-3. 9)
H
O
0
-------
Table 4-4. Case-control studies of formaldehyde exposure and nasal and paranasal (sinonasal) cancers (continued)
Reference
Study design
Exposure assessment
Results
Studies with specification by cell types
Hayes etal. (1986)
Olsen and Asnaes
(1986)
Case-control study of 91 men
with squamous cell carcinoma
of the nasal cavity and
paranasal sinuses, from clinical
records of six medical
institutions in the Netherlands.
195 controls from living and
deceased males from municipal
residence registries, from
1978-1981.
Case-control study of
histologically confirmed cases
of squamous cell
carcinoma/lymphoepithelioma
of the sinonasal cavities and
paranasal cancers in 215 men
and adenocarcinomas of the
sinonasal cavities and paranasal
cancers in 39 men matched with
2,465 controls with other
cancers from the Danish Cancer
Registry, 1970-1982.
Industrial hygienists
evaluated job histories
according to probability
of exposure based on job
records.
Employment histories
after 1964 from files
maintained by Danish
Cancer Registry
estimated by industrial
hygienists.
Squ
Indi
Indi
Squ
Eve
10 c
amous cell carcinoma
astrial hygienist A
Any exposure
Moderate exposure
High exposure
astrial hygienist B
Any exposure
Moderate exposure
High exposure
amous cell carcinoma/lyn
rvs. never
Formaldehyde only
Formaldehyde + wood
dust
>r more years since first exj
Formaldehyde only
Formaldehyde + wood
dust
Adenocarcinoma
Ever vs. Never
Formaldehyde only
Formaldehyde + wood
dust
RR
RR
nphoepit
RR
RR
josure
RR
RR
RR
RR
3.0
2.7
3.1
1.9
1.4
2.4
helioma
2.0
1.6
1.4
1.8
7.0
40.0
(90% CI: 1.3-6.4)
(90% CI: 1.0-7.2)
(90% CI: 0.9-10.0)
(90% CI: 1.0-3.6)
(90% CI: 0.5-3.4)
(90% CI: 1.1-5.1)
(95% CI: 0.7-5.9)
(95% CI: 0.8-3.3)
(95% CI: 0.3-6.4)
(95% CI: 0.7-4.4)
(95% CI: 1.1-44)
(95% CI: 22-71)
& s
II
5'
s ,
J^.
to
fe
H
U
O
-------
Table 4-4. Case-control studies of formaldehyde exposure and nasal and paranasal (sinonasal) cancers (continued)
Reference
Olsen and Asnaes
(1986) (continued)
Luce etal. (1993)
Study design
Case-control study of men with
sinonasal cancer (histologically
confirmed), 77 with
adenocarcinoma, 59 with
squamous cell carcinomas, and
25 tumors of other types,
matched with 409 controls from
27 French hospitals and
from lists of names supplied by
patients.bAll had medium to
high exposure to wood dust.
Adjusted for age and exposure
to glues and adhesives. adjusted
for age and study.
Exposure assessment
Industrial hygienist
estimation based on job
histories from personal
interviews. Exposure
classification based on
probability, frequency,
and concentration level.
Frequency categories (%
in normal week): 1 =
<5%, 2 = 5-30%, 3 =
>30%. Concentration
categories: low (0.0-0.1
ppm), medium (0.1-1
ppm), high (>1 ppm).
Exposure index =
concentration x
frequency. Cumulative
level = sum of exposure
indices. Average level =
cumulative
level/duration and ranged
from 1 to 9. Nearly all
cases had had wood dust
exposure.
Results
10 c
>r more years since first exj
Formaldehyde only
Formaldehyde + wood
dust
josure
RR
RR
9.5
44.0
(95% CI: 1.6-58)
(95% CI: 22-88)
Adenocarcinoma
Possible exposure
OR
1.28
(95% CI: 0.16-10)
Probable/definite exposure
Average level
<2
>2
OR
4.15
5.33
(95% CI: 0.96-18)
(95% CI: 1.28-22)
Duration (years')
\v - /
<20
>20
OR
1.03
6.86
Cumulative level (years)
<30
30-60
>60
OR
1.13
2.66
6.91
(95% CI: 0.18-5.77)
(95% CI: 1.69-28)
(95% CI: 0.19-6.95)
(95% CI: 0.38-19)
(95% CI: 1.69-28)
Age 1st exposed (years)
<15
16-20
>20
OR
9.99
4.12
2.74
(95% CI: 1.85-54)
(95% CI: 0.95-18)
(95% CI: 0.58-13)
Date 1st exposed (years)
After 1954
Before 1954
OR
6.02
4.26
(95% CI: 1.18-31)
(95% CI: 1.06-17)
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
s ,
J^.
to
fe
H
o
o
-------
Table 4-4. Case-control studies of formaldehyde exposure and nasal and paranasal (sinonasal) cancers (continued)
•
>j 01
(^ s
' § S'
l-o
a ,2
Reference
Luce etal. (1993)
(continued)
Study design
Exposure assessment
Results
Other cell type carcinoma
Possible exposure
OR
0.81
(95% CI: 0.15-4.36)
Probable/definite exposure
Average level
<2
>2
OR
1.67
3.04
(95% CI: 0.51-5.42)
(95% CI: 0.95-9.7)
Duration (years)
<20
>20
OR
2.82
1.62
(95% CI: 0.94-8.4)
(95% CI: 0.48-5.51)
Cumulative level (years)
<30
>30
OR
2.18
2.21
(95% CI: 0.65-7.31)
(95% CI: 0.73 -6.73)
Age 1st exposed (years)
<20
>20
OR
2.03
2.36
(95% CI: 0.63 -6.54)
(95% CI: 0.76-7.33)
Date 1st exposed (years)
After 1954
Before 1954
OR
0.48
3.27
(95% CI: 0.05-4.35)
(95% CI: 1.15-9.33)
to
Oi
fe
H
-------
This documen
*eview purpose,
not constitute ,
•(9 ° 5'
a ||. a
^ § 1"
O w^ ^J5
^
Table 4-4. Case-control studies of formaldehyde exposure and nasal and paranasal (sinonasal) cancers (continued)
Reference
Luce et al. (2002)
Study design
Pooled analysis of 195
adenocarcinomas and 432
squamous cell carcinomas of
the sinus/nasal cavity matched
with 3,136 controls from 12
case-control studies.
Adenocarcinoma results in men
adjusted for age, study, and
cumulative exposure to wood
and leather dust. All other
results adjusted for age and
study.
Exposure assessment
Job exposure matrix
based on interview data
developed for pooled
analysis. Industrial
hygiene data used to
develop indices of
exposure. 11
formaldehyde cases
reported no exposure to
wood dust.
Results
Adenocarcinoma
High probability of exposure
Men
Women
ORd
ORe
3.0
6.2
(95% CI: 1.5-5.7)
(95% CI: 2.0-20)
Squamous cell carcinoma
High probability of exposure
Men
Women
ORe
ORe
1.2
1.5
(95% CI: 0.8-1.8)
(95% CI: 0.6-3. 8)
-1^
to
fe
H
O
O
-------
1 4.1.2.1.2.3. Summary of nasal and paranasal cancers.
2 Rare outcomes are difficult to study due to a lack of statistical power to identify risks.
3 Most cohort studies observed no more than one case of sinonasal cancer and may not have had
4 sufficient statistical power to show an association. The exception is the national population-
5 based study linking the Danish cancer registry and industry data in which 13 cases were
6 identified for a SPIR of 2.3 (95% CI 1.3-4.0) (Hansen and Olsen, 1995). The case-control
7 studies of sinonasal cancer in which no excess risk was seen did not distinguish cancer type and
8 thus may have aggregated a truly causal relationship with adenocarcinoma and a noncausal
9 relationship with squamous cell carcinoma. Among all the studies of sinonasal cancer, the
10 pooled case-control study of Luce et al. (2002) provides the strongest causal evidence of an
11 association between formaldehyde exposure and increased risk of sinonasal adenocarcinoma. In
12 summary, there appears to be increased risk of sinonasal cancer associated with formaldehyde
13 exposure with or without exposure to wood dust (Luce et al., 2002). The effect appears to be
14 stronger when the risk is stratified by cancer type with higher risks of adenocarcinoma compared
15 with squamous cell carcinoma. Taken together with the nasopharyngeal cancer findings in the
16 neighboring tissue, EPA concludes that there is evidence of a causal association of sinonasal
17 cancer associated with exposure to formaldehyde.
18
19 4.1.2.1.3. Cohort studies of lung cancer.
20 Several industrial cohort studies (Andjelkovich et al., 1995; Stayner et al., 1988; Edling
21 et al., 1987) reported no significant excess risks of lung cancer from exposure to formaldehyde.
22 Bertazzi et al. (1986) studies a cohort of 1322 workers producing formaldehyde resins in Italy
23 and reported an increased risk of lung cancer in the initial analysis with an SMR of 1.86 (95%
24 CL1.10-2.93) based on 18 cases; however, the reanalysis of this cohort with six additional years
25 of follow-up (Bertazzi et al., 1989) showed no excess risk with 24 cases of lung cancer compared
26 to 23.9 expected. No consistent association between formaldehyde exposure and lung cancer
27 was found in several reports of the NCI 10-plant cohort study originally investigated by Blair et
28 al. (1987, 1986). Hauptmann et al. (2004) give the most recent report on this cohort, which has
29 been studied in part or in its entirety by several others (Marsh et al., 1994, 1992a, b; Sterling and
30 Weinkam, 1994, 1989a, b, 1988; Robins et al., 1988; Liebling et al., 1984; Fayerweather et al.,
31 1983; Wong, 1983; Marsh, 1982). Hauptmann et al. (2004) reported 641 cases of lung cancer
32 against 660.8 expected for an SMR of 0.97 (95% CI: 0.90-1.05).
33 Gardner et al. (1993) found a modest association between lung cancer and formaldehyde
34 exposure in the British chemical industry cohort study (n = 14,016) among men hired before
35 1965 with a SMR of 1.23 (95% CI: 1.10-1.36) based on national rates of expected deaths and a
This document is a draft for review purposes only and does not constitute Agency policy.
4-128 DRAFT—DO NOT CITE OR QUOTE
-------
1 SMR of 1.12 (95% CI: 1.00-1.24) based on local deaths rates. In workers hired after 1964, the
2 corresponding SMRs were 1.14 (95% CI: 0.85-1.48) and 1.13 (95% CI: 0.85-1.47). No trends by
3 level or duration of exposure were found (Gardner et al., 1993). Coggon et al. (2003) updated
4 the Gardner et al. (1993) cohort study with additional follow-up through the end of 2000 and
5 reported the effects of formaldehyde exposure on mortality from lung cancer among
6 14,014 workers. Coggon et al. (2003) reported an SMR of 1.22 (95% CI: 1.12-1.32) for lung
7 cancer in the entire cohort. In the analysis of the combined data from the 6 factories, an excess
8 risk of lung cancer was seen in the high-exposure category when compared with British national
9 mortality rates (SMR =1.58 [95% CI: 1.40-1.78]) and with local mortality rates (SMR = 1.28
10 [95% CI: 1.13-1.44]). Pinkerton et al. (2004) and Stayner et al. (1988) studied a cohort of
11 11,030 workers in three garment plants and found an SMR of 1.14 for lung cancer (95% CI:
12 0.81-1.56). Callas et al. (1996) reanalyzed the cumulative exposure of 279 lung cancer cases
13 among white male workers from the NCI study, which comprised 80% of the NCI cohort (Blair
14 et al., 1986). The analysis revealed RRs of 1.46, 1.27, and 1.38 for lung cancer in the cumulative
15 exposure categories of 0.05 to 0.5 ppm-years, 0.51 to 5.5 ppm-years, and > 5.5 ppm-years,
16 respectively. None of these RRs were statistically significant.
17 None of the cohort studies of workers in specific professions indicated excess risks of
18 lung cancer. Of the professional studies reviewed, the RRs range from an extremely low SMR
19 value of 0.2 in Hall et al. (1991), based on nine deaths, to an RR (proportional mortality ratio) of
20 1.1, based on 70 lung cancer deaths in Walrath and Fraumeni (1983). Matanoski (1991) reported
21 a significant deficit in the risk of respiratory cancer (SMR = 0.56 [95% CI: 0.44-0.70]; 77
22 observed) in pathologists presumably exposed to formaldehyde based on U.S. mortality rates.
23
24 4.1.2.1.4. Case-control studies of lung cancer.
25 Several case-control studies of lung cancer (Partanen et al., 1990; Gerin et al., 1989;
26 Bond et al., 1986; Coggon et al., 1984; Fayerweather et al., 1983; Anderson et al., 1982) showed
27 no excess lung cancer risk associated with potential exposure to formaldehyde when analyzed by
28 length of exposure, intensity, and potential exposure 5, 10, or 15 years before death or by
29 combinations of these factors. By contrast, Coggon et al. (1984) reported an increased in risk of
30 lung cancer among male patients with any potential exposure to formaldehyde based on
31 occupations listed on death certificates (SMR =1.5 [95% CI: 1.2-1.8]). De Stefani et al. (2005)
32 conducted a case-control study of 338 adenocarcinomas of the lung in male patients admitted to
33 four Montevideo hospitals in Uruguay from 1994 to 2000. Three agents (i.e., asbestos, silica
34 dust, and formaldehyde) indicated excess risks of lung adenocarcinoma after adjusting for
35 smoking history. The association seen with 21 or more years of exposure to formaldehyde was
This document is a draft for review purposes only and does not constitute Agency policy.
4-129 DRAFT—DO NOT CITE OR QUOTE
-------
1 OR = 3.0 (95% CI: 1.6-5.8), with an exposure-response trend seen across exposure levels (trend
2 p = 0.004).
O
4 Summary of lung cancer. Evidence of a relationship between formaldehyde exposure and lung
5 cancer is relatively weak and conflicting, with some studies showing modest increases (e.g.,
6 relative risks of 1.2, as seen in Coggon et al., 2003) while others show inverse associations
7 between exposure and risk. In all studies of formaldehyde and lung cancer, smoking remains an
8 important confounder and possibly an effect modifier. Residual confounding of smoking or
9 other respiratory exposures (e.g., wood dust or chemical or particular exposures) must always be
10 considered. A meta-analysis of workers exposed to formaldehyde by Bosetti et al. (2008)
11 reported a summary effect estimate of RR = 1.06 (95% CI: 0.92-1.23) for industry workers and
12 RR = 0.63 (95% CI: 0.47-0.84) for professionals. The meta-analysis by Blair et al. (1990)
13 reported a RR = 0.9 for professional and nonoccupational exposures and RR =1.1 among
14 industrial workers. The meta-analysis by Partanen (1993) largely confirmed the results of Blair
15 et al. (1990). Except for the findings of De Stefani et al. (2005), other studies of lung cancer and
16 exposure to formaldehyde have not supported the finding by Coggon et al. (2003), including
17 several well-done cohort studies that were specifically designed to evaluate lung cancer.
18
19 4.1.2.1.5. Other respiratory tract cancers.
20 Other cancers of the upper respiratory tract include buccal cavity, salivary gland, floor of
21 the mouth, other mouth, larynx, pharynx (including oropharynx and hypopharynx)
22
23 4.1.2.1.5.1. Cohort studies of other respiratory tract cancers.
24 In studies of industrial worker cohorts where buccal/pharyngeal cancer was examined
25 (Andjelkovich et al., 1995; Stayner et al., 1988; Blair et al., 1986), only one (Stayner et al., 1988)
26 reported an excess risk of death from this tumor (SMR = 3.4, based on four deaths in a cohort of
27 6,741 white women). Of six cohort studies of buccal/pharyngeal cancer in studies of
28 professionals reviewed by IARC (Hayes et al., 1990; Logue et al., 1986; Stroup et al., 1986;
29 Levine et al., 1984; Walrath and Fraumeni, 1984, 1983), no evidence of a risk associated with
30 exposure to formaldehyde was reported.
31 Hansen and Olsen (1995) reported a 10% increase in the risk of cancer of the buccal
32 cavity and pharynx (SPIR =1.1, 95% CI 0.7, 1.7) in their proportional incidence study of Danish
33 workers. Marsh et al. (1996) reported no significant excess risk of buccal cavity cancer cases
34 (SMR = 1.31) based on U. S. mortality rates and no excess based on State mortality rates
35 (SMR = 1.0). For oropharyngeal cancer, the SMR was 1.84 (based on two cases), the SMR for
This document is a draft for review purposes only and does not constitute Agency policy.
4-130 DRAFT—DO NOT CITE OR QUOTE
-------
1 hypopharyngeal cancer was 1.41 (based on one case), and the SMR for laryngeal cancer was
2 1.47 (based on six cases). The latter risks were elevated even when SMRs were derived from
3 Connecticut mortality rates.
4 A nested case-control study, based on the 22 pharyngeal cancer (including the 7
5 nasopharyngeal cancers) deaths in the Wallingford, Connecticut, plant cohort was also reported
6 in Marsh et al. (2002). Each of the pharyngeal cancer deaths was matched on race, sex, age, and
7 year of birth to four controls from the cohort. Data on smoking status and date of hire were
8 collected and controlled for in the analyses of the pharyngeal cancers but apparently not for
9 nasopharyngeal cancer due to sparse data. Twenty of the 22 cases were exposed to
10 formaldehyde, yielding an odds ratio (OR) of 3.04 (95% CI: 0.36-145.58) after adjustment for
11 smoking and year of hire. There was little or no association of pharyngeal cancer incidence in
12 these workers with either average or cumulative exposure, based on the exposure estimates in
13 this study. There was a suggested trend of increasing OR with increasing duration of exposure
14 for any formaldehyde exposure as well as for formaldehyde exposure >0.2 ppm. The relatively
15 flat dose-response curve in the nested case-control study contradicts the positive dose-response
16 curves reported (particularly for nasopharyngeal cancers) in the same study, based on SMRs
17 derived from county and U.S. death rates in the cohort analysis.
18 Hauptmann et al. (2004) combined upper respiratory tract cancers (cancers of the salivary
19 gland, mouth, nasopharynx, nasal cavity, and larynx; n=\\ observed deaths). For average
20 intensity of exposure, the RR was 1.47, 1.69, and 2.21 for the low (> 0-0.5 ppm), medium (0.5
21 to <1.0 ppm) and high (> 1.0 ppm) exposure categories, respectively, (trend/? = 0.122).
22 Confidence intervals were not provided for these RR; however, a footnote states that for the
23 RR = 2.21 for the highest exposure category (> 1.0 ppm), the 95% confidence interval does not
24 include 1.00. For peak exposure, the RR was 1.32, 1.24, and 1.65 for the low (> 0- <2.0 ppm),
25 medium (2.0- <4.0 ppm) and high (>4.0 ppm) exposure categories, respectively (trend
26 p = 0.142). For cumulative exposure, the RR was 1.24, 1.92 and 0.86 for the low (> 0-1.5 ppm-
27 yrs), medium (1.5- <5.5 ppm-yrs) and high (>5.5 ppm-yrs) exposure categories (trend
28 p = 0.0.765). Increased relative risks of cancer of the buccal cavity (which included salivary
29 glands, the floor of the mouth, other mouth and the nasopharynx) were reported for the medium
30 and high exposure categories with 95% confidence intervals that did not include 1.00 for the
31 medium exposure category. Hauptmann et al. (2004) concluded that in spite of the small number
32 of deaths from these rare cancers of the upper respiratory tract, the positive associations of
33 increased cancer risk with increased formaldehyde exposure were consistent with the
34 carcinogenicity of formaldehyde at these sites of first contact.
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-131 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.1.2.1.5.2. Case-control studies of other respiratory tract cancers.
2 Three case-control studies (Merletti et al., 1991; Vaughan et al., 1986a, b) did not find an
3 association between oral, oropharyngeal, and hypopharyngeal cancers and formaldehyde
4 exposure. However, Merletti et al. (1991) found an elevated OR of 1.8 associated with probable
5 or definite exposure to formaldehyde in a study of 86 patients with oral or oropharyngeal cancer
6 matched with 373 controls. There was no increased risk of laryngeal cancer associated with
7 formaldehyde in a case-control study (Wortley et al., 1992) of 235 patients with laryngeal cancer
8 and 547 controls. The OR in that study was 1.0 (adjusted for age, smoking, drinking, and level
9 of education). IARC (1995) concluded that there was little evidence of an increased risk of
10 laryngeal cancer.
11 Three additional studies, published since the IARC (1995) report of other respiratory
12 cancers (e.g., oral cavity, oropharynx, and/or larynx), are available (Shangina et al., 2006;
13 Laforest et al., 2000; Gustavsson et al., 1998). Gustavsson et al. (1998) conducted a case-control
14 study of 545 cases of squamous cell carcinoma of the oral cavity, oropharynx, hypopharynx,
15 larynx, and esophagus and frequency-matched by age and region with 641 controls. Regression
16 analyses among 545 male cases showed elevated but nonsignificant risks of squamous cell
17 carcinoma of the oral cavity (OR = 1.28), esophagus (OR = 1.90), and larynx (OR = 1.45)
18 associated with formaldehyde exposure. However, several of the carcinoma types were
19 statistically significantly associated with exposure to welding fumes, polyaromatic hydrocarbons,
20 asbestos, and metal dust. Laforest et al. (2000) examined 201 patients with squamous cell
21 carcinoma of hypopharynx and 296 patients with squamous cell carcinoma of larynx, who were
22 matched to 296 controls with cancers of other sites in 15 French hospitals. Adjusting for
23 potential confounders, the OR of hypopharyngeal cancer in patients with a high probability of
24 exposure to formaldehyde was 3.78 (95% CI: 1.50-9.49). The ORs were significantly increased
25 with both exposure durations and high cumulative level of exposure. Shangina et al. (2006)
26 conducted a multicentered case-control study of 34 cases of hypopharyngeal cancer, 316 cases of
27 laryngeal cancer, and 728 hospital-based controls. The study was based on Romania, Poland,
28 Russia and Slovakia. Data was collected through a structured interview and included a detailed
29 work history including information on all jobs held for at least one year and more detailed
30 modules for specific jobs. These data were evaluated by local industrial hygienists, chemists,
31 and physicians, blinded to disease status. Exposure classifications were developed for 73 agents;
32 frequency was estimated as the proportion of time a worker was exposed. Linear trends were
33 examined for duration in years, weighted duration in hours, and cumulative exposure. The OR
34 for formaldehyde exposure and laryngeal cancer was 1.68 (95% CI: 0.85-3.31). Trends over
35 increasing exposure were found for duration of exposure in years (p = 0.06) and for cumulative
This document is a draft for review purposes only and does not constitute Agency policy.
4-132 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposures (p = 0.07). The investigators reported an OR of 3.12 (95% CI: 1.23-7.91) for the
2 highest cumulative exposure group (>22,700 mg/m3-hours) compared with the unexposed group.
O
4 4.1.2.1.5.3. Summary of other respiratory tract cancers.
5 The evidence for a compound-specific effect on the risk of buccal/pharynx, oral cavity,
6 oropharynx, hypopharynx, and laryngeal cancers as a result of exposure to formaldehyde is slight
7 when examining each rare endpoint on its own. The study by Hauptmann et al. (2004) did show
8 an increased risk of cancers of the buccal cavity and the upper respiratory tract as a group that
9 had confidence interval that did not include 1.00. Further evidence comes from the results of the
10 study by Laforest et al. (2000) showing nearly four-fold increased risk of hypopharyngeal cancer
11 and, to a lesser extent, that by Shangina et al. (2006) who demonstrated an exposure-response
12 relationship for duration of formaldehyde exposure and risk of laryngeal cancer. However, taken
13 together with the causal evidence of an association between formaldehyde and nasopharyngeal
14 and sinonasal cancers in neighboring tissues of the upper respiratory tract and sites of first
15 contact, these sporadic results in humans may indicate a broader pattern of carcinogen!city
16 within the upper respiratory tract.
17
18 4.1.2.1.5.4. Summary of respiratory tract cancers.
19 The epidemiologic studies of nasopharyngeal cancer provide strong support for a causal
20 role of formaldehyde exposure to the etiology of this rare cancer. The Hauptmann et al. (2004)
21 analysis of the large NCI cohort study of 10 worksites in the United States reported a doubling of
22 nasopharyngeal cancer mortality risk in workers (SMR = 2.1, 95% CI: 1.05-4.21). Hauptmann
23 et al. demonstrated that relative risks increased with average exposure intensity (p-
24 trend = 0.066), cumulative exposure (p-trend = 0.025), highest peak exposure (p-trend < 0.001)
25 and duration of exposure to formaldehyde (p-trend = 0.147). The case-control studies also
26 provide a robust collection of data supporting a similar association between formaldehyde
27 exposure and nasopharyngeal cancer (i.e., relative risk estimates in the range of 1.5-3.0). The
28 association between formaldehyde and nasopharyngeal cancer persisted when adjusted for the
29 effect of potential confounders. Data from several of these studies have identified a
30 concentration-response relationship (Hauptmann et al., 2004; Marsh et al., 2002; Vaughan et al.,
31 2000; West et al., 1993), adding weight to the causal evidence of an association between
32 formaldehyde exposure and cancer. The studies of the single Wallingford plant by Marsh et al.
33 (2002, 1996, 1994) and Marsh and Youk (2005) also revealed a dose-response trend, although
34 the absolute exposure level estimates were much lower according to Marsh et al. (2002).
This document is a draft for review purposes only and does not constitute Agency policy.
4-133 DRAFT—DO NOT CITE OR QUOTE
-------
1 Sinonasal cancer is also a rare cancer, and most cohort studies have had insufficient
2 statistical power to show an association. The exception is the national population-based study
3 linking the Danish Cancer Registry and industry data in which 13 cases were identified for a
4 SPIR of 2.3 (95% CI 1.3-4.0) (Hansen and Olsen, 1995). The case-control studies of sinonasal
5 cancer in which no excess risk was seen, did not distinguish cancer type and thus may have
6 aggregated a truly causal relationship with adenocarcinoma with a noncausal relationship with
7 squamous cell carcinoma. Among all the studies of sinonasal cancer, the pooled analysis by
8 Luce et al. (2002) provides the strongest evidence of a relationship between formaldehyde and
9 sinonasal cancer, particularly adenocarcinoma, which did not appear to be confounded by
10 concurrent exposure to wood dust. The effect appears to be stronger when the risk is stratified
11 by cancer type with higher risks of adenocarcinoma compared with squamous cell carcinoma.
12 Taken together with the nasopharyngeal cancer findings in the neighboring tissue, EPA
13 concludes that there is evidence of a causal association of sinonasal cancer associated with
14 exposure to formaldehyde.
15 Evidence of a relationship between formaldehyde exposure and lung cancer is relatively
16 weak and conflicting, with some studies showing modest increases (e.g., relative risks of 1.2, as
17 seen in Coggon et al., 2003) while others show inverse associations between exposure and risk.
18 The evidence for a compound-specific effect on the risk of buccal/pharyngeal, oral,
19 oropharyngeal, hypopharyngeal, and laryngeal cancers as a result of exposure to formaldehyde is
20 slight when examining each rare endpoint on its own. The study by Hauptmann et al. (2004) did
21 show an increased risk of cancers of the buccal cavity and the upper respiratory tract as a group
22 that had confidence interval that did not include 1.00. Further evidence comes from the results
23 of the study by Laforest et al. (2000) showing nearly four-fold increased risk of hypopharyngeal
24 cancer and, to a lesser extent, that by Shangina et al. (2006) who demonstrated an exposure-
25 response relationship for duration of formaldehyde exposure and risk of laryngeal cancer.
26 However, taken together with the causal evidence of an association between formaldehyde and
27 nasopharyngeal cancer and sinonasal cancer in neighboring tissues of the upper respiratory tract
28 and sites of first contact, these sporadic results in humans may indicate a broader pattern of
29 carcinogenicity within the upper respiratory tract.
30
31 4.1.2.2. Nonrespiratory Tract Cancer
32 4.1.2.2.1. Lymphohematopoietic (LHP) cancers.
33 Cancers of the hematopoietic system include lymphosarcoma, reticulosarcoma,
34 Hodgkin's disease, non-Hodgkin's disease, multiple myeloma, and all types of leukemia,
35 including lymphoid and myeloid. Virtually all of the studies of LHP cancers and formaldehyde
36 are cohort studies and are divided into two groups: professional and industrial. Several of the
This document is a draft for review purposes only and does not constitute Agency policy.
4-134 DRAFT—DO NOT CITE OR QUOTE
-------
1 studies of professional groups were reviewed in an IARC (1995) monograph and are briefly
2 discussed in the next section regarding their findings on cancer of the LHP system. One case-
3 control study of non-Hodgkin's lymphoma is discussed at the end of this section.
4
5 4.1.2.2.1.1. Professional cohort studies.
6 Several cohort studies have been undertaken by professional groups (i.e., anatomists,
7 pathologists, embalmers, and funeral directors) because their careers are likely to bring them into
8 contact with formaldehyde. Some studies have reported an increase in the risk of myelogenous
9 leukemia and other LHP cancers (see Table 4-5). A few of the increased risks were statistically
10 significant. None of the studies of professionals have used personal exposure measurements of
11 formaldehyde or other chemicals, making specificity for any single exposure difficult to
12 determine.
13 Harrington and Shannon (1975) conducted a cohort mortality study of 2,079 British
14 pathologists (1955-1973) and 12,944 British medical laboratory technicians (1963-1973).
15 When compared with death rates for England and Wales, the all-cause SMR for the pathologists
16 was 0.60 versus 0.67 for the laboratory technicians. There was a significant increase in the risk
17 of lymphatic and hematopoietic neoplasia (SMR 2.0; 8 observed with 3.3 expected;/? < 0.01)
18 among male pathologists. However, the SMR for technicians was only 0.6 (3 observed). The
19 low SMRs suggest that these professionals have a healthier profile compared with the British
20 population. No actual exposure estimates are available.
21 Harrington and Oakes (1984) expanded the above study to include 2,307 male and 413
22 female pathologists. Mortality was only examined from 1973 until 1980; deaths that occurred
23 before 1974 were not included in the update. The SMR for leukemia was 0.91 in men and 9.26
24 (based on one case) in women. Although the earlier LHP cancer deaths were not included in this
25 analysis, the investigators say in their conclusion that their previous suggestion of an increase in
26 certain lymphatic neoplasia was not confirmed in the present study because of small numbers.
27 The exceptionally low SMRs suggest that this group of professionals enjoyed a healthier lifestyle
28 compared with the British population as a whole. Just as in the earlier studies of these
29 professionals, no exposure estimates are available.
30 Hall et al. (1991) expanded the above study by including the newest members of the
31 Royal College of Pathologists. The cohort totaled 4,512 individuals, although only 3,069 males
32 and 803 females were included in the analysis. The reasons for this discrepancy were not
33 specified, although the authors mentioned that an unknown number of expected deaths for
34 Northern Irish and female Scottish pathologists were not calculated, 32 pathologists were lost in
35 follow-up, and cause of death was unknown for 9 individuals. Follow-up was extended from
This document is a draft for review purposes only and does not constitute Agency policy.
4-135 DRAFT—DO NOT CITE OR QUOTE
-------
ts
s ,
Table 4-5. Epidemiologic studies of formaldehyde and pharyngeal cancer (includes nasopharyngeal cancer)
Reference
Marsh et al. (2002)
Study design
Retrospective cohort
mortality study of 7,328
workers hired up to 1984
and followed until 1998 in
one plant from Blair et al.
(1986, 1987) and
Hauptmann et al. (2004).
Mortality was compared
with death rates in two
Connecticut counties and
U.S. A nested case-control
analysis was also conducted
with 4 controls matched on
age, year of birth, race, and
sex randomly selected from
cohort. Conditional logistic
model was used for nested
case-control analysis.
Exposure assessment
Worker-specific exposures
from job exposure matrix
were based on available
sporadic sampling data from
1965-1987, job descriptions,
and verbal job descriptions
by plant personnel and
industrial hygienists.
Exposures were then ranked
on a 7-point scale. An
exposure range was assigned
to each rank. 17% of jobs
validated with company
monitoring data, remaining
83% based on professional
judgment. Pre-1965 levels of
formaldehyde were assumed
to be the same as post- 1965
levels.
Results, statistical significance (number observed deaths for cohort study)
Cohort study
Overall
U.S.
County
SMR
SMR
Short-term worker (<1 year)
SMR
2.63
2.23
2.35
Long-term worker (1 or more years)
SMR
2.10
Cumulative exp. (ppm-years) county
Unexposed
>0 to <0.004
0.004-0.219
0.22+
SMR
SMR
SMR
SMR
1.24
3.31
2.06
2.30
(95% CI: 1.65-3.98)
(95% CI: 1.40-3.38)
(95% CI: 1.22-4.11)
(95% CI: 1.01-3.86)
(95% CI: 0.15-4.49)
(95% CI: 1.22-7.21)
(95% CI: 0.83-4.24)
(95% CI: 0.92-4.73)
(22)
(22)
(12)
(10)
(2)
(6)
(7)
(7)
Average Exposure (ppm) county
Unexposed
>0 to O.03
0.03-0.159
0.16+
Exposure to <0.2 ppm
Exposure to >0.2 ppm
Exposure to <0.7 ppm
SMR
SMR
SMR
SMR
SMR
SMR
SMR
Nested case-control analysis
Cumulative exp. (ppm-years)
O.004
0.004-0.219
0.22+
OR
OR
OR
1.24
2.02
3.82
2.03
1.72
2.68
2.12
1.00
0.71
0.79
(95% CI: 0.15-4.49)
(95% CI: 0.74-4.40)
(95% CI: 1.54-7.88)
(95% CI: 0.82-4. 19)
(95% CI: 0.74-3.39)
(95% CI: 1.46-4.49)
(95% CI: 1.21-3.45)
(95% CI: 0.20-2.43)
(95% CI: 0.18-3.20)
(2)
(6)
(7)
(7)
(8)
(14)
(16)
(8)
(7)
(7)
fe
H
O
o
-------
ts
^ 8
1!
f^ S
3 ,
Table 4-5. Epidemiologic studies of formaldehyde and pharyngeal cancer (includes nasopharyngeal cancer)
(continued)
Reference
Marsh et al. (2002)
(continued)
Coggon et al. (2003)
Shangina et al.
(2006)
Study design
Cohort mortality study of
14,014 chemical workers
employed in 6 British
factories.
Multicentered, hospital-
based case-control study in
four European countries;
men only. Cancer cases: 34
hypopharyngeal; 316
laryngeal. Controls: 72%
hospital patients with
various conditions.
Exposure assessment
Based on data abstracted
from company records. Each
job was categorized as having
background, low, moderate,
high, or unknown levels of
formaldehyde.
Exposures determined by
local industrial hygienists,
chemists, and physicians.
Coding was established and
standardized. Categories
were developed for 73
agents; frequency was
estimated as the proportion of
time a worker was exposed.
Linear trends were examined
for duration in years,
weighted duration in hours,
and cumulative exposure.
Results, statistical significance (number observed deaths for cohort study)
Average exposure (ppm)
0.03
0.03-0.159
0.16+
Exposure to >0.2 ppm
Exposure to >0.7 ppm
Overall
High exposure
Larvnseal cancer:
Formaldehyde
Ever vs. never
Highest cumulative
(>22,100 mg/m3-hours)
vs. lowest
Tests of trends:
Years exposed
Cumulative exposure
OR
OR
OR
OR
OR
SMR
SMR
OR
OR
p = 0.06
p = 0.07
1.00
1.71
0.99
1.35
1.60
1.55
1.91
1.68
3.12
(95% CI: 0.47-6. 10)
(95% CI: 0.27-3. 55)
(95% CI: 0.45-4.25)
(95% CI: 0.15-9.77)
(95% CI: 0.87-2.56)
(95% CI: 0.70-4. 17)
(95% CI: 0.85-3.31)
(95% CI: 1.23-7.91)
(8)
(7)
(7)
(14)
(6)
(15)
(6)
fe
H
o
o
-------
1 1980 to 1986. Mortality was enumerated from 1974 to 1987, a period of time that differed from
2 both of the earlier studies described above. There were statistically not significant excess risks
3 for lymphatic and hematopoietic cancer (SMR 1.44; 10 observed) and leukemia (SMR 1.52;
4 4 observed) for both sexes combined, based on mortality rates in England and Wales.
5 Separately, there was 1 female death in the lymphatic and hematopoietic cancer category (0.57
6 expected). The most striking observation in this study is that, despite the low cancer mortality
7 (SMR 0.45 for all cancer; 53 observed but 118.19 expected), there was still an excess (but not
8 statistically significant) risk of hematopoietic cancers. This finding of an extremely low risk for
9 all cancers suggests that population death rates may not be appropriate as a referent group—for
10 example, the SMRs for lung cancer (0.19) and nonneoplastic respiratory diseases (0.23) were
11 significantly decreased, suggesting a lower prevalence of smoking among the pathologists
12 compared with the general population of England and Wales. However, the finding of a possibly
13 increased risk of LHP cancers should be analyzed further by selecting a more appropriate
14 reference population (another professional group without exposure to formaldehyde) or by
15 utilizing internal comparisons.
16 Walrath and Fraumeni (1983) conducted a proportionate mortality study of all embalmers
17 and funeral directors licensed in the state of New York between 1902 and 1980 who were known
18 to have died between 1925 and 1980. While the investigators called their study a proportionate
19 mortality ratio (PMR) study, the methods described the comparison population with which
20 cancer deaths as the general population and not other deaths in the study population. The authors
21 computed their expected number of cause-specific deaths based on the US male population
22 stratified by 5-year age and calendar time periods. As such, their PMR meets the modern
23 definition of the more methodologically appropriate standardized mortality ration (SMR). The
24 authors did also compute what they called the proportionate cancer mortality ratio (PCMR)
25 which is not standardized and may overestimate effects. The investigators requested death
26 certificates for 1,678 persons but received only 1,263 (75%). The investigators restricted their
27 analysis to 1,132 males. The distribution of the causes of death was compared with the age-,
28 race-, and calendar-year-specific proportions of deaths for each cause among the male U.S.
29 population. Duration of exposure was approximated by time since first license. While the
30 methodology could not be applied in all calculations because of data gaps, excess risks were
31 found for lymphatic and hematopoietic cancers, with a SMR of 1.2 (observed 25), and for
32 leukemia, with a SMR of 1.32 and proportionate cancer mortality ratio (PCMR) of 1.19
33 (12 observed). The SMRs were not affected when the estimates were stratified by latency
34 (<35 years or 35 years since first license) or by age at first license. Because the cause of death
35 could not be determined for nearly 25% of the study group, the risk estimates could be
This document is a draft for review purposes only and does not constitute Agency policy.
4-138 DRAFT—DO NOT CITE OR QUOTE
-------
1 underestimated. When the mortality from leukemia was further examined, the investigators
2 identified six of the 12 leukemias as myeloid and five as multiple myeloma. Walrath and
3 Fraumeni (1983) reported that 4.1 myeloid leukemias were expected and therefore a true SMR
4 for myeloid leukemia of can be calculated as SMR = 1.46 (0.21-2.72).
5 Using what was actually the standardized mortality ratio methods (although referring to it
6 as the proportionate mortality method), Walrath and Fraumeni (1984) studied 1,007 deceased
7 white male embalmers, members of the California Bureau of Funeral Directing and Embalming,
8 whose deaths occurred between 1925 and 1980. The decedents had to have been licensed to
9 practice between 1916 and 1978. For lymphatic and hematopoietic cancer, the PMR (actually an
10 SMR) was 1.22 (19 deaths observed). For leukemia alone, the PMR was 1.75 and significant
11 (12 deaths observed, p < 0.05). Among embalmers licensed for 20 years or longer, the risk of
12 leukemia increased and was also significant (PMR (actually an SMR) 2.21; 8 observed;
13 /?<0.05).
14 Levine et al. (1984) conducted a cohort mortality study of 1,477 male Ontario
15 undertakers first licensed between 1928 and 1957 and followed until the end of 1977. Out of
16 359 subjects who had died, there were 8 deaths from lymphatic and hematopoietic cancers
17 compared with 6.5 expected. Additionally, there were 4 deaths from leukemia versus
18 2.5 expected. Because death rates were not available for Ontario before 1950, person-years and
19 deaths before 1950 could not be counted. No actual exposure estimates are available for these
20 undertakers.
21 Stroup et al. (1986) conducted an historic cohort mortality study of 2,317 men who were
22 members of the American Association of Anatomists between 1888 and 1969. The investigators
23 derived SMRs from the U.S. white male population and used members of the American
24 Psychiatric Association (APA) as a comparison group. Vital status was ascertained between
25 1925 and 1979. Women were excluded from analysis because of the small numbers. Only
26 738 deaths were observed versus 1,133.9 expected, based on U.S. death rates (SMR 0.65),
27 possibly indicating a sizable HWE. However, a slight increase in the risk of lymphatic and
28 hematopoietic cancers (SMR 1.2; 18 observed) and the risk of leukemia (SMR 1.5; 10 observed)
29 was evident. A significant increase in the risk of brain cancer (SMR 2.7; 10 observed; p < 0.05)
30 was also reported. When the leukemia analysis was restricted to the myeloid type, the SMR
31 increased to 8.8, based on five deaths (p < 0.05). The analysis using the APA group was
32 restricted to deaths that occurred between 1900 and 1969. This restriction removed five
33 leukemia deaths and person-years from the analysis because they likely died after 1969. Because
34 of this, there were only 3 leukemia deaths versus 3.6 expected, based on APA death rates. The
35 investigators concluded that the etiological agent had not been definitively identified, mentioning
This document is a draft for review purposes only and does not constitute Agency policy.
4-139 DRAFT—DO NOT CITE OR QUOTE
-------
1 that a wide range of solvents, stains, and preservatives, including formaldehyde, are used to
2 prepare biological specimens.
3 Logue et al. (1986) conducted a cohort study of male radiologists and pathologists
4 registered with the Radiation Registry of Physicians and the College of American Pathologists
5 (CAP) between 1962 and 1977. Although the main focus was on determining mortality in
6 radiologists from exposure to ionizing radiation, mortality was also ascertained for pathologists
7 alone. To derive SMRs, expected deaths were the sum of the products of person-years times
8 death rates for both cohorts during the follow-up period in white males only. However, there
9 were no exposure measurements, and the SMRs were not adjusted for calendar time. Of 5,585
10 members of the CAP, 496 had died by December 31, 1977. Although the SMR was 0.48 for
11 pathologists for cancer of the lymphatic and hematopoietic system, for the more specific
12 category of leukemia and aleukemia the SMR was 1.06 (neither was significant). For
13 radiologists, the SMRs were 0.78 and 1.55, respectively, also not significant. Cause of death
14 could not be determined for 8% of the deaths. Although age-adjusted rates for leukemia were
15 also calculated for each cohort, they were only used for comparison between the two separate
16 professional groups.
17 Hayes et al. (1990) conducted a cohort study of 3,649 deceased white and 397 deceased
18 nonwhite U.S. male embalmers and funeral directors who had died between 1975 and 1985,
19 using records from local licensing boards, state funeral directors' associations in 32 states and
20 the District of Columbia, the National Funeral Directors' Association, and state offices of vital
21 statistics (n = 894). Expected deaths by cause were derived from 5-year age- and calendar-year-
22 specific proportions of deaths among appropriate race groups from the U.S. population. No
23 measured exposure data were available. A PCMR would be derived by excluding noncancer
24 causes of death. Statistically significant excesses in hematopoietic and lymphatic cancers were
25 found in white (PMR (actually SMR)1.31 [95% CI: 1.06-1.59]; 100 observed) and nonwhite
26 (PMR (actually SMR) 2.41 [95% CI: 1.35-3.97]; 15 observed) embalmers and funeral directors.
27 The combined PMR (actually SMR) was 1.39 (95% CI: 1.15-1.63). The excess risk for all men
28 was higher for myeloid leukemia (ML) (PMR (actually SMR) 1.57 [95% CI: 1.01-2.34]; 24
29 observed) and for other unspecified leukemias (PMR (actually SMR) 2.28 [95% CI: 1.39-3.52];
30 20 observed) in white males.
31 Matanoski (1991) conducted a study of 6,111 male pathologists for NIOSH. Members of
32 the cohorts were part of an earlier unpublished study. Twenty-nine thousand psychiatrists were
33 used as a comparison group. Both samples were selected from the membership rolls of
34 professional associations. A total of 3,787 pathologists died between 1940 and 1978. Women
35 were excluded from the analysis. Of the population of psychiatrists, 4,788 died by 1980. U.S.
This document is a draft for review purposes only and does not constitute Agency policy.
4-140 DRAFT—DO NOT CITE OR QUOTE
-------
1 age- and calendar-time-specific death rates from 1925 were used to develop SMRs. Separate
2 SMRs were based on psychiatrists' death rates. The risk of hematopoietic cancer (excluding
3 Hodgkin's disease) was elevated (SMR 1.25; 57 observed) based on U.S. white males. For
4 leukemia, the SMR was 1.35 (31 observed). The SMR for leukemia among psychiatrists was
5 0.83 (35 observed). Compared with leukemia in psychiatrists, the SMR for pathologists was
6 1.68 (95% CI: 1.14-2.38). The SMR for other lymphatic cancers was 1.53 (16 observed) and for
7 LHP cancer 1.22 (64 observed). Comparing the pathologists' death rates to those of psychiatrists
8 could be thought to have greater validity than if death rates for the U.S. population as a whole
9 had been used, because of shared socioeconomic circumstances and access to medical care
10 between the two professional groups. Differences in access to health care might have been
11 greater for subjects in the earlier part of the study, because improved diagnosis and medical care
12 for LHP cancers became more broadly available later in the study period. By using SMRs based
13 on U.S. death rates, which include those who do not have adequate access to medical care, the
14 difference between expected and observed deaths would be reduced. This is less likely to occur
15 when one professional group is compared with another professional group, assuming
16 psychiatrists and pathologists have equal access to care.
17 Hauptmann et al. (2009) conducted a nested case-control study of lymphohematopoietic,
18 brain and nasopharyngeal cancers that included embalmers and funeral directors from previous
19 cohort mortality studies (Walrath and Fraumeni 1983, 1984; Hayes et al. 1990). Death
20 certificates for 6,808 embalmers and funeral directors who had died between January 1, 1960 and
21 January 1, 1986 were coded for underlying and contributory causes of death. There were 168
22 deaths attributed to lymphohematopoietic cancers, with 99 of lymphoid origin and 48 of
23 nonlymphoid origin, including 34 myeloid leukemias. Cases were matched to control subjects
24 randomly selected from cohort members, died of other causes, excluding cancers of the buccal
25 cavity and pharynx, respiratory system, and eye, brain and other parts of the nervous system.
26 Controls (N= 265) were matched on data source, sex, and dates of birth and death.
27 Extensive interviews with the next of kin and coworkers of the cases and controls
28 provided detailed information on the funeral homes and work practices of the study subjects.
29 The authors noted that since the funeral industry is often a family business, these study subjects'
30 next of kin were believed to be unusually knowledgeable of funeral home work practices. The
31 work history component of these interviews provided data on the frequency and duration of
32 embalmings for jobs held at least five years, as well as information on ventilation of the premises
33 and the frequency of spills. These data were linked to data from an exposure-reconstruction
34 experiment which sought to replicate standard funeral home practices while measuring exposures
35 to formaldehyde. These exposure data were not specific to any of the workplaces, but do provide
This document is a draft for review purposes only and does not constitute Agency policy.
4-141 DRAFT—DO NOT CITE OR QUOTE
-------
1 a general idea of exposures during embalming. Specific work practices, facilities, and
2 ventilation systems may vary widely between funeral homes potentially impacting actual
3 exposure levels for any location (Stewart et al., 1992). An exposure model was constructed to
4 retrospectively estimate study subjects' exposure. Multiple exposure metrics were estimated,
5 including lifetime 8-hour time weighted average (TWA), cumulative, and peak exposures to
6 formaldehyde.
7 Odds ratios were estimated for all lymphohematopoietic cancers combined, cancers of
8 lymphoid as well as nonlymphoid origin, and specifically for myeloid leukemia using the
9 following exposure metrics: ever embalming, number of embalmings, number of years of
10 working in embalming, and four quantitative estimates of exposure to formaldehyde (see
11 Table 4-5). Exposure to formaldehyde as estimated as "ever embalming" was not associated
12 with an increased risk for all lymphohematopoietic cancers combined (OR = 1.4, 95% CI:
13 0.8-2.6). The ORs for lymphohematopoietic cancers of nonlymphoid origin which includes
14 Hodgkin disease as well as other specific subtype were generally elevated but there did not
15 appear to be any clear patterns of association. However there was an increased risk for
16 lymphohematopoietic cancers of nonlymphoid origin (OR = 3.0, 95% CI: 1.0-9.5). An
17 exposure-response trend showed that more years of working in embalming was associated with
18 an increase in risk of lymphohematopoietic cancers of nonlymphoid origin (p < 0.05); the other
19 exposure metrics did not demonstrate significant exposure-response relationships. Statistically
20 significant increases were observed for lymphohematopoietic cancers of nonlymphoid origin for
21 the highest exposure category for durations of working in embalming jobs (OR = 3.7, 95%
22 CI: 1.1-12.2), for the number of embalmings (OR = 3.9, 95% CI: 1.2-12.8), for cumulative
23 exposure (OR = 4.0, 95% CI: 1.2-13.2), as well as for the cumulative mid-level exposures
24 (OR = 4.2, 95% CI: 1.2-14.3) and high-level exposures (OR = 3.4, 95% CI: 1.0-11.8) for 8-hour
25 TWA intensity of exposure, and for the highest exposure category for peak exposure (OR =3.8,
26 95% CI: 1.1-12.7).
27 The highest reported increases in risk for ever embalming were for myeloid leukemia
28 (OR = 11.2, 95% CI: 1.3-95.6). Duration of employment in jobs with embalming demonstrated
29 an exposure-response relationship with increased risk of myeloid leukemia (p = 0.02). The
30 number of embalming was also significantly associated with increased risk of myeloid leukemia
31 in the middle (OR = 12.7, 95%: 1.4-116.7) and highest exposure levels (OR = 12.7, 95% CI: 1.6-
32 119.7). Cumulative exposure was associated with increased risk of myeloid leukemia, with the
33 highest category of exposure showing OR = 13.2 (95% CI: 1.5-115.4).
34 All three categories of average formaldehyde intensity while embalming were very
35 strongly and significantly associated with the risk of myeloid leukemia mortality (OR = 11.1,
This document is a draft for review purposes only and does not constitute Agency policy.
4-142 DRAFT—DO NOT CITE OR QUOTE
-------
1 14.8, and 9.5), and the test for an exposure-response trend was of borderline statistical
2 significance (p = 0.058). The risk of myeloid leukemia mortality was also very strongly and
3 significantly associated with 8-hour TWA exposure for mid-level exposures (OR = 13.6, 95%
4 CI: 1.5-125.8) and high-level exposures (OR = 12.0, 95% CI: 1.3-107.4), as well as for peak
5 exposure in two of three categories (0-7 ppm OR = 15.2, 95% CI: 1.6-141.6; 7-9.3 ppm
6 OR =8.0, 95% CI: 0.9-74.0; and >9.3 ppm OR = 13.0, 95% CI: 1.4-116.9). For peak exposures,
7 there was also a statistically significant exposure-response tend (p = 0.036).
8 The study by Hauptmann et al. (2009) stands out among the studies of embalmers and
9 professionals in the funeral industry based on the strength of the quantitative exposure data and
10 the demonstration of exposure-response relationships which provide causal evidence of an
11 association between formaldehyde exposure and increased risk of myeloid leukemia. The results
12 also show an association between the broader categories of lymphohematopoietic cancers of
13 nonlymphoid origin which includes myeloid leukemia. These results were internally consistent
14 and demonstrated statistically significant associations that were unlikely the result of chance. As
15 this nested case-control study was based on the cohorts of Hayes et al. (1990) and those of
16 Walrath and Fraumeni (1983, 1984), the potential for selection bias is considered to be low.
17 Further, the controls in Hauptmann et al. (2009) were carefully selected to avoid individuals who
18 died of any causes that were thought to even possibly be related to formaldehyde exposure.
19 Confounding is also unlikely to be an alternative explanation for the observed results as there
20 were clear and convincing exposure-responses and the magnitude of the effect estimates were
21 extremely large.
22
23 4.1.2.2.1.2. Industry worker cohort studies.
24 This section discusses updated industrial worker studies that show associations between
25 LHP cancer and formaldehyde. The studies by Marsh et al. (1994), Blair et al. (1986), and
26 Acheson et al. (1984) and the later update by Gardner et al. (1993) provide estimates of exposure
27 to formaldehyde. The remaining studies generally rely either on duration of exposure (number
28 of years in the job) as a surrogate (Pinkerton et al., 2004) or provide no exposure assessment.
29 Blair et al. (1986) reported on 4,396 deaths from all causes in the 10 formaldehyde-
30 associated factories that made up the NCI cohort of 26,561 workers employed before January 1,
31 1966. There was little evidence of an association with LHP system cancer (SMR 0.91; 56
32 observed) in exposed white men, who dominated the cohort. Marsh et al. (1994), in an early
33 study of the Wallingford plant, which is also part of the Hauptmann et al. (2004, 2003) and Blair
34 et al. (1986) studies, found SMRs of 0.89 and 0.91, based on U.S. and county death rates,
This document is a draft for review purposes only and does not constitute Agency policy.
4-143 DRAFT—DO NOT CITE OR QUOTE
-------
1 respectively (25 observed deaths). The authors did not further discuss this cancer site until after
2 Hauptmann et al. (2003) was published.
3 Hauptmann et al. (2003) updated the cohort mortality study of Blair et al. (1986) that
4 consisted of predominantly the same (25,619) workers from 10 plants. The primary focus of this
5 analysis was cancer of the LHP system, including leukemia. The description and demographics
6 of the current study are the same as those reported by Blair et al. (1986) and Stewart et al.
7 (1986). In this update, follow-up was extended through December 31, 1994. The additional 15
8 years of follow-up increased the number of deaths from 4,349 to 8,486. However, in the course
9 of updating this cohort with follow-up through 2004, Beane Freeman et al. (2009) (including
10 Hauptmann) discovered that the matching of death certificates in Hauptmann et al. (2003) had
11 inexplicatively missed 1,006 deaths. Beane Freeman et al. (2009) reanalyzed the cohort with
12 follow-up through 1994 in order to replicate the analyses in Hauptmann et al. (2003) and
13 provided those results in supplementary tables. Nonetheless, the focus of the most recent re-
14 analysis of the NCI cohort was on follow-up completed through 2004.
15 While the results of the earlier Hauptmann et al. (2003) analysis are called into question
16 based on the missing deaths, the authors concluded that formaldehyde may cause leukemia,
17 particularly ML, in humans. However, because results from other studies were inconsistent, they
18 suggested caution in drawing definite conclusions as they could not provide a biological basis for
19 the significant excess risk of LHP. The authors pointed out several studies that indicate changes
20 that are consistent with chromosomal changes in formaldehyde-exposed persons, such as
21 increased frequencies of (MN (He et al., 1998; Kitaeva et al., 1996; Suruda et al., 1993), sister
22 chromatid exchanges (SCEs) (Shaham et al., 2002, 1997; Yager et al., 1986), chromosomal
23 aberrations (CAs) (He et al., 1998; Bauchinger and Schmid, 1985), and DNA-protein cross-links
24 (DPXs) (Shaham et al., 1997, 1996) in peripheral lymphocytes of humans exposed to
25 formaldehyde.
26 In the latest update to the large NCI cohort, Beane Freeman et al. (2009) extended the
27 follow-up through 2004. This cohort is based on workers from 10 manufacturers of
28 formaldehyde, formaldehyde resins, molding compounds, plastic products, film or plywood who
29 were first employed prior to 1966. The median follow-up time for workers was 42 years,
30 representing 998,106 person-years of follow-up among 25,619 workers (Beane Freeman et al.,
31 2009). Exposure to formaldehyde was estimated for each individual job category from work
32 histories, with calendar-time and plant-specific estimates based on assessments of job titles and
33 tasks associated with those jobs using plant visits by industrial hygienists and monitoring data
34 (Blair et al. 1986; Stewart et al. 1986; Blair and Stewart 1990). Exposures were categorized by
35 peak exposure, average intensity of exposure and cumulative exposure. Peak exposure
This document is a draft for review purposes only and does not constitute Agency policy.
4-144 DRAFT—DO NOT CITE OR QUOTE
-------
1 categories were defined as nonexposed, low (0.1-1.9 ppm), medium (2.0-3.9 ppm), and high
2 (4.0 ppm or greater). Average intensity categories of exposure were defined as nonexposed, low
3 (0.1-0.4 ppm), medium (0.5 to <0.9 ppm), and high (>1.0 ppm). Cumulative exposure was
4 defined as nonexposed, low (0.1-1.4 ppm-years), medium (1.5-5.4 ppm-years), and high (>5.5
5 ppm-years). Duration of exposure was defined as 0, 0.1-4.9 years, 5.0-14.9 years, and >15
6 years. The presence of formaldehyde-containing particulates and other potential chemical
7 coexposures in the plants were indentified. Information on potential formaldehyde exposures
8 after 1980 was unavailable.
9 Among those workers who were exposed, the median estimated time-weighted average
10 exposure to formaldehyde was 0.3 ppm (Range: 0.01-4.3 ppm). There were 4359 workers (17%)
11 who were classified as never exposed. Standardized mortality ratios (SMRs) were calculated
12 using sex-, race, age-, and calendar-year-specific U.S. mortality rates. Relative risks based on
13 internal comparisons within the cohort were estimated using Poisson regression, controlling for
14 calendar year, age (5-year categories), sex, race, and pay category.
15 A total of 319 deaths from all lymphohematopoietic cancers were identified. Among
16 them, there were 286 ever-exposed and 33 never-exposed workers. SMR analyses based on
17 external comparison of populations showed that mortality from lymphohematopoietic cancers
18 was not elevated in either the ever-exposed (SMR = 0.94, 95% CI: 0.84-1.06) or never-exposed
19 workers (SMR = 0.86, 95% CI: 0.61-1.21). The ratio of the exposed to unexposed SMR was
20 1.09 (95% CI: 0.76-1.57)5. Table 4-6 shows the ratios of the SMRs in the exposed workers to
21 the SMRs in the unexposed workers. An elevated risk for Hodgkin lymphoma among exposed
22 workers was observed (SMR = 1.42, 95% CI: 0.96-2.10) which was twice the SMR for the
23 unexposed workers but not significantly different from the null. The SMR for lymphatic
24 leukemia was only 1.15 among exposed workers but this was more than four times higher than
25 the SMR for lymphatic leukemia among unexposed worker although not significantly so.
26 Mortality from other subtypes of lymphohematopoietic cancers among the exposed workers did
27 not show increased mortality rates compared with the U.S. population. It is important to note
28 that workers are generally healthier than the general population and, as such, are expected to
29 have lower baseline risk for cancer and mortality. This healthy worker effect effectively biases
30 many external rate and risk comparisons to yield SMR values below unity. Internal analyses can
31 provide a more appropriate comparison, as exposed and unexposed workers are likely to have
32 background risks of cancer and mortality that are more similar.
33
34
1 Variance(SMRExpoSed/SMRUnexposed) ~ [l/ObservedExpoSed + l/ObservedUnexpoSed]
This document is a draft for review purposes only and does not constitute Agency policy.
4-145 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-6. Comparison of SMRs in the exposed workers to the SMRs in the
unexposed workers from NCI cohort reported by Beane Freeman et al. 2009.
Cause of Death
All LHP
NHL
Hodgkin Disease
Multiple Myeloma
All Leukemia
Lymphatic Leukemia
Myeloid Leukemia
Nonexposed
N
33
12
2
11
7
1
4
SMR
0.86
0.86
0.7
1.78
0.48
0.26
0.65
Exposed
N
286
94
25
48
116
36
44
SMR
0.94
0.85
1.42
0.94
1.02
1.15
0.9
Ratio of
SMRs
1.09
0.99
2.03
0.53
2.13
4.42
1.38
95% CIa
0.76-1.57
0.54-1.80
0.48-8.56
0.27-1.02
0.99-4.56
0.61-32.26
0.50-3.85
4
5 aVariance(SMRExposed/SMRUnexposed) ~ [l/ObservedExposed + l/ObservedUnexposed].
6
7 Internal analyses of exposed workers indicated that peak exposures in the highest
8 exposure category were associated with a significant increase in all lymphohematopoietic deaths
9 comparing death rates among workers with peaks of >4 ppm to those with >0 to 2.0 ppm
10 (RR = 1.37, 95% CI: 1.03-1.81). Across the three categories of peak exposure (i.e., exposure
11 >0 ppm), there was also a statistically significant exposure-response trend (p = 0.02). The
12 exposure-response trend including the never exposed workers was also statistically significant
13 (p = 0.04). No association was observed for all lymphohematopoietic cancers for average
14 intensity or cumulative exposure.
15 Among the specific subtypes of lymphohematopoietic cancer mortality, Hodgkin
16 lymphoma and multiple myeloma were both shown to be at increased risk associated with peak
17 exposure concentrations. Peak exposures in the highest exposure category were associated with
18 a significant increase in Hodgkin lymphoma deaths comparing death rates among workers with
19 peaks of > 4 ppm to those with > 0 to 2.0 ppm (RR = 3.96, 95% CI: 1.31-12.02). Across the
20 three categories of peak exposure, there was a statistically significant exposure-response trend
21 (p = 0.01). The exposure-response trend including the never-exposed workers was also
22 statistically significant (p = 0.004).
23 Peak exposures in the highest exposure category were associated with a significant
24 increase in multiple myeloma deaths comparing death rates among workers with peaks of > 4
25 ppm to those with > 0 to 2.0 ppm (RR = 2.04, 95% CI: 1.01-4.12). Across the three categories of
26 peak exposure, there was some evidence of an exposure-response trend (p = 0.08); however,
This document is a draft for review purposes only and does not constitute Agency policy.
4-146 DRAFT—DO NOT CITE OR QUOTE
-------
1 there was no evidence of an exposure-response trend including the never-exposed workers. The
2 association of multiple myeloma with formaldehyde exposure was also shown throughout the
3 cohort experience (see Figure 4-3). Regarding leukemia mortality, there was indication of a
4 modest exposure-response trend of peak exposure concentrations (p = 0.12); likewise for
5 myeloid leukemia specifically (p = 0.13). At the highest exposure category of peak exposure
6 (peaks > 4 ppm vs. > 0 to 2.0 ppm), the RR = 1.42 (95% CI: 0.92-2.18) for leukemia and
7 RR = 1.78 (95% CI: 0.87-3.64) for myeloid leukemia.
8 Beane Freeman et al. (2009) further showed plots presenting the RR from the internal
9 analyses for each endpoint and for each year of follow-up. The association of Hodgkin
10 lymphoma with formaldehyde exposure is not only seen for the complete 2004 follow-up, but
11 throughout the cohort experience (see Figure 4-3). These plots show that during the 1970's, the
12 relative risk of Hodgkin lymphoma with peak exposure was greater than 10 and diminished to
13 approximately RR = 8 in the 1980's and remained at about RR = 4 through the end of follow-up
14 in 2004. Similarly, the relative risks for myeloid leukemia were greater than 10 in the early
15 1970's and diminished with extended follow-up. Beane Freeman et al. (2009) stated that "this
16 pattern could reflect the increased precision of the relative risk estimates with accrual of
17 additional person-years and myeloid leukemia or could reflect a relatively short induction-
18 incubation time for myeloid leukemia because by time since first exposure and first high peak
19 both indicate highest risks within the first 25 years."
20 The exposure metric based on the average intensity of formaldehyde exposure was
21 associated with increased risks of Hodgkin lymphoma mortality but not with other specific
22 subtypes of lymphohematopoietic cancer mortality. Average exposures in the highest exposure
23 category were associated with an increase in Hodgkin lymphoma deaths (RR = 2.48, 95% CI:
24 0.84-7.32), comparing death rates among workers with peaks of > 4 ppm to those with > 0 to 2.0
25 ppm. Across the three categories of average intensity, there was a statistically significant
26 exposure-response trend (p = 0.05). The exposure-response trend including the never exposed
27 workers was also statistically significant (p = 0.03). As with peak exposure, there is a consistent
28 trend across the years of follow-up for RR for myeloid leukemia and multiple myeloma of the
29 mid and high exposed individuals be elevated, although only nearing significance at a few points
30 (see Figure 4-4).
31
This document is a draft for review purposes only and does not constitute Agency policy.
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
• : .M
i: ill
•I'1]!
:,.'•
•in.
NHL
IS'C
•: :-,'
i .:•,
J '
- = -
""*•••"*•' WW
ton ^
ton
L".
MYL
Kexvcnskr. l=r
* Hijr,
acco 1963
yn,nr of nn3 nF
Figure 4-3. Association between peak formaldehyde exposure and the risk of
lymphohematopoietic malignancy.
Relative risks for medium-peak (2.0 to <4.0 ppm) and high-peak (>4.0 ppm)
formaldehyde exposure categories compared with the low exposed category (>0
to <2.0 ppm) and P values for trend tests among the exposed person-years for
lymphohematopoietic malignancies are shown by year of end of follow-up, 1965-
2004. Values plotted at 0.1 represent RR = 0 due to no cases in the exposure
category values plotted at 20 represent RR = infinity due to no cases in the
referent category. The small graphs above the relative risk plots represent the
exposure-response trend P values based on two-sided likelihood ratio tests (1 df)
of zero slope for continuous formaldehyde exposure among exposed person-years
only. The points represent the relative risk estimates based on the cumulative
number of cases and person-years accrued from the start of the study to that point
in time and for 2004 are equivalent to the relative risk estimates presented in
Table 2 (Beane Freeman et al., 2009). HLP = lymphohematopoietic
malignancies, NHL = non-Hodgkin lymphoma, HDG = Hodgkin lymphoma,
MM = multiple myeloma, LEU = leukemia, LYL = lymphatic leukemia,
MYL = myeloid leukemia, RR = relative risk.
Source: Beane Freeman et al. (2009)
This document is a draft for review purposes only and does not constitute Agency policy.
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
1«0 19TO 1980 1990 H»0 1980 1flJO 1980 1990 2000 1«0 1970 19TO 1990 2000
Calendar year of end of loHew-up
Figure 4-4. Association between average intensity of formaldehyde exposure
and the risk of lymphohematopoietic malignancy.
Relative risks for medium (0.5-0.9 ppm) and high ( > 1.0 ppm) average-intensity
formaldehyde exposure categories compared with the low exposed category
(0.1-0.4 ppm) and P values for trend tests among the exposed person-years for
lymphohematopoietic malignancies by year of end of follow-up, 1965-2004.
Values plotted at 0.1 represent RR •• = due to no cases in the exposure category.
The small graphs above the relative risk plots represent the exposure—response
trend P values based on two-sided likelihood ratio tests (1 df) of zero slope for
continuous formaldehyde exposure among exposed person-years only. The
points represent the relative risk estimates based on the cumulative number of
cases and person-years accrued from the start of the study to that point in time and
for 2004 are equivalent to the relative risk estimates presented in Table 3, (Beane
Freeman et al., 2009). HLP = lymphohematopoietic malignancies, NHL = non-
Hodgkin lymphoma, HDG = Hodgkin lymphoma, MM = multiple myeloma, LEU
leukemia, LYL = lymphatic leukemia, MYL = myeloid leukemia; RR = relative
risk.
Source: Beane Freeman et al. (2009)
This document is a draft for review purposes only and does not constitute Agency policy.
4-149 DRAFT—DO NOT CITE OR QUOTE
-------
1 The third exposure metric that was assessed was cumulative exposure. There was no
2 clear trend toward increasing risk with increasing cumulative exposure to formaldehyde for all
3 lymphohematopoietic deaths. Among the subtype-specific causes of deaths, Hodgkin lymphoma
4 showed some evidence of an exposure-response association among the exposed workers
5 (p = 0.08) and among all workers (p = 0.06) which was consistent with the more clearly
6 demonstrated and stronger statistical findings of increased risk of Hodgkin lymphoma with
7 higher peak and average intensity. Leukemia also showed some evidence of an exposure-
8 response association among the exposed workers (p = 0.12) and among all workers (p = 0.08)
9 which was consistent with the findings of increased risk of leukemia with higher peak; however
10 there did not appear to be an association of with average intensity.
11 The strongest causal evidence of an association between exposure to formaldehyde and
12 lymphohematopoietic cancer mortality was demonstrated by the exposure-response gradient for
13 peak exposure and mortality from Hodgkin lymphoma and leukemia and the nonlinear increase
14 in RR for multiple myeloma. The plots showing a consistent increased in the relative risk over
15 time are further supportive of a causal association. The results for Hodgkin lymphoma are
16 supported by the demonstration of an exposure-response relationship for increased average
17 intensity of formaldehyde exposures and a consistent pattern of increasing risk with increasing
18 cumulative exposure to formaldehyde. The results for leukemia are supported by the exposure-
19 response relationship for cumulative exposure to formaldehyde.
20 This well-conducted study (Beane Freeman et al., 2009) showed adverse effects of
21 exposure to formaldehyde based on extensive and laborious reconstruction of individuals'
22 exposures (Blair et al. 1986; Stewart et al. 1986; Blair and Stewart 1990). The internal analyses
23 reflect control for potential selection bias akin to the healthy worker effect, and the Poisson
24 regression results controlled for numerous potential confounders. Potential confounding was
25 evaluated for exposure to 11 concomitant occupational substances (ever/never), as well as
26 working as a chemist or lab technician (for several years). In a follow-up analysis, although
27 Beane Freeman et al. (2009) excluded 586 individuals with possible exposure to benzene, a
28 known leukemogen, and the results did not change the relative risk for myeloid or lymphatic
29 leukemia in the highest peak exposure category (RR = 1.77; 95% CI = 0.85 to 3.69 and
30 RR = 1.16; 95% CI = 0.54 to 2.48, respectively) or any other cancer (data not shown). Exposure
31 lags ranging from 2 to 20 years were considered to account for latency; all exposures were
32 subsequently calculated using a 2-year lag interval for the analyses of lymphohematopoietic
33 malignancies (Beane Freeman et al. 2009, Hauptmann et al. 2003) and a 15-year lag interval for
34 the analyses of solid cancers (Hauptmann et al. 2004). The authors did note that smoking
35 information was unavailable for most of the cohort. However, smoking was not considered to be
This document is a draft for review purposes only and does not constitute Agency policy.
4-150 DRAFT—DO NOT CITE OR QUOTE
-------
1 a source of confounding in internal analyses since analysis of a sample of workers revealed no
2 major differences in smoking prevalence by cumulative formaldehyde exposure. Further, an
3 earlier analysis of the association between formaldehyde exposure and lung cancer in the same
4 cohort (Blair et al., 1990b) identified a small subset of workers with information on smoking
5 from medical records and reported that the prevalence of smoking did not appear to be strongly
6 associated with exposure to formaldehyde.
7 Beane Freeman et al. (2009) concluded that the pattern of attenuating
8 lymphohematopoietic risk over time was "consistent with a causal association within the
9 relatively short induction-incubation periods, characteristic of leukemogenesis." To understand
10 the implication of this conclusion, it should be noted that if the etiologically-relevant window of
11 time for formaldehyde-induced lymphohematopoietic cancer mortality is relatively short, then
12 extending a follow-up of the cohort well beyond the upper bound of that window will not allow
13 for the additional inclusion of more formaldehyde-induced lymphohematopoietic cancer deaths.
14 Rather, it allows for the inclusion of unrelated lymphohematopoietic cancer mortality which
15 serves to dilute any true effect of formaldehyde exposure. The median follow-up time in the
16 Beane Freeman et al. (2009) was 42 years and the analyses reported that exposure lags between 2
17 and 25 years were tested for fit, with an 18-year lag having the best fit. Since exposures were
18 not considered to have continued beyond 1980 and the cohort mortality was complete through
19 2004, the mortality in the later years of follow-up was unlikely to be related to formaldehyde as
20 the deaths may have been outside of any causal exposure window. It should be noted that the
21 analyses present in Beane Freeman et al. (2009) were based on a 2-year lag, as there was not
22 strong enough support for assuming either a longer or a shorter lag.
23 Two sensitivity analyses evaluated the assumption of no formaldehyde exposure after
24 1980. If exposure was considered to continue at 1980 levels until age 65 years or death, risk
25 patterns for all lymphohematopoietic cancers were reported to be similar to those observed in the
26 primary analysis. If the cohort follow-up was censored two years after the last job for the
27 2,810 individuals who were still exposed in 1979 and alive two years later, the association for
28 myeloid leukemia with peak and average intensity of exposure was stronger than that observed in
29 the primary analyses. The investigators did not report the results of the sensitivity analyses for
30 each subtype but did report that the patterns for the other subtypes were unchanged.
This document is a draft for review purposes only and does not constitute Agency policy.
4-151 DRAFT—DO NOT CITE OR QUOTE
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers
Reference
Harrington and
Shannon (1975)
Harrington and Oakes
(1984)
Study design
Cohort mortality study of 2,079
pathologists and 12,944 laboratory
technicians from the Royal
College of Pathologists and the
Pathological Society of Great
Britain from 1955-1973. The
comparison population came from
national mortality data.
Cohort mortality study of 2,720
pathologists from the Royal
College of Pathologists and the
Pathological Society of Great
Britain from 1974-1980. Vital
status obtained from the census, a
national health registry, and other
sources. SMRs developed from
the English, Scottish, Irish, and
Welsh populations.
Exposure assessment
Presumed exposure to
formaldehyde tissue
fixative.
Presumed exposure to
formaldehyde tissue
fixative.
Results, statistical significance (number of observed deaths for cohort
study)
Patholosists
All cause mortality
Tec
LHP cancers
Hodgkin's disease
Leukemia
micians
All cause mortality
All
Leu
LHP cancers
Hodgkin's disease
Leukemia
causes
Men
Women
iemia
Men
Women
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
0.60
2.0
1.4
0.6
0.67
0.5
-
0.5
0.56
0.99
0.91
9.26
Other LHP cancers
Men
Women
SMR
SMR
0.53
-
NR
p
-------
ts
^ 8
1!
f^ S
3 ,
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
Reference
Hall etal. (1991)
Levineetal. (1984)
Stroup etal. (1986)
Study design
Cohort mortality study of 4,5 12
pathologists from the Royal
College of Pathologists and the
Pathological Society of Great
Britain from 1974-1987. Vital
status obtained from the census, a
national health registry, and other
sources. SMRs developed from
the English and Welsh
populations.
Cohort mortality study of 1,477
male Ontario undertakers first
licensed 1928-1957, followed
from 1950 to 1977. SMRs
developed from Ontario mortality
rates.
Cohort mortality study of 2,3 17
white male members of the
American Association of
Anatomists from 1888 to 1969
who died 1925-1979. SMRs
developed using U.S. population
mortality rates.
Exposure assessment
Presumed exposure to
formaldehyde tissue
fixative.
Presumed exposure to
formaldehyde tissue
fixative.
Presumed exposure to
formaldehyde tissue
fixative.
Results, statistical significance (number of observed deaths for cohort
study)
All cause mortality
Men
Women
Hodgkin's disease
All cancers
Leukemia
All LHP cancers
Leukemia
All cause mortality
All LHP cancers
Lymphosarcoma and
reticulosarcoma
Hodgkin's disease
Leukemia
Other lymphatic
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
0.43
0.65
1.21
1.44
1.52
1.24
1.60
0.65
1.2
0.7
—
1.5
2.0
(95% CI: 0.37-0.50)
(95% CI: 0.38-1.03)
(95% CI: 0.03-6.71)
(95% CI: 0.69-2.63)
(95% CI: 0.41-3.89)
NR
NR
(95% CI: 0.60-0.70)
(95% CI: 0.7-2.0)
(95% CI: 0.1-2.5)
—
(95% CI: 0.7-2.7)
(95% CI: 0.7-4.4)
(176)
(18)
(1)
(10)
(4)
(8)
(4)
(738)
(18)
(2)
(0)
(10)
(6)
fe
H
o
o
-------
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
Reference
Logueetal. (1986)
Matanoski(1991)
Beane Freeman et al.
(2009)
Previous reports:
Hauptmann et al.
(2003)
Blair etal. (1986)
Study design
Cohort mortality study of 5,585
pathologists who were members of
the College of American
Pathologists, 1962-1972,
followed for mortality through
1977. SMRs developed from U.S.
population mortality rates.
Cohort mortality study of 6, 1 1 1
male pathologists from
membership rolls of the American
Medical Association 1912-1950.
Mortality was followed through
1978. SMRs developed from U.S.
population white male mortality
rates.
Retrospective cohort mortality
study of 25,619 workers employed
at 10 formaldehyde plants in the
U.S. followed from either the
plant start-up or first employment
through 2004.
SMRs calculated using sex-, age-,
race-, and calendar-year-specific
U.S. mortality rates.
RRs estimated using Poisson
regression stratified by calendar
year, age, sex, and race; adjusted
for pay category.
Exposure assessment
Presumed exposure to
formaldehyde tissue
fixative.
Presumed exposure to
formaldehyde tissue
fixative.
Exposure estimates based
on job titles, tasks, visits
to plants by study
industrial hygienists, and
monitoring data through
1980. Peak exposure
defined as short-term
excursions exceeding the
8-hour TWA
formaldehyde intensity
and knowledge of job
tasks. Exposures to 11
other compounds were
Results, statistical significance (number of observed deaths for cohort
study)
LHP cancer other than
leukemia
Leukemia
All cancer
All LHP cancers
Lymphosarcoma and
reticulosarcoma
Hodgkin's disease
Leukemia
Other lymphatic
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
0.48
1.06
0.78
1.25
1.31
0.36
1.35
1.54
NR
NR
(95% CI: 0.71-0.85)
(95% CI: 0.95-1.62)
(95% CI: 0.66-2.35)
(95% CI: 0.04-1.31)
(95% CI: 0.92-1.92)
(95% CI: 0.82-2.63)
(NR)
(NR)
(508)
(57)
(H)
(2)
(31)
(13)
All LHP cancers
Exposed
Unexposed
Peak exposure (ppm)
0
0.1-1.9
2.0 to <4.0
4.0 or greater
SMR
SMR
RR
RR
RR
RR
0.94
0.86
1.07
1.00
1.17
1.37
(95% CI: 0.84-1.06)
(95% CI: 0.61-1.52)
(95% CI: 0.70-1.62)
Reference value
(95% CI: 0.86-1.59)
(95% CI: 1.03-1.81)
(286)
(33)
(33)
(103)
(75)
(108)
Trend p = 0.02
Average exposure (ppm)
& S n
5'
3 ,
fe
H
O
o
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
Reference
Beane Freeman et al.
(2009) (continued)
Study design
Exposure assessment
identified. Workers
contributed pre-exposure
person-time to
nonexposed category.
Poisson regression
models used a 2-year lag
to account for tumor
latency.
Results, statistical significance (number of observed deaths for cohort
study)
0
0.1-0.4
0.5to<1.0
1.0 or greater
RR
RR
RR
RR
0.99
1.00
1.29
1.07
(95% CI: 0.66-1. 48)
Reference value
(95% CI: 0.97-1. 73)
(95% CI: 0.78-1. 47)
(33)
(164)
(67)
(55)
Trend p > 0.50
Cumulative exposure(ppm-vears)
0
0.1-1.4
1.5 to 5.4
5.5 or greater
RR
RR
RR
RR
0.89
1.00
0.77
1.07
(95% CI: 0.59-1.34)
Reference value
(95% CI: 0.56-1.07)
(95% CI: 0.80-1.42)
(33)
(168)
(49)
(69)
Trend p = 0.25
Leukemia
Peak exposure (ppm)
0
0.1-1.9
2.0 to <4.0
4.0 or greater
RR
RR
RR
RR
0.59
1.00
0.98
1.42
(95% CI: 0.25-1.36)
Reference value
(95% CI: 0.60-1.62)
(95% CI: 0.92-2. 18)
(7)
(41)
(27)
(48)
Trend p = 0.012
Average exposure (ppm)
Oppm
0.1-0.4
0.5to<1.0
1.0 or greater
RR
RR
RR
RR
0.54
1.00
1.13
1.10
(95% CI: 0.24-1.22)
Reference value
(95% CI: 0.71-1.79)
(95% CI: 0.68-1.78)
(7)
(67)
(25)
(24)
s
"3
o
5'
s ,
fe
H
O
O
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
s
"3
o
5'
s ,
Reference
Beane Freeman et al.
(2009) (continued)
Study design
Exposure assessment
Results, statistical significance (number of observed deaths for cohort
study)
Trend p > 0.50
Cumulative exposure (ppm-vears)
0
0.1-1.4
1.5-5.4
5.5 or greater
RR
RR
RR
RR
0.53
1.00
0.96
1.11
(95% CI: 0.23-1.21)
Reference value
(95% CI: 0.60-1.56)
(95% CI: 0.70-1.74)
(7)
(63)
(24)
(29)
Trend p = 0.12
Hodgkin Lymphoma
Peak exposure (ppm)
0
0.1-1.9
2.0 to <4.0
4.0 or greater
RR
RR
RR
RR
0.67
1.00
3.30
3.96
(95% CI: 0.12-3. 60)
Reference value
(95% CI: 1.04-10.50)
(95% CI: 1.31-12.02)
(2)
(6)
(8)
(11)
Trend p = 0.01
Average exposure (ppm)
0
0.1-0.4
0.5to<1.0
1.0 or greater
RR
RR
RR
RR
0.46
1.00
3.62
2.48
(95% CI: 0.05-3. 93)
Reference value
(95% CI: 1.41-9.31)
(95% CI: 0.84-7.32)
(2)
(10)
(9)
(6)
Trend p = 0.05
Cumulative (ppm-vears)
0
0.1-1.4
RR
RR
0.42
1.00
(95% CI: 0.09-2.05)
Reference value
(2)
(14)
fe
H
O
O
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
s
"3
o
5'
s ,
Reference
Beane Freeman et al.
(2009) (continued)
Study design
Exposure assessment
Results, statistical significance (number of observed deaths for cohort
study)
1.5-5.4
5.5 or greater
RR
RR
1.71
1.30
(95% CI: 0.66-4.38)
(95% CI: 0.40-4. 19)
(7)
(4)
Trend p = 0.08
Myeloid Leukemia
Peak exposure (ppm)
0
0.1 to 1.9
2.0 to <4.0
4.0 or greater
RR
RR
RR
RR
0.82
1.00
1.30
1.78
(95% CI: 0.25-2.67)
Reference value
(95% CI: 0.58-2.92)
(95% CI: 0.87-3.64)
(4)
(14)
(11)
(19)
Trend p = 0.13
Average exposure (ppm)
0
0.1 to 0.4
0.5to<1.0
1.0 or greater
RR
RR
RR
RR
0.70
1.00
1.21
1.61
(95% CI: 0.23-2. 16)
Reference value
(95% CI: 0.56-2.62)
(95% CI: 0.76-3.39)
(4)
(24)
(9)
(11)
p = 0.43
Cumulative (ppm-vears)
0
0.1-1.4
1.5-5.4
5.5 or greater
RR
RR
RR
RR
0.61
1.00
0.82
1.02
(95% CI: 0.20-1.91)
Reference value
(95% CI: 0.36-1.83)
(95% CI: 0.48-2. 16)
(4)
(26)
(8)
(10)
Trend p > 0.50
fe
H
o
o
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
s
"3
o
5'
s ,
Reference
Hauptmann et al.
(2009)
Previous reports:
Hayes etal. (1990)
Walrath and Fraumeni
(1983)
Walrath and Fraumeni
(1984)
Related re-analvses:
Marsh et al. (2007a)
Marsh et al. (2007b)
Marsh and Youk
(2005)
Marsh etal. (1996)
Study design
Nested case-control study within
cohort mortality study of 6,808
A +1-. f 1 C\£f\ 4- 1 OO/T
deaths Irom 1960 to 1986.
Identified from registries of the
National Funeral Director
Association, licensing boards and
state funeral directors'
associations, NY State Bureau of
Funeral Directors and CA Funeral
Directors and Embalmers.
Odds ratios calculated using
unconditional logistic regression
Exposure assessment
Occupational history
obtained by interviews
with next of kin and
coworkers using detail
questionnaires.
Exposure was assessed
by linking questionnaire
responses to an exposure
assessment experiment.
Exposure levels (peak,
intensity and cumulative)
were assigned to each
individual using a
predictive model based
on the exposure data.
Results, statistical significance (number of observed deaths for cohort
study)
All LHP cancers
Embalming
Never
Ever
OR
OR
1.0
1.4
Reference value
(95% CI: 0.8-2.6)
(24)
(144)
Duration of working in iobs with embalming (years)
0
> 0 to 20
> 20 to 34
>34
OR
OR
OR
OR
1.0
0.8
1.5
1.8
Reference value
(95% CI: 0.4-1. 8)
(95% CI: 0.8-2.8)
(95% CI: 1.0-3.4)
(24)
(28)
(50)
(66)
Trend p = 0.058
Number of embalmings
0
> 0 to 1422
> 1422 to 3068
>3068
OR
OR
OR
OR
1.0
0.9
1.9
1.5
Reference value
(95% CI: 0.6-1. 8)
(95% CI: 1.0-3.6)
(95% CI: 0.8-2.9)
(24)
(29)
(62)
(53)
Trend p = 0.477
Cumulative exposure (ppm-hours)
0
> 0 to 4058
> 4058 to 9253
>9253
OR
OR
OR
OR
1.0
1.3
1.4
1.6
Reference value
(95% CI: 0.6-2.5)
(95% CI: 0.8-2.8)
(95% CI: 0.8-3.0)
(24)
(40)
(49)
(55)
Trend p = 0.422
oo
fe
H
o
o
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
s
"3
o
5'
s ,
Reference
Hauptmann et al.
(2009) (continued)
Study design
Exposure assessment
Results, statistical significance (number of observed deaths for cohort
study)
Average formaldehyde intensity while embalming (ppm)
0
>0tol.4
> 1.4 to 1.9
>1.9
OR
OR
OR
OR
1.0
1.6
1.4
1.3
Reference value
(95% CI: 0.9-3.2)
(95% CI: 0.7-2.7)
(95% CI: 0.7-2.5)
(24)
(53)
(47)
(44)
Trend p = 0.591
Time-weighted average exposure over 8 hours (ppm)
0
>0 to 0.10
> 0.1 to 0.18
>0.18
OR
OR
OR
OR
1.0
1.3
1.6
1.4
Reference value
(95% CI: 0.7-2.6)
(95% CI: 0.8-3.1)
(95% CI: 0.7-2.8)
(24)
(47)
(52)
(45)
Trend p = 0.635
Peak formaldehyde exposure (ppm)
0
> 0 to 7.0
> 7.0 to 9.3
>9.3
OR
OR
OR
OR
1.0
1.6
1.6
1.2
Reference value
(95% CI: 0.8-3.2)
(95% CI: 0.9-3.1)
(95% CI: 0.6-2.3)
(24)
(48)
(55)
(41)
Trend p = 0.555
Myeloid leukemia
Embalming
Never
Ever
OR
OR
1.0
11.2
Reference value
(95% CI: 1.3-95.6)
(1)
(33)
J^.
\D
fe
H
o
o
-------
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
Reference
Hauptmann et al.
(2009) (continued)
Study design
Exposure assessment
Results, statistical significance (number of observed deaths for cohort
study)
Duration of working in iobs with embalming (years)
0
> 0 to 20
> 20 to 34
>34
OR
OR
OR
OR
1.0
5.0
12.9
13.6
Reference value
(95% CI: 0.5-51.6)
(95% CI: 1.4-117.1)
(95% CI: 1.6-119.7)
(1)
(6)
(13)
(14)
Trend p = 0.02
Number of embalmings
0
> 0 to 1422
> 1422 to 3068
>3068
OR
OR
OR
OR
1.0
7.6
12.7
12.7
Reference value
(95% CI: 0.8-73.5)
(95% CI: 1.4-116.7)
(95% CI: 1.4-112.8)
(1)
(7)
(12)
(14)
Trend p = 0.314
Cumulative exposure (ppm-hours)
0
> 0 to 4058
> 4058 to 9253
>9253
OR
OR
OR
OR
1.0
10.2
9.4
13.2
Reference value
(95% CI: 1.1-95.6)
(95% CI: 1.0-85.7)
(95% CI: 1.5-115.4)
(1)
(9)
(10)
(14)
Trend p = 0.192
Average formaldehyde intensity while embalming (ppm)
0
>0to 1.4
> 1.4 to 1.9
>1.9
OR
OR
OR
OR
1.0
11.1
14.8
9.5
Reference value
(95% CI: 1.2-106.3)
(95% CI: 1.6-136.9)
(95% CI: 1.1-86.0)
(1)
(10)
(13)
(10)
Trend p = 0.058
& s
II
s ,
a\
o
fe
H
O
O
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
s
"3
o
5'
s ,
Reference
Hauptmann et al.
(2009) (continued)
Wang et al. (2009)
Study design
Population-based case-control
study of incident cases of non-
Hodgkin lymphoma diagnoses
1996-2000.
Exposure assessment
Exposures classified
using job exposure
matrix based on
occupational and industry
data obtained from
personal interviews.
Results, statistical significance (number of observed deaths for cohort
study)
Time-weighted average exposure over 8 hours (ppm)
0
>0 to 0.10
> 0.1 to 0.18
>0.18
OR
OR
OR
OR
1.0
8.4
13.6
12.0
Reference value
(95% CI: 0.8-79.3)
(95% CI: 1.5-125.8)
(95% CI: 1.3-107.4)
(1)
(8)
(13)
(12)
Trend p = 0.396
Peak formaldehyde exposure (ppm)
0
> 0 to 7.0
> 7.0 to 9.3
>9.3
OR
OR
OR
OR
1.0
15.2
8.0
13.0
Reference value
(95% CI: 1.6-141.6)
(95% CI: 0.9-74.0)
(95% CI: 1.4-116.9)
(1)
(12)
(9)
(12)
Trend p = 0.036
Non
i-Hodgkin lymphoma
Formaldehyde
Never
Ever
Average exposure inte
OR
OR
nsitv
Never
Low
Medium-High
OR
OR
OR
1.0
1.3
1.0
1.4
1.2
Reference value
(95% CI: 1.0-1.7)
Reference value
(95% CI: 1.0-1.8)
(95% CI: 0.8-1.7)
(398)
(203)
(398)
(129)
(74)
Trend p = 0.21
o
o
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
s
"3
o
5'
s ,
Reference
Wang et al. (2009)
(continued)
Study design
Exposure assessment
Results, statistical significance (number of observed deaths for cohort
study)
Average exposure probability
Never
Low
Medium-High
OR
OR
OR
1.0
1.3
1.4
Reference value
(95% CI: 1.0-1.7)
(95% CI: 0.9-2.3)
(398)
(129)
(74)
Trend p > 0.50
Both average exposure intensity and average exposure probability
Low Intensity and
Low Probability
Med-High Intensity
and Low Probability
Med-High Intensity
and Med-High Prob.
Med-High Intensity
and Med-High Prob.
OR
OR
OR
OR
Leukemia
Peak exposure (ppm)
0
0.1-1.9
2.0 to <4.0
4.0 or greater
RR
RR
RR
RR
1.4
1.0
1.1
1.6
0.59
1.00
0.98
1.42
(95% CI: 1.1-1.9)
(95% CI: 0.7-1. 6)
(95% CI: 0.5-2.4)
(95% CI: 0.9-3.1)
(95% CI: 0.25-1.36)
Reference value
(95% CI: 0.60-1. 62)
(95% CI: 0.92 -2. 18)
(115)
(50)
(14)
(24)
(7)
(41)
(27)
(48)
Trend p = 0.012
Average exposure (ppm)
Oppm
0.1-0.4
RR
RR
0.54
1.00
(95% CI: 0.24-1. 22)
Reference value
(7)
(67)
a\
to
o
o
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
s
"3
o
5'
s ,
Reference
Wang et al. (2009)
(continued)
Study design
Exposure assessment
Results, statistical significance (number of observed deaths for cohort
study)
0.5to<1.0
1.0 or greater
RR
RR
1.13
1.10
(95% CI: 0.71-1.79)
(95% CI: 0.68-1. 78)
(25)
(24)
Trend p > 0.50
Cumulative exposure (ppm-vears)
0
0.1-1.4
1.5-5.4
5.5 or greater
RR
RR
RR
RR
0.53
1.00
0.96
1.11
(95% CI: 0.23-1.21)
Reference value
(95% CI: 0.60-1. 56)
(95% CI: 0.70-1. 74)
(7)
(63)
(24)
(29)
Trend p = 0.12
Hodgkin Lymphoma
Peak exposure (ppm)
0
0.1-1.9
2.0 to <4.0
4.0 or greater
RR
RR
RR
RR
0.67
1.00
3.30
3.96
(95% CI: 0.12-3. 60)
Reference value
(95% CI: 1.04-10.50)
(95% CI: 1.31-12.02)
(2)
(6)
(8)
(11)
Trend p = 0.01
Average exposure (ppm)
0
0.1-0.4
0.5to<1.0
1.0 or greater
RR
RR
RR
RR
0.46
1.00
3.62
2.48
(95% CI: 0.05-3.93)
Reference value
(95% CI: 1.41-9.31)
(95% CI: 0.84-7.32)
(2)
(10)
(9)
(6)
Trend p = 0.05
J^.
a\
fe
H
o
o
-------
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
Reference
Wang et al. (2009)
(continued)
Study design
Exposure assessment
Results, statistical significance (number of observed deaths for cohort
study)
Cumulative (ppm-vears)
0
0.1-1.4
1.5-5.4
5.5 or greater
RR
RR
RR
RR
0.42
1.00
1.71
1.30
(95% CI: 0.09-2.05)
Reference value
(95% CI: 0.66-4.38)
(95% CI: 0.40-4. 19)
(2)
(14)
(7)
(4)
Trend p = 0.08
Myeloid Leukemia
Peak exposure (ppm)
0
0.1 to 1.9
2.0 to <4.0
4.0 or greater
RR
RR
RR
RR
0.82
1.00
1.30
1.78
(95% CI: 0.25-2.67)
Reference value
(95% CI: 0.58-2.92)
(95% CI: 0.87-3.64)
(4)
(14)
(11)
(19)
Trend p = 0.13
Average exposure (ppm)
0
0.1 to 0.4
0.5to<1.0
1.0 or greater
RR
RR
RR
RR
0.70
1.00
1.21
1.61
(95% CI: 0.23 -2. 16)
Reference value
(95% CI: 0.56-2.62)
(95% CI: 0.76-3.39)
(4)
(24)
(9)
(11)
p = 0.43
Cumulative (ppm-vears)
0
0.1-1.4
1.5-5.4
5.5 or greater
RR
RR
RR
RR
0.61
1.00
0.82
1.02
(95% CI: 0.20-1. 91)
Reference value
(95% CI: 0.36-1.83)
(95% CI: 0.48-2. 16)
(4)
(26)
(8)
(10)
Trend p > 0.50
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
s ,
J^.
ON
o
o
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
Reference
Pinkerton et al. (2004)
Update of Stayner et
al. (1988)
Study design
Cohort mortality study of 11,098
workers in 3 garment plants
exposed >3 months after
formaldehyde was introduced.
Women comprised 81.7% of the
cohort. Vital status was followed
through 1998. SMRswere
calculated by using sex-, age-,
race-, and calendar-year-specific
U.S. mortality rates. Multiple
cause SMRs were derived from all
contributing causes from death
certificates.
Exposure assessment
Data for 549 randomly
selected employees in 5
departments in 1981 and
1984 used to estimate
overall exposure levels.
Levels presumed to be
0.09-0.20 ppm.
Results, statistical significance (number of observed deaths for cohort
study)
All LHP cancers
Lymphosarcoma and
reticulosarcoma
Hodgkin's disease
Other lymphatic
Leukemia
Mortality since 1960
Lymphocytic leukemia
ML
10+ years of
exposure
20+ years since 1st
exposure
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
0.97
0.85
0.55
0.97
1.09
0.60
1.44
2.19
1.91
Multiple cause leukemia
10+ years of
exposure and 20+
years since 1st
exposure
SMR
1.92
Multiple cause ML
20+ years since 1st
exposure
10+ years of
exposure and 20+
years since 1st
exposure
SMR
SMR
2.02
2.55
(95% CI: 0.74-1. 26)
(95% CI: 0.28-1. 99)
(95% CI: 0.07-1. 98)
(95% CI: 0.64-1. 40)
(95% CI: 0.70-1. 62)
(95% CI: 0.12-1. 75)
(95% CI: 0.80-2.37)
NS
^>0.05
(95% CI: 1.08-3.17)
(95% CI: 1.13-3.34)
(95% CI: 1.10-5.03)
(59)
(5)
(2)
(28)
(24)
(3)
(15)
(8)
(13)
(15)
(15)
(8)
s
"3
o
5'
s ,
fe
H
o
o
-------
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
Reference
Coggon et al. (2003)
Update of Gardner et
al. (1993)
Andjelkovich et al.
(1995)
Bertazzi et al. (1986)
Study design
Cohort mortality study of 14,014
men employed in 6 factories of the
chemical industry in Great Britain
from periods during which
formaldehyde was produced.
Cohort mortality followed from
1 94 1 through 2000 . SMRs based
on English and Welsh age- and
calendar-year-specific mortality
rates.
Cohort mortality study of 3,929
automotive industry iron foundry
workers exposed from 1960-1987
and followed through 1989.
SMRs calculated using sex-, age-,
race-, and calendar-year-specific
U.S. mortality rates.
Cohort mortality study of 1,330
male workers in an Italian resin
plant. Subjects were employed
any time between 1959 and 1980
for at least 30 days. Vital status
followed through 1986. SMRs
calculated using sex-, age-, race-,
and calendar-year-specific
national and local mortality rates.
Exposure assessment
Exposure assessment
based on data abstracted
from company records.
Jobs categorized as
background, low,
moderate, high, or
unknown levels.
Exposure assessment
based on review of work
histories by an industrial
hygienist.
Exposure assessment
based on reconstruction
of work history.
Exposure levels were
0.16to3.1ppm
Results, statistical significance (number of observed deaths for cohort
study)
Non-Hodgkin's lymphoma
Overall
High exposure
T PlllcPTTlifl
-Lit LUVtllllCl
Overall
High exposure
Multiple myeloma
Overall
High exposure
All LHP cancers
Leukemia
All LHP cancers
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
SMR
0.98
0.89
0.91
0.71
0.86
1.18
0.59
0.43
2.01
(95% CI: 0.67-1.39)
(95% CI: 0.41-1.70)
(95% CI: 0.62-1.29)
(95% CI: 0.31-1.39)
(95% CI: 0.48-1.41)
(95% CI: 0.48-2.44)
(95% CI: 0.23-1.21)
(95% CI: 0.05-1.57)
(31)
(9)
(31)
(8)
(15)
(7)
(7)
(2)
(5)
& s
II
s ,
Oi
Oi
fe
H
o
o
-------
o 1 a Tab\e 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
"""^ (V ~-.
& s R
111
f^ XJ '.\l
k ^ S
•<§ ° 5'
"a a ^2
O W^ ^J5
^ o4 ^
>!
J^.
ON
o
Reference
Edlingetal. (1987)
Dell and Teta (1995)
Walrath and Fraumeni
(1983)
Study design
Cohort mortality and incidence
study of 521 Swedish workers in
an abrasive production plant with
at least 5 years of employment
between 1955 and 1983, followed
through 1983.
Cohort mortality study of 5,932
male employees of a New Jersey
plastics manufacturing, research
and development facility that
produced phenol-formaldehyde
resins.
Cohort study of 1,132 white male
embalmers licensed to practice
between 1902 and 1980 in New
York who died between 1925 and
1980 identified from registration
files. Deaths were compared with
age-, race-, and calendar-year-
expected numbers of deaths from
the U.S. population.
Exposure assessment
Exposure level of
0.1-lmg/m3.
Examination of work
histories to identify jobs
where formaldehyde was
involved.
No direct measurements.
Presumed exposure to
formaldehyde tissue
fixative.
Results, statistical significance (number of observed deaths for cohort
study)
Lymphoma
Multiple myeloma
SMR
SMR
2.0
4.0
(95% CI: 0.2-7.2)
(95% CI: 0.5-14)
All LHP cancers
Hourly workers
Salaried workers
SMR
SMR
0.93
1.69
Leukemia
Hourly workers
Salaried workers
SMR
SMR
0.98
1.98
(2)
(2)
(2%)
(23)
(12)
(H)
All LHP cancers
SMRb
1.15
(21)
Lymphosarcoma and reticulosarcoma
SMRb
1.08
(4)
Hodgkin's disease
SMRb
1 0
(2)
Other lymphatic lymphoma
SMRb
1.18
(5)
Leukemia
£
H
o
0
SMRb
1.32
(10)
-------
ts
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
5 8
1!
f^ S
2 5'
S ,
Reference
Walrath and Fraumeni
(1984)
Hayes etal. (1990)
Study design
Cohort study of 1,007 white male
embalmers from the California
Bureau of Funeral Directing and
Embalming who died between
1925 and 1980. Deaths were
compared with age- and calendar-
year-expected numbers
of deaths from the U. S.
population.
Proportionate mortality cohort
study of 3,649 deceased white and
397 deceased nonwhite U.S. male
embalmers and funeral directors,
derived from licensing boards and
funeral director associations in the
32 states and the District of
Columbia. Occupation was
confirmed on death certificate.
Deaths were compared with age-
arid calendar-year-expected
numbers of deaths from the U.S.
population.
Exposure assessment
No direct measurements.
Presumed exposure to
formaldehyde tissue
fixative.
No direct measurements.
Presumed exposure to
formaldehyde tissue
fixative.
Results, statistical significance (number of observed deaths for cohort
study)
All LHP cancers
Lymphosarcoma and
reticulosarcoma
Hodgkin's disease
Other lymphatic
lymphoma
Leukemia
Licensed <20 years
Licensed >20 years
SMRb
SMRb
SMRb
SMRb
SMRb
SMRb
SMRb
1.22
0.97
-
1.33
1.75
1.24
2.21
^<0.05
^<0.05
(19)
(3)
(0)
(4)
(12)
(4)
(8)
All LHP cancers
Race
White
Nonwhite
SMRb
SMRb
SMRb
Occupation on death certificate
Embalmer
Funeral director
Other
Ase at death
<60
60-74
>75
SMRb
SMRb
SMRb
SMRb
SMRb
SMRb
1.39
1.31
2.41
1.23
1.56
1.30
1.35
1.72
1.16
(95% CI: 1.15-1.67)
(95% CI: 1.06-1.59)
(95% CI: 1.35-3.97)
(95% CI: 0.78-1.85)
(95% CI: 1.23-1.94)
(95% CI: 0.67-2.28)
(95% CI: 0.88-1.98)
(95% CI: 1.33-2.19)
(95% CI: 0.74-1. 74)
(115)
(100)
(15)
(23)
(78)
(12)
(26)
(66)
(23)
oo
fe
H
o
o
-------
Table 4-7. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers (continued)
Reference
Hayes etal. (1990)
(continued)
Blair etal. (1993)
Study design
Population-based case-control
study of 622 white men with LHP
cancers. Cancers selected from
Iowa and Minnesota cancer
surveillance networks diagnosed
between 10/80 and 9/82. 1,245
matched controls for living cases
selected by random digit dialing if
younger than age 65 and from
Medicare records if 65 or older.
Study focused on agricultural
exposures.
Exposure assessment
Personal interviews of
subjects or next of kin
included job histories,
agricultural exposures,
and chemical exposures.
Job titles used to create
job exposure matrix.
Industrial hygienist
estimated probability and
intensity of exposures to
large numbers of
substances.
Results, statistical significance (number of observed deaths for cohort
study)
Hodgkin's disease
Non-Hodgkin's
lymphoma
Lymphosarcoma and
reticulosarcoma
Multiple myeloma
Other lymphatic
lymphoma
Lymphatic leukemia
ML
Other leukemia
Non-Hodgkin's
lymphoma (formaldehyde
exposure)
Funeral service worker
SMRb
SMRb
SMRb
SMRb
SMRb
SMRb
SMRb
SMRb
ORa
ORa
0.72
1.26
1.12
1.37
1.35
0.74
1.57
2.28
1.2
2.1
(95% CI: 0.15-2. 10)
(95% CI: 0.87-1.76)
(95% CI: 0.58-1.96)
(95% CI: 0.84 -2. 12)
(95% CI: 0.85-2.01)
(95% CI: 0.29-1.53)
(95% CI: 1.01-2.34)
(95% CI: 1.39-3.52)
(95% CI: 0.9-1.7)
(95% CI: 0.5-7.9)
(3)
(34)
(12)
(20)
(22)
(7)
(24)
(20)
(6)
& s
II
5'
s ,
"Adjusted for age, state, smoking, family history of malignant proliferative disease, agricultural exposure to pesticides, use of dye, and direct/surrogate response
to interview.
bWalrath and Fraumeni (1983, 1984). These studies are referred to by the authors as proportionate mortality studies and report proportional mortality ratios
which are known to be potentially biased. However, review of the actual methods described clear shows that the expected numbers of cause-specific deaths
were based on a standardized general population and therefore the reported PMRs are more accurately called SMRs.
O
O
-------
1 Stayner et al. (1988) conducted a cohort study of 11,030 workers (82% female) followed
2 from 1955 or the beginning date of exposure through 1982 in three garment factories. Personnel
3 records from three garment manufacturing facilities, one in Pennsylvania and two in Georgia,
4 were used to assemble a cohort of workers who attained a minimum of 3 months of exposure
5 after the introduction of formaldehyde into these facilities. Formaldehyde resins were used to
6 treat fabrics, beginning in 1955 and 1959. Although formaldehyde levels were available on a
7 subset of the employees from monitoring data available from surveys completed in 1981 and
8 1984, they were not used in this analysis. Instead, the results were stratified by duration and
9 latency. SMRs were based on U.S. population mortality rates. Based on six cases, the SMRs for
10 leukemia were 2.43 and 3.81 among workers with 20 or more years since first exposure or at
11 least 10 years of exposure, respectively. In their conclusions, the authors suggested that,
12 although the numbers of deaths from LHP cancers were small, the risks were related to duration
13 and latency.
14 Pinkerton et al. (2004) updated the Stayner et al. (1988) study by adding 16 years of
15 follow-up. No new exposure information was added. The mean TWA exposure in 1981-1984
16 for the three plants was 0.15 ppm. No additional information regarding earlier industrial hygiene
17 data was available, although the authors stated that the levels of exposure to formaldehyde were
18 greater in the years before 1980. Stayner et al. (1988) cited independent studies of exposure
19 levels in similar garment factories in the 1960s that seemed to indicate that the formaldehyde
20 levels during that period ranged from 0.9 to 2.7 ppm (Blejer and Miller, 1966) in one garment
21 manufacturing area. Another report (Shipkovitz, 1966) of 10-minute personal exposure samples
22 indicated a range from 0.3 to 2.7 ppm in eight garment plants. In another study (Ahmad and
23 Whitson, 1973), the levels ranged from 2 to 10 ppm. Goldstein (1973) calculated that
24 concentrations in the cutting rooms of garment plants dropped from 10 ppm in 1968 to less than
25 2 ppm in 1973 because of an improvement in the resin treating process. The authors assumed
26 that exposure ceased in 1981 and 1983. This produced an underestimate of exposure based on
27 duration of employment for about 11% of the cohort who were still actively employed after those
28 dates. Stayner et al. (1988) speculated that the risks of cancer of the buccal cavity, leukemia, and
29 other LHP neoplasia may have been due to exposure to the highest potential formaldehyde levels
30 in the industry between 1955 and 1962, because the resin used to treat permanent press fabrics
31 still contained a relatively large amount of formaldehyde.
32 The SMRs were derived from age-, race-, and calendar-time-adjusted U.S. mortality
33 rates. The analysis was repeated using Georgia or Pennsylvania mortality rates. In addition to
34 the primary analysis of the underlying cause of death, the analysis used all causes listed on the
35 death certificates to evaluate multiple cause mortality. As a referent for this, the analysis relied
This document is a draft for review purposes only and does not constitute Agency policy.
4-170 DRAFT—DO NOT CITE OR QUOTE
-------
1 on multiple cause death rates available since 1960 from the National Death Index maintained by
2 the U.S. Centers for Disease Control and Prevention (CDC).
3 Altogether, 608 cancer deaths were observed (Pinkerton et al. 2004). The SMR for all
4 cancer was 0.89 (95% CI: 0.82-0.97). The overall SMR for leukemia was 1.09 (24 deaths) and
5 1.44 (15 deaths) for ML. After 10 years of exposure, the risk for ML was 2.19. Exposure prior
6 to 1963 was associated with a risk of 1.61. Among garment workers followed for 20 or more
7 years from initial exposure, the SMR was significantly elevated for ML (1.91; p < 0.05; 13
8 deaths), as was the SMR for multiple cause leukemia (1.92 [95% CI: 1.08-3.17]; 15 deaths) in
9 the subgroup with 10 or more years of exposure to formaldehyde and who were followed for 20
10 or more years after first exposure. The multiple cause mortality for ML for this subgroup of
11 workers was also significant (SMR 2.55 [95% CI: 1.10-5.03]; 8 deaths).
12 The study by Stayner et al. (1988) has only limited power to detect excess risks of rare
13 cancers, such as NPC and nasal cancer (13 and 16%, respectively). Limitations to the
14 interpretations of the findings include a lack of any monitoring data before 1981, particularly
15 during the critical time period 1955 to 1962, and lack of personal exposure estimates for any
16 members of the cohort. The possibility exists that misclassification may still be present because
17 the intensity of exposure to formaldehyde decreased as improvements were made in the resin
18 systems used to treat fabrics (e.g., a person who worked 5 years beginning in 1955 might have
19 been subject to greater exposure than a person who worked 5 years beginning in 1993).
20 However, workers from the 1950s and 1990s were both placed in the same category of having
21 worked fewer than 10 years. The median duration of exposure was 3.3 years. Work histories
22 were not updated in the follow-up study; however, the low or background exposure levels that
23 probably existed after 1981 were not likely to contribute substantially to the risk of cancer. The
24 use of mortality data to estimate risk, when the case fatality rate was less than 100% for most
25 cancer sites evaluated, could potentially produce an underestimate of the actual risk. Despite
26 these limitations, this study provides additional evidence of an association between leukemia,
27 especially ML, and formaldehyde in comparison with the general population.
28 Gardner et al. (1993) reported that the risk of leukemia was not statistically significant
29 (SMR 0.9) based on 15 deaths among workers employed before 1965. Only four leukemia
30 deaths were observed after 1964 through 1989, producing an SMR of 0.9.
31 When Coggon et al. (2003) updated the above cohort study of 14,014 men first employed
32 before 1965 in six factories by adding 11 additional years of follow-up (ending December 31,
33 2000), no increase in the risk of leukemia or related cancers of the hematopoietic system was
34 reported, either in the entire cohort (SMR 0.91; 31 observed) or in the group with the highest
35 formaldehyde exposure (>2 ppm) (SMR 0.71; 8 observed). Similar results were obtained for
This document is a draft for review purposes only and does not constitute Agency policy.
4-171 DRAFT—DO NOT CITE OR QUOTE
-------
1 Hodgkin's disease, non-Hodgkin's lymphoma, and multiple myeloma. No other cancers of the
2 hematopoietic system were evaluated, and no additional analyses were performed to assess a
3 possible leukemia risk. However, the main finding from this study was a marked association of
4 lung cancer with formaldehyde (discussed in the lung cancer section). This study's main focus
5 was respiratory disease, lung cancer, and stomach cancer, not LHP cancers. For cancers of the
6 LHP system, there was neither latency evaluation nor internal comparisons. The HWE is also
7 potentially a problem.
8 Andjelkovich et al. (1995) studied a cohort of 3,929 male iron foundry workers
9 potentially exposed to formaldehyde between January 1, 1960, and December 31, 1989, in which
10 127 cancer deaths had occurred during the observation period. An industrial hygienist, after
11 reviewing work histories, categorized formaldehyde exposure into four levels corresponding to
12 the approximate midpoint of the ranges: none, low (0.05 ppm), medium (0.55 ppm), and high
13 (1.5 ppm) for exposure to formaldehyde. Boundaries of these exposure categories were not
14 given. The authors warned that the assignment of exposure levels was not perfect because
15 "subjective judgment had to be applied in many instances." SMRs were based on U.S. male
16 mortality rates, but actual ranges were not specified. The authors also compared the exposed to
17 2,032 nonexposed workers from the same company. The population-based SMR for
18 hematopoietic cancer in the exposed population was 0.59 (based on seven observed deaths). For
19 leukemia the SMR was 0.43, based on two deaths. There were no other analyses for leukemia or
20 LHP cancers in this study. Because of the uncertainty about workers' true formaldehyde
21 exposure, there was no analysis by level of exposure, duration, or latency. There were also very
22 few LHP cancers in the cohort. Thus, these results neither support nor refute an association of
23 formaldehyde exposure with LHP cancers. The main focus of this paper was on lung cancer risk.
24 Bertazzi et al. (1989, 1986), in a cohort mortality study, followed 1,330 male workers
25 from 1959 through 1986 at a formaldehyde resin plant in Italy. The workers had to have been
26 employed for at least 30 days at the plant sometime between 1959 and 1980 to be included in the
27 study. Their mortality was compared with national and local rates adjusted for age and calendar
28 time period. No individual exposure estimates were available, but mean levels were estimated to
29 be between 0.2 and 3.8 mg/m3 (0.16 and 3.1 ppm) during the period 1974-1979. The authors
30 found an SMR of 2.01 (five deaths observed) for cancer of the lymphatic and hematopoietic
31 system. The study's limitations included incomplete work histories, small numbers of deaths,
32 and a follow-up period that may not have been sufficient to allow for a latency period for the
33 development of LHP cancers. As before, the results neither support nor refute an association of
34 formaldehyde exposure with LHP cancers.
This document is a draft for review purposes only and does not constitute Agency policy.
4-172 DRAFT—DO NOT CITE OR QUOTE
-------
1 Edling et al. (1987) reported on the incidence of disease in a cohort of 521 blue collar
2 Swedish workers in plants where abrasives bound with formaldehyde resins were manufactured.
3 Formaldehyde levels ranged from 0.1 to 1.0 mg/m3 (0.08-0.8 ppm). The workers in the cohort
4 were employed between 1955 and 1983, and incidence rates were calculated from 1958 through
5 1981. There were only 24 total cancer cases (28.5 expected) of which 2 (1.0 expected) were
6 lymphomas and 2 (0.5 expected) were multiple myelomas. Expected cases were determined
7 through the Swedish National Cancer Register. No other LHP cancers were observed. This
8 study lacked the power to detect any significant associations between LHP cancer and exposure
9 to formaldehyde.
10 Dell and Teta (1995) conducted a cohort mortality study of 5,932 male employees of a
11 New Jersey plastics manufacturing, research, and development facility that produced phenol-
12 formaldehyde resins. The workers, who had been employed during the period 1946-1967, were
13 followed-up for an average of 32 years. SMRs were based on U.S. and New Jersey mortality
14 rates. Hourly workers (n = 3,853) were analyzed separately from the 2,079 salaried employees.
15 Although no excess risk was evident for hematopoietic cancer in hourly workers (SMR 0.93;
16 28 observed), there was an SMR of 1.69 (95% CI: 1.07-2.53; 23 observed) among salaried
17 workers. This association was further narrowed to mainly research and development workers
18 (eight leukemia deaths observed with three expected, for an SMR of 2.67). No common
19 exposure was found when work history records were examined. The decedents were mostly
20 associated with process development in two research pilot plants, where chemical engineers, lab
21 technicians, and plant operators executed small-scale product development. Although notebooks
22 referred to benzene and toluene solvents, no definite connection was made with formaldehyde or
23 any of the solvents. No ambient air measurements of formaldehyde were available. The
24 findings cannot be assumed to be due to formaldehyde exposure because of the presence of other
25 potential leukemogens.
26 Blair et al. (1993) conducted a study that evaluated the risk of non-Hodgkin's lymphoma
27 from exposure to formaldehyde. This was a population-based, case-control, interview-based
28 study of 1,867 white males of whom 622 cases had the disease and 1,245 were controls.
29 Subjects had lived in Iowa and Minnesota between 1980 and 1983. This study was exploratory
30 and designed to find associations with any environmental exposures and non-Hodgkin's
31 lymphoma. Subjects or next of kin were interviewed to determine what exposures the cases and
32 controls may have received based on agricultural exposures, work histories, medical conditions,
33 and family history. Extra effort was made to collect information about occupation, industrial
34 exposures, and other selected exposures. The analysis revealed an OR of 1.2 for exposure to
35 formaldehyde. Similar associations were found for metals and other substances in the study.
This document is a draft for review purposes only and does not constitute Agency policy.
4-173 DRAFT—DO NOT CITE OR QUOTE
-------
1 This study, because it did not select cases and controls from a population with possible
2 formaldehyde exposure, could not detect specific relationships between formaldehyde and
3 non-Hodgkin's lymphoma.
4 Wang et al. (2009) assessed the effect of formaldehyde exposure on the risk of
5 non-Hodgkin lymphoma in a population-based case-control study among women in Connecticut.
6 Incident cases (N = 601) were frequency matched to 717 controls by age. A standardized
7 questionnaire was used to gather information on lifetime occupational history and other risk
8 factors. Exposure to organic solvents and formaldehyde for each job was assessed by linking
9 study participant's occupational data to a job-exposure matrix created by industrial hygienists at
10 the National Cancer Institute (Dosemeci et al., 1994; Gomez et al., 1994). Semiquantitative
11 exposure metrics included average exposure intensity and average exposure probability which
12 were evaluated individually and together. Analyses use unconditional logistic regression and
13 controlled for age, family history of hematopoietic cancers, alcohol consumption and race. For
14 the low average exposure intensity category the OR =1.4 (95% CI: 1.0-1.8), while for the
15 Medium-High category the OR =1.2 (95% CI: 0.8-1.7). For the Low average exposure
16 probability category the OR =1.3 (95% CI: 1.0-1.7), while for the Medium-High category the
17 OR =1.4 (95% CI: 0.9-2.3). When both factors were considered jointly, the Low intensity and
18 Low probability exposure category had OR =1.4 (95% CI: 1.1-1.9); Medium-High intensity and
19 Low probability exposure category had OR =1.0 (95% CI: 0.7-1.6); Low intensity and Medium-
20 High probability exposure category had OR = 1.1 (95% CI: 0.5-2.4); Medium-High and
21 Medium-High probability exposure category had OR =1.6 (95% CI: 0.9-3.1). The investigators
22 also examined the risk of non-Hodgkin lymphoma among major subtypes. The risk of follicular
23 lymphoma and chronic lymphocytic leukemia/Small lymphocytic lymphoma was slightly
24 elevated but the risk of diffuse large B-cell lymphoma was OR =1.9 (95% CI: 1.3-2.6) for ever
25 having been exposed to formaldehyde. For Low average intensity exposure, the risk was
26 OR = 2.1 (95% CI: 1.3-2.6) while for Medium-High average intensity exposure, the risk was
27 OR =1.5 (95% CI: 0.9-2.4). Even so, an exposure-response relationship was demonstrated using
28 the continuous parameterization of average intensity rather that the categorical (p = 0.03).
29 Likewise, an exposure-response relationship was demonstrated using the continuous
30 parameterization of average probability of exposure rather that the categorical (p = 0.01).
31 The findings of Wang et al. (2009) provide some support for an association between
32 formaldehyde exposure and non-Hodgkin lymphoma. It should be noted that a population-based
33 case-control study, where incidence rather than mortality defines the case—may be more
34 appropriate for cancers with relatively low mortality (i.e., CCL, large B-cell lymphoma, Small
35 lymphocytic lymphoma).
This document is a draft for review purposes only and does not constitute Agency policy.
4-174 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.1.2.2.1.3. Meta-analyses of epidemiological studies for lymphohematopoietic malignancies.
2 Several meta-analyses have reported risks for all lymphohematopoietic cancers and
3 leukemia (Collins and Lineker, 2004; Bosetti et al., 2008; Zhang et al., 2009). The meta-analysis
4 conducted by Collins and Lineker (2004) was based on 18 studies. Fixed-effects models were
5 used to obtain summary relative risk values and 95% confidence intervals, and random effects
6 models were used to evaluate heterogeneity across studies. The summary RR across all studies
7 was 1.1 (95% CI = 1.0 to 1.2) for leukemia. The effect estimates varied by type of study,
8 country of study population, type of industry, year of publication, and study size. The cohort
9 studies had a summary RR =1.0 (95% CI: 0.9-1.2) while the summary estimate for the case-
10 control studies was RR = 2.4 (95% CI: 0.9-6.5). For industrial type jobs the RR was 0.9 (95%
11 CI: 0.8-1.0) while for embalmers the RR was 1.6 (95% CI: 1.2-2.0) and for pathologists and
12 anatomists the RR was 1.4 95% CI: 1.0-1.9).
13 Bosetti et al. (2008) reviewed cohort studies of industry workers and health professional
14 using fixed-effect and random-effect models depending on the degree of heterogeneity among
15 the individual cohorts. They did not indentify sufficient heterogeneity among the studies and
16 used the fixed effect models to estimate the relative risk of lymphatic and hematopoietic cancers
17 among four studies of industrial workers exposed to formaldehyde with RR = 0.85 (95% CI:
18 0.74-0.96) and among eight studies of health professionals with RR = 1.31 (95% CI: 1.16-1.47).
19 Likewise for leukemia, they used the fixed effect model to the estimate the RR among four
20 studies of industry workers of lymphatic and hematopoietic cancers among workers exposed to
21 formaldehyde with RR = 0.90 (95% CI: 0.75-1.07) and among eight studies of health
22 professional with RR= 1.39 (95% CI: 1.15-1.68).
23 Zhang et al. (2009) reviewed many of the same studies of lymphohematopoietic cancers
24 and all leukemia that were included in the Bosetti et al. (2008) meta-analysis using a different
25 methodology that focused on the highest exposure groups in each study in order to increase the
26 statistical power of the meta-analysis by minimizing bias from type II error. This method of
27 focusing on the highest exposure groups also reduces the likelihood that any individual study
28 results were confounded or otherwise biased (Greenland, 1998). The authors noted that in the
29 presence of a true causal association, "combining workers with very low exposures with workers
30 with high exposures into one overall exposed group can dilute relative risk estimates towards the
31 null." The authors also preferentially selected the results for myeloid leukemia over all leukemia
32 when these data were reported. This methodology is appropriate if the underlying hypothesis is
33 that formaldehyde causes myeloid leukemia which has the ancillary effect of increasing the risk
34 of all leukemias. By increasing the specificity of the endpoint definition, this meta-analytic
35 method does not create any bias but rather increases the statistical power of the study to detect an
This document is a draft for review purposes only and does not constitute Agency policy.
4-175 DRAFT—DO NOT CITE OR QUOTE
-------
1 association with formaldehyde that is driven by myeloid leukemia. If the any association
2 between formaldehyde exposure and all leukemia were driven by nonmyeloid leukemias, then
3 this method would have less statistical power.
4 Zhang et al. (2009) also used fixed-effects and random-effect models depending on the
5 degree of heterogeneity among the individual cohorts. Zhang et al. (2009) found the
6 heterogeneity among the studies sufficiently high to warrant use of the random-effects model and
7 reported a summary RR = 1.25 (95% CI: 1.09-1.43) for the 19 studies with data on all types of
8 lymphohematopoietic cancers combined. For the 15 studies reporting data on all leukemias, the
9 summary RR= 1.54 (95% CI: 1.18-2.00). For the six studies of myeloid leukemia, the summary
10 RR = 1.90 (95% CI: 1.31-2.76). The authors concluded that the primary reason that their meta-
11 analytic results differed from those previously reported, was attributable to their differential use
12 of the results from the studies by Hauptmann et al. (2004, Stroup et al. (1986) and Pinkerton et
13 al. (2004) for which they used either the relative risks for myeloid leukemia or the highest
14 exposure category in each study.
15 A sensitivity analysis showed that had Zhang et al. (2009) replaced the results of those
16 three studies with the results used in Bosetti et al (2008) meta-analyses, the summary relative
17 risk for leukemia dropped from 1.54 (95% CI: 1.18-2.00) to RR= 1.10 (95% CI: 0.93-1.31).
18 The authors state that the primary reason for the different results is the use of myeloid-specific
19 data from Hauptmann et al. (2003), Stroup et al. (1986) and Pinkerton et al. (2004). Bosetti et al.
20 (2008) use the SMR results from Hauptmann et al. (2003) while Zhang et al. (2009) use the
21 highest peak exposure results for myeloid leukemia. The use of the result from the internal
22 comparison of Hauptmann et al. (2003) appears to be more appropriate. However, those data
23 from Hauptmann et al. (2003) were missing 1,006 deaths and were reanalyzed by Beane
24 Freeman et al. (2009) and were judged to be similar. The correct RR for the highest exposure
25 level of peak exposure should have been 2.79 rather than the value of 3.46 used by Zhang et al.
26 (2009). Likewise, the value of the SMR used by Bosetti et al. (2008) was 0.93 rather than
27 SMR = 0.85 which they used. In a similar fashion, the Bosetti et al. (2008) meta-analysis used
28 the overall SMR of 1.19 from Stroup et al. (1986) while Zhang et al. (2009) used the myeloid
29 leukemia SMR for longest duration which was 2.19. Zhang et al. (2009) also used the SMR
30 from Coggon et al. (2003) for the highest exposure group which was 0.71 while Bosetti et al.
31 (2008) used the overall SMR of 0.91. The Zhang et al. (2009) also excluded five studies that
32 were included by Collins and Lineker (2004). Had the Zhang et al. (2009) analysis included
33 these five studies, their summary relative risk would have been RR = 1.38 (95% CI: 1.15-1.65).
34 Zhang et al. (2009) also report results of the meta-analysis of formaldehyde and other
35 lymphohematopoietic cancers. The RR for Hodgkin lymphoma was 1.23 (95% CI: 0.67-2.29)
This document is a draft for review purposes only and does not constitute Agency policy.
4-176 DRAFT—DO NOT CITE OR QUOTE
-------
1 and was based on 8 studies. The RR for non-Hodgkin lymphoma was 1.08 (95% CI: 0.86-1.35)
2 and was based on 11 studies while the RR for multiple myeloma was 1.31 (95% CI: 1.02-1.67).
3 The criteria for study inclusion and exclusion applied by Zhang et al. (2009) appear to be
4 appropriate and the methodology for using myeloid-specific results where possible also appears
5 to be appropriate. This study found statistically significant increases in risk of all
6 lymphohematopoietic cancers with RR = 1.25(95% CI: 1.12-1.39), all leukemia with RR = 1.54
7 (95% CI: 1.24-1.91), myeloid leukemia with RR = 1.90 (95% CI: 1.41-2.55) and multiple
8 myeloma with RR = 1.31 (95% CI: 1.02-1.67).
9 Of the three meta-analyses, the study by Zhang et al. (2009) appears to be the most
10 rigorous methodologically. One difference between them is that while Zhang et al. (2009)
11 combined across industry, Collins and Lineker (2003) and Bosetti et al. (2008) stratified by
12 industry which may have reduced the statistical power of those analyses. Nonetheless, both
13 Collins and Lineker (2003) and Bosetti et al. (2008) both reported statistically significant
14 increases in risk of leukemia mortality among the nonindustrial workers. And Bosetti et al.
15 (2008) also reported statistically significant increases in risk of all lymphohematopoietic cancers
16 mortality among the nonindustrial workers.
17
18 4.1.2.2.1.4. Summary oflymphohematopoietic cancers.
19 All LHP Malignancies
20 The majority of studies of all LHP malignancies (as a group) have been based on
21 comparison of the risk of cancer incidence or mortality in the studied population compared to the
22 risk in a general population. These external analyses rely on the assumption that cancer
23 incidence or mortality rates are expected to be similar between the general population and the
24 study population in the absence of exposure. However, the 'healthy worker effect' is well
25 known to bias rates in workers downwards compared to general populations, and there may be
26 differences in the magnitude of this selection bias by industry or profession.
27 Positive associations between formaldehyde exposure and LHP cancers have been
28 reported based on external comparison groups for chemical workers (Wong, 1983; Bertazzi et
29 al., 1986), embalmers (Walrath and Fraumeni, 1983, 1984; Hayes et al., 1990), anatomists and
30 pathologists (Harrington and Shannon, 1975; Hall et al., 1991; Levine at al., 1984; Stroup et al.,
31 1986; Matanoski et al., 1989). However, associations were not reported in external comparison
32 group analyses for garment workers, iron-foundry workers, and a large US industrial cohort
33 (Pinkerton et al., 2004; Andjelkovich et al., 1995; Beane Freeman et al., 2009; Marsh et al.,
34 1996), although associations were observed in some of these studies when exposure-response
35 relationships were considered.
This document is a draft for review purposes only and does not constitute Agency policy.
4-177 DRAFT—DO NOT CITE OR QUOTE
-------
1 Several published meta-analyses are available which more formally assess the strength of
2 association between formaldehyde exposure and mortality from all LHP cancers. Bosetti et al.
3 (2008) reviewed cohort studies of industry workers and health professionals (i.e., embalmers,
4 anatomists, and pathologists). They used fixed effect models to estimate the relative risk of
5 lymphatic and hematopoietic cancers among four studies of industrial workers exposed to
6 formaldehyde with RR = 0.85 (95% CI: 0.74-0.96) and among eight studies of health
7 professionals with RR =1.31 (95% CI: 1.16-1.47). A subsequent meta-analysis by Zhang et al.
8 (2009) reports a summary relative risk (RR) of 1.25 (95% CI 1.09-1.43) for both professional
9 and industry workers for all LHP cancers (ICD 9 codes 200-209). These researchers identified
10 19 cohort study analyses, including cohort study updates. Zhang et al. (2009) used the reported
11 RR from the highest exposure category to increase statistical power and reduce uncertainty
12 regarding confounding or other bias. These two summary evaluations of the weight of evidence
13 of a causal association between exposure to formaldehyde and increased risk of mortality from
14 all LHP malignancies both showed statistically significant association. Therefore it is judged
15 that the collective evidence shows a causal association for all LHP as a group.
16 While most of the studies reporting SMRs in the meta-analyses were limited in their
17 exposure assessment to work in an industry known or thought to be exposed to formaldehyde,
18 some of the studies did conduct detailed exposure assessment and were able to compute internal
19 comparison of risk within the group of studied workers. These internal analyses are less likely to
20 be biased by the selection pressures for hiring and retaining healthy workers and represent
21 stronger methodol ogi e s.
22 The two individual studies with the relatively strongest exposure assessment reported
23 results that are consistent with the results of the meta-analyses (Hauptmann et al., 2009; Beane
24 Freeman et al. 2009). Hauptmann et al. (2009) conducted extensive interviews with the next of
25 kin and coworkers of the cases and controls provided detailed information on the funeral homes
26 and work practices of the study subjects. The authors noted that since the funeral industry is
27 often a family business, these study subjects' next of kin were believed to be unusually
28 knowledgeable of funeral home work practices. The work history component of these interviews
29 provided data on the frequency and duration of embalmings for jobs held at least five years, as
30 well as information on ventilation of the premises and the frequency of spills. These data were
31 linked to data from an exposure-reconstruction experiment which sought to replicate standard
32 funeral home practices while measuring exposures to formaldehyde.
33 While the results of Hauptmann et al. (2009) are not conclusive of an association on a
34 stand-alone basis, they do show consistently elevated odds ratios for all LHP associated with
35 ever embalming, duration of working in embalming jobs, number of embalmings and cumulative
This document is a draft for review purposes only and does not constitute Agency policy.
4-178 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde exposure. Of the multiple OR results presented for all LHP, 17 of 19 were
2 elevated but only two were statistically significantly elevated. For the middle category of
3 duration, the OR = 1.5 (95% CI: 0.8-2.8) and for the highest category of duration, the OR = 1.8
4 (95% CI: 1.0-3.4) with a borderline significant test for trend (p = 0.058). For the middle
5 category of number of embalming, the OR =1.9 (95% CI: 1.0-3.6). Based on a binomial
6 distribution assuming that half the ORs would be elevated and half below unity, the probability
7 of 17 OR being elevated out of 19 isp = 0.0003. While only two of the OR results were
8 statistically significant elevated, the overall findings were consistent with the results of the meta-
9 analyses. Hauptmann et al. (2009) show elevated risks of all LHP of a somewhat higher
10 magnitude than the summary effects in the meta-analyses that are, however, less statistically
11 precise as they are based on a single study rather than on multiple studies.
12 The individual study with the most detailed and objectively ascertained exposure
13 assessment was that of Beane Freeman et al. (2009). Exposure to formaldehyde was estimated
14 for each individual job category from work histories, with calendar-time and plant-specific
15 estimates based on assessments of job titles and tasks associated with those jobs using plant
16 visits by industrial hygienists and monitoring data (Blair et al. 1986; Stewart et al. 1986; Blair
17 and Stewart 1990). Exposures were categorized by peak exposure, average intensity of exposure
18 and cumulative exposure. The median follow-up time for workers was 42 years, representing
19 998,106 person-years of follow-up among 25,619 workers (Beane Freeman et al., 2009).
20 Internal analyses of exposed workers indicated that peak exposures in the highest
21 exposure category were associated with a significant increase in all lymphohematopoietic deaths
22 comparing death rates among workers with peaks of > 4 ppm to those with > 0 to 2.0 ppm
23 (RR = 1.37, 95% CI: 1.03-1.81). Across the three categories of peak exposure (i.e., exposure
24 >0 ppm), there was also a statistically significant exposure-response trend (p = 0.02). The
25 exposure-response trend including the never exposed workers was also statistically significant
26 (p = 0.04). However, no association was observed for all lymphohematopoietic cancers for
27 average intensity or cumulative exposure.
28 The results of Beane Freeman et al. (2009) confirm those of the meta-analyses showing
29 statistically significant increased risk at the highest category of peak exposure that was
30 associated with all LHP mortality. These results do not appear to be confounded by any known
31 coexposures as the investigators reported that controlling for duration of exposure to 11
32 potentially confounding coexposures did not meaningfully change results. As these results are
33 based on a large and well-followed occupational cohort, these results are unlikely to be
34 influenced by selection bias. The estimated magnitude of the identified association (OR = 1.37)
This document is a draft for review purposes only and does not constitute Agency policy.
4-179 DRAFT—DO NOT CITE OR QUOTE
-------
1 was approximately equivalent to the summary RR of Zhang et al. (2009) which was 1.25 and the
2 summary RR =1.31 for health professionals analyzed by Bosetti et al. (2008).
3 Given the consistency and strength of the positive associations for all LHP cancer
4 mortality in professional cohorts (embalmers, anatomists and pathologists) taken together with
5 the strong positive results of the NCI cohort, human epidemiologic evidence are sufficient to
6 conclude that there is a causal association between formaldehyde exposure and mortality from all
7 LHP malignancies (as a group).
8
9 All Leukemia
10 Like the studies of all LHP (as a group), the majority of studies of all leukemia
11 malignancies (as a group) have been based on comparison of the risk of cancer incidence or
12 mortality in the studies population compared to the risk in a general population and may be
13 influenced by the healthy worker effect which effective depresses effect estimates. In spite of
14 this potential bias which may mask a true effect of exposure, an association between
15 formaldehyde exposure and leukemia mortality is supported by cohort analyses of embalmers,
16 pathologists and anatomists (Hayes et al., 1990; Walrath and Fraumeni, 1983; Walrath and
17 Fraumeni 1984; Hall et al., 1991; Levine et al., 1984; Stroup et al., 1986; Matanoski et al., 1989).
18 Formaldehyde exposure and formaldehyde-related occupation have also been shown to be
19 associated with leukemia diagnosis in a case-control study (RR = 5.79 (95% CI 1.44-23.25), but
20 not formaldehyde exposure alone (RR = 0.96; 95% CI 0.54-1.71) (Stellman et al., 1998).
21 However, SMR analyses of the large industrial cohorts do not indicate a similar
22 association (Coggon et al., 2000; Beane Freeman et al., 2009, Pinkerton et al, 2004). Although
23 the SMR analysis provided for the NCI cohort does not indicate a positive association for all
24 leukemia using an external reference group (Beane Freeman et al., 2009), the SMR for exposed
25 versus unexposed workers within the cohort suggests all leukemia is elevated 2.13-fold with this
26 type of internal comparison (95% CI 0.99-4.56).
27 Several meta-analyses have been conducted for formaldehyde exposure and all leukemia
28 (as a group) which indicate a positive association. Collins and Lineker (2004) reported an
29 overall RR for 18 available studies of 1.1 (95% CI: 1.0-1.2), suggesting an association of
30 leukemia with formaldehyde exposure. This association was stronger for both
31 pathologists/anatomists (1.4; 95% CI: 1.0-1.9) and embalmers (RR =1.6; 95% CI: 1.2-2.0) than
32 for industrial workers (RR = 0.9; 95% CI: 0.8-1.0). Study design also impacted the apparent
33 strength of association, with stronger associations seen in case-control studies (RR = 2.4; 95%
34 CI: 0.9-6.5) versus cohort studies (RR = 1.0; 95% CI: 0.9-1.2). Bosetti et al. (2008) reported an
35 association between formaldehyde exposure and leukemia mortality with a pooled RR of 1.39
This document is a draft for review purposes only and does not constitute Agency policy.
4-180 DRAFT—DO NOT CITE OR QUOTE
-------
1 (95% CI 1.15-1.68) for 8 groups of professional workers. In the same analysis, the pooled RR
2 for the 4 industrial cohorts was 0.90 (95% CI: 0.75-1.07). Zhang et al. (2009) reported a pooled
3 RR of 1.54 (95% CI: 1.18-2.00) for all cohorts identified in their meta-analysis, although this
4 pooled RR should be considered with the knowledge that when the source studies specifically
5 identified results for myeloid leukemia those results and not the all leukemia results were
6 included (Zhang et al., 2009). As discussed earlier, this methodology is appropriate if the
7 underlying hypothesis is that formaldehyde causes myeloid leukemia which has the ancillary
8 effect of increasing the risk of all leukemias. By increasing the specificity of the endpoint
9 definition, this meta-analytic method does not create any bias but rather increases the statistical
10 power of the study to detect an association with formaldehyde that is driven by myeloid
11 leukemia. If the association between formaldehyde exposure and all leukemia as a group were
12 driven by nonmyeloid leukemias, then this method would have less statistical power.
13 Of two individual studies with the relatively strongest exposure assessment (Hauptmann
14 et al., 2009; Beane Freeman et al. 2009), Hauptmann et al. (2009) did not report specifically on
15 leukemia (ICD-8: 205-207) but rather on lymphohematopoietic cancers of nonlymphoid origin
16 (ICD-8: 205, 206, 208 or 209) and on myeloid leukemia specifically (ICD-8: 205). Statistically
17 significant increased risk were reported for the larger grouping for 'ever embalming' with
18 RR = 3.0 (95% CI: 1.0-9.5), for the highest category of duration with RR = 3.7 (95% CI:
19 1.1-12.2), for the highest category of number of embalmings with RR = 3.9 (95% CI: 1.2-12.8)
20 and for the highest category of cumulative exposure with RR = 4.0 (95% CI: 1.2-13.2).
21 However, none of the associated linear tests for tend were significant. Statistically significant
22 increased risks were also reported for the highest category TWA formaldehyde intensity with
23 RR = 3.4 (95% CI: 1.0-11.8) and the highest category of peak formaldehyde exposure with
24 RR = 3.8 (1.1-12.7), both without showing a linear trend. These results were considered to be
25 consistent with the meta-analytic findings of an association between formaldehyde exposure and
26 all leukemia as a group.
27 While not the same as all leukemia as a group, Hauptmann et al. (2009) did look
28 specifically at myeloid leukemia and reported extremely strong associations of each exposure
29 metrics with the majority of RRs greater than 10 and statistically significant as detailed in the
30 following subsection devoted to myeloid leukemia specifically. These myeloid leukemia results
31 are strongly supportive of a causal association with formaldehyde and therefore are, in turn,
32 supportive of a causal relationship with all leukemia as a group. The study by Beane Freeman et
33 al. (2009) did look specifically at all leukemia as a group (ICD-8: 204-207) and reported an
34 elevated RR = 1.42 (95% CLO.92-2.18) for the highest category of peak exposure that was not
35 statistically significant. Two trend tests were reported for peak exposures with the trend within
This document is a draft for review purposes only and does not constitute Agency policy.
4-181 DRAFT—DO NOT CITE OR QUOTE
-------
1 only the exposure groups having ap = 0.12 and the trend across all exposure groups including
2 the unexposed having a significant trend (p = 0.02). Results from the analyses of average
3 formaldehyde intensity did not show an association while results from the analyses of cumulative
4 exposure did show some indication of a trend both within the exposed workers (p = 0.12) and
5 across all the workers (p = 0.08).
6 While the epidemiologic evidence for a causal association between formaldehyde and all
7 leukemia as a group is not at strong as for all LHP as a group, the repeated identification of an
8 association in multiple meta-analyses taken together with the clear causal association between
9 myeloid leukemia demonstrated by Hauptmann et al. (2009) and the consistent evidence reported
10 by Beane Freeman et al. (2009) are sufficient to conclude that there is a causal association
11 between formaldehyde exposure and mortality from all leukemia as a group.
12
13 Myeloid Leukemia
14 The associations between myeloid leukemia and formaldehyde exposure are strong and
15 consistent. Of the six studies which formally assess myeloid leukemia mortality, five are
16 positive, including cohorts of both professional and industrial workers (Hauptmann et al., 2009;
17 Pinkerton at al., 2003; Hayes et al., 1990; Stroup et al., 1986; Walrath and Fraumeni, 1984,
18 Walrath and Fraumeni, 1983; but not Beane Freeman et al., 2009).
19 Walrath and Fraumeni (1983) studied embalmers in NY and reported 6 deaths from
20 myeloid leukemia compared to 4.1 expected based on the US male population stratified by age
21 and calendar time which is equivalent to a SMR= 1.46 (exact 95% CI: 0.54-3.19). Walrath and
22 Fraumeni (1984) studied embalmers in California and reported 6 deaths from myeloid leukemia
23 compared to 4.0 expected based on the U.S. male population stratified by age and calendar time
24 which is equivalent to a SMR = 1.50 (exact 95% CI: 0.55-3.26). While these studies were called
25 proportionate mortality studies by the authors, the reported PMRs are more actually SMRs and
26 should be interpreted as such.
27 Stroup et al. (1986) conducted an historic cohort mortality study of men who were
28 members of the American Association of Anatomists between 1888 and 1969. Only 738 deaths
29 were observed versus 1,133.9 expected, based on U.S. death rates (SMR 0.65), possibly
30 indicating a sizable HWE. However, a slight increase in the risk of lymphatic and hematopoietic
31 cancers (SMR 1.2; 18 observed) and the risk of leukemia (SMR 1.5; 10 observed) was evident.
32 When the leukemia analysis was restricted to the myeloid type, the SMR increased to 8.8, based
33 on five deaths (p < 0.05).
34 Hayes et al. (1990) conducted a cohort study of deceased U.S. male embalmers and
35 funeral directors who had died between 1975 and 1985, using records from local licensing
This document is a draft for review purposes only and does not constitute Agency policy.
4-182 DRAFT—DO NOT CITE OR QUOTE
-------
1 boards, state funeral directors' associations in 32 states and the District of Columbia, the
2 National Funeral Directors' Association, and state offices of vital statistics (n = 894). Expected
3 deaths by cause were derived from 5-year age- and calendar-year-specific proportions of deaths
4 among appropriate race groups from the U.S. population. No measured exposure data were
5 available. Statistically significant excesses in hematopoietic and lymphatic cancers were found
6 in embalmers and funeral directors. The PMR (actually SMR) was 1.39 (95% CI: 1.15-1.63).
7 The excess risk in all males was higher for myeloid leukemia (ML) (PMR (actually SMR) 1.57
8 [95% CI: 1.01-2.34]; 24 observed).
9 The Pinkerton et al. (2004) occupational cohort study of garment workers also reported
10 SMRs derived from age-, race-, and calendar-time-adjusted U.S. mortality rates. The overall
11 SMR for leukemia was 1.09 (24 deaths) and 1.44 (95% CI: 0.80-2.37; 15 deaths) for ML. After
12 10 years of exposure, the risk for ML was 2.19 (exact 95% CI: 0.95-4.32). Exposure prior to
13 1963 was associated with a risk of 1.61 (exact 95% CI: 0.80-2.88). Among garment workers
14 followed for 20 or more years from initial exposure, the SMR was significantly elevated for ML
15 (1.91 [exact 95% CI: 1.02-3.26]; 13 deaths), as was the SMR for multiple cause leukemia (1.92
16 [95% CI: 1.08-3.17]; 15 deaths) in the subgroup with 10 or more years of exposure to
17 formaldehyde and who were followed for 20 or more years after first exposure. The multiple
18 cause mortality for ML for this subgroup of workers was also significant (SMR 2.55 [95% CI:
19 1.10-5.03]; 8 deaths).
20 The meta-analysis by Zhang et al. (2009) evaluated the studies of formaldehyde exposure
21 and myeloid leukemia available at the time including Hauptmann et al. (2003), Pinkerton at al.,
22 2003; Hayes et al., 1990; Stroup et al., 1986; Walrath and Fraumeni, 1984, Walrath and
23 Fraumeni, 1983. While the findings of Hauptmann et al. (2003) on the NCI cohort have been
24 recently updates by those of Beane Freeman et al. (2009) who updated the cohort, the Zhang et
25 al. (2009) analysis provide the only formal meta-analysis specific to myeloid leukemia. Zhang et
26 al., 2009 reported a statistically significant summary RR of 1.90 (95% CI: 1.31-2.76).
27 As previously mentioned, the two individual studies with the relatively strongest
28 exposure assessment reported results that are also consistent with the results of the meta-analyses
29 (Hauptmann et al., 2009; Beane Freeman et al. 2009). Because the results of Beane Freeman et
30 al. (2009) are based on an update of the Hauptmann et al. (2003) study which was included in the
31 meta-analysis by Zhang et al. (2009) it is not surprising that they are consistent with those of the
32 meta-analyses showing statistically significant increased risk at the highest category of peak
33 exposure that was associated with all LHP mortality. However, Beane Freeman et al. (2009) did
34 not report statistically significant associations between formaldehyde exposure and mortality
35 from myeloid leukemia. At the highest exposure category of peak exposure (peaks >4 ppm vs.
This document is a draft for review purposes only and does not constitute Agency policy.
4-183 DRAFT—DO NOT CITE OR QUOTE
-------
1 >0 to 2.0 ppm), the RR = 1.42 (95% CI: 0.92-2.18) for myeloid leukemia. Two test for trend
2 were report for myeloid leukemia which provide some support for a concentration-response
3 relationship with peak exposure but neither was statistically significant (p = 0.13 andp = 0.07).
4 Hauptmann et al. (2009) reported increases in risk for ever embalming with myeloid
5 leukemia (OR = 11.2, 95% CI: 1.3-95.6). Duration of employment in jobs with embalming
6 demonstrated an exposure-response relationship with increased risk of myeloid leukemia
7 (p = 0.02). The number of embalming was also significantly associated with increased risk of
8 myeloid leukemia in the middle (OR = 12.7, 95%: 1.4-116.7) and highest exposure levels
9 (OR = 12.7, 95% CI: 1.6-119.7). Cumulative exposure was associated with increased risk of
10 myeloid leukemia, with the highest category of exposure showing OR =13.2 (95% CI:
11 1.5-115.4).
12 Hauptmann et al. (2009) showed that all three categories of average formaldehyde
13 intensity while embalming were very strongly and significantly associated with the risk of
14 myeloid leukemia mortality (OR =11.1, 14.8, and 9.5), and the test for an exposure-response
15 trend was of borderline statistical significance (p = 0.058). The risk of myeloid leukemia
16 mortality was also very strongly and significantly associated with 8-hour TWA exposure for
17 mid-level exposures (OR = 13.6, 95% CI: 1.5-125.8) and high-level exposures (OR = 12.0, 95%
18 CI: 1.3-107.4), as well as for peak exposure in two of three categories (0-7 ppm OR = 15.2, 95%
19 CI: 1.6-141.6; 7-9.3 ppm OR =8.0, 95% CI: 0.9-74.0; and >9.3 ppm OR = 13.0, 95% CI: 1.4-
20 116.9). For peak exposures, there was also a statistically significant exposure-response tend
21 (p = 0.036).
22 The study by Hauptmann et al. (2009) stands out among the studies of embalmers and
23 professionals in the funeral industry based on the strength of the quantitative exposure data and
24 the demonstration of exposure-response relationships which provide causal evidence of an
25 association between formaldehyde exposure and increased risk of myeloid leukemia. These
26 results were internally consistent and demonstrated statistically significant associations that were
27 unlikely the result of chance. As this nested case-control study was based on the cohorts of
28 Hayes et al. (1990) and those of Walrath and Fraumeni (1983, 1984), the potential for selection
29 bias is considered to be low. Further, the controls in Hauptmann et al. (2009) were carefully
30 selected to avoid individuals who died of any causes that were thought to even possibly be
31 related to formaldehyde exposure. Confounding is also unlikely to be an alternative explanation
32 for the observed results as there were clear and convincing exposure-responses and the
33 magnitude of the effect estimates were extremely large.
34 Given the consistency of the positive associations for formaldehyde with myeloid
35 leukemia cancer mortality across five of the six studies (Hauptmann et al., 2009; 2009; Pinkerton
This document is a draft for review purposes only and does not constitute Agency policy.
4-184 DRAFT—DO NOT CITE OR QUOTE
-------
1 at al., 2003; Hayes et al., 1990; Stroup et al., 1986; Walrath and Fraumeni, 1984, Walrath and
2 Fraumeni, 1983; but not Beane Freeman et al., 2009), the statistically significant meta analysis
3 by Zhang et al. (2009) and the convincing results from Hauptmann et al. (2009), the human
4 epidemiologic evidence is sufficient to conclude that there is a causal association between
5 formaldehyde exposure and mortality from myeloid leukemia.
6
7 Hodgkin Lymphoma
8 The only meta-analysis to specifically address Hodgkin lymphoma was conducted by
9 Zhang et al. (2009) and included eight studies (Anjelkovich et al., 1995; Coggon et al., 2003;
10 Harrington and Shannon, 1975; Hauptmann et al., 2003; Hayes et al., 1990; Pinkerton et al.,
11 2004; Walrath and Fraumeni, 1983, 1984; and Wong, 1983). Zhang et al. (2009) reported a
12 summary RR = 1.23 (95% CI 0.67-2.29). This elevated, but nonstatistically significant finding is
13 consistent with the large variance on reported results among the individual studies as well as the
14 wide confidence intervals of the results which were based on small numbers of cases—even from
15 the large cohort studies. Six of the eight studies observed three or fewer deaths from Hodgkin
16 lymphoma. Coggon et al. (2003) reported 6 deaths from Hodgkin lymphoma against 8.5
17 expected for an SMT = 0.70 (95% CI: 0.26-1.53) and Hauptmann et al. (2003) reported
18 21 observed deaths with 20 deaths among the exposed workers who has an SMR = 1.26 (95%
19 CI: 0.81-1.95). However, the Beane Freeman et al. (2009) update of the Hauptmann et al. (2003)
20 study had the largest number of observed cases (n = 27) and was not included in the Zhang et al.
21 (2009) meta-analysis. In fact, the Beane Freeman et al. (2009) study describes more deaths from
22 Hodgkin lymphoma than all the other studies in Zhang et al. (2009) combined. Excluding the
23 Hauptmann et al. (2003) results from the list of studies in the meta-analysis leaves 19 cases.
24 There is evidence for an exposure-response relationship for Hodgkin lymphoma in the
25 NCI industrial cohort among exposed workers (Beane Freeman et al., 2009). Clear exposure
26 response relationships for Hodgkin lymphoma are defined with all three metrics of exposure,
27 peak average intensity and cumulative exposure (p = 0.01,/> = 0.05 and/? = 0.08 respectively for
28 mortality through 2004). These associations have been evident from first follow-up through the
29 current publication, and statistically significant for the majority of the follow-up period
30 demonstrating that this is a strong and consistent finding in the NCI cohort (see Figures 4-3 and
31 4-4) (Beane Freeman et al., 2009).
32 As the majority of the studies reported specific data on Hodgkin lymphoma report on just
33 three or fewer cases, the best epidemiologic evidence is obtained from the most recent evaluation
34 of the NCI cohort by Beane Freeman et al. (2009). This cohort study, on its own, reported on
35 more deaths from Hodgkin lymphoma that the remainder of the epidemiologic literature.
This document is a draft for review purposes only and does not constitute Agency policy.
4-185 DRAFT—DO NOT CITE OR QUOTE
-------
1 Hodgkin lymphoma was both shown to be at increased risk associated with peak exposure
2 concentrations. Peak exposures in the highest exposure category were associated with a
3 significant increase in Hodgkin lymphoma deaths comparing death rates among workers with
4 peaks of > 4 ppm to those with > 0 to 2.0 ppm (RR = 3.96, 95% CI: 1.31-12.02). Across the
5 three categories of peak exposure, there was a statistically significant exposure-response trend
6 (p = 0.01). The exposure-response trend including the never-exposed workers was also
7 statistically significant (p = 0.004). The RR was also elevated for average intensity of
8 formaldehyde exposure with RR = 2.48 (95% CI: 0.84-7.32) and there were significant tests for
9 trend among only the exposed workers (p = 0.05) and all workers (p = 0.03). Similarly, there
10 were nearly significant tests for trend with cumulative exposure among only the exposed workers
11 (p = 0.08) and all workers (p = 0.06).
12 The majority of the studies reporting on Hodgkin lymphoma did not have sufficient
13 statistical power describe any potential association with formaldehyde as the numbers of
14 observed and expect cases were small and the resulting effects estimates were imprecise. As the
15 Beane Freeman et al. (2009) study reported on the largest number of cases and was the
16 individual study with the most detailed and objectively ascertained exposure assessment and
17 demonstrated significant exposure-response gradients, it is judged that this epidemiologic
18 evidence is supportive of a causal association between formaldehyde and Hodgkin lymphoma.
19
20 Non-Hodgkin Lymphoma
21 The only meta-analysis to specifically address non Hodgkin lymphoma was conducted by
22 Zhang et al. (2009) and included eleven studies. Zhang et al. (2009) reported a summary
23 RR = 1.08 (95% CI 0.86-1.35). Hauptmann et al. (2009) did not specifically report on
24 non-Hodgkin lymphoma. Beane Freeman et al. (2009) did report on 106 deaths from
25 non-Hodgkin lymphoma but did not identify any significant association in their categorical
26 analyses or in their tests for trend for either peak exposure, average intensity of exposure or for
27 cumulative exposure.
28 Wang et al. (2009) assessed the effect of formaldehyde exposure on the risk of
29 non-Hodgkin lymphoma in a population-based case-control study. Semi quantitative exposure
30 metrics included average exposure intensity and average exposure probability which were
31 evaluated individually and together. Analyses use unconditional logistic regression and
32 controlled for age, family history of hematopoietic cancers, alcohol consumption, and race. For
33 the low average exposure intensity category the OR =1.4 (95% CI: 1.0-1.8), while for the
34 Medium-High category the OR =1.2 (95% CI: 0.8-1.7). For the Low average exposure
35 probability category the OR =1.3 (95% CI: 1.0-1.7), while for the Medium-High category the
This document is a draft for review purposes only and does not constitute Agency policy.
4-186 DRAFT—DO NOT CITE OR QUOTE
-------
1 OR =1.4 (95% CI: 0.9-2.3). The investigators also examined the risk of non-Hodgkin
2 lymphoma among major subtypes. The risk of follicular lymphoma and chronic lymphocytic
3 leukemia/Small lymphocytic lymphoma was slightly elevated but the risk of diffuse large B-cell
4 lymphoma was OR =1.9 (95% CI: 1.3-2.6) for ever having been exposed to formaldehyde. For
5 Low average intensity exposure, the risk was OR = 2.1 (95% CI: 1.3-2.6) while for
6 Medium-High average intensity exposure, the risk was OR =1.5 (95% CI: 0.9-2.4). Even so, an
7 exposure-response relationship was demonstrated using the continuous parameterization of
8 average intensity rather that the categorical (p = 0.03). Likewise, an exposure-response
9 relationship was demonstrated using the continuous parameterization of average probability of
10 exposure rather that the categorical (p = 0.01).
11 The findings of Wang et al. (2009) provide some support for an association between
12 formaldehyde exposure and non-Hodgkin lymphoma. It should be noted that a population-based
13 case-control study, where incidence rather than mortality defines the case—may be more
14 appropriate for cancers with relatively low mortality (i.e., CCL, large B-cell lymphoma, Small
15 lymphocytic lymphoma).
16 Aside from the semi quantitative study by Wang et al (2009), non-Hodgkin lymphoma
17 does not appear to be associated with formaldehyde exposure. There is not sufficient evidence of
18 a causal association between formaldehyde exposure and non-Hodgkin lymphoma.
19
20 Multiple Myeloma
21 The only meta-analysis to specifically address Hodgkin lymphoma was conducted by
22 Zhang et al. (2009) and included nine studies (Boffetta et al., 1989; Coggon et al., 2003; Dell and
23 Teta, 1995; Edling et al. 1987; Hauptmann et al. 2003; Hayes et al. 1990; Heineman et al. 1982;
24 Pottern et al. 1992; Stellman et al. 1998). Zhang et al. (2009) reported a summary RR = 1.31
25 (95% CI 1.02-1.67). This statistically significant finding is consistent with the findings of Beane
26 Freeman et al. (2009) who reported that peak exposures in the highest exposure category were
27 associated with a significant increase in multiple myeloma deaths comparing death rates among
28 workers with peaks of > 4 ppm to those with > 0 to 2.0 ppm (RR = 2.04, 95% CI: 1.01-4.12).
29 Across the three categories of peak exposure, there was also some evidence of an exposure-
30 response trend (p = 0.08); however, there was no evidence of an exposure-response trend
31 including the never-exposed workers. The association of multiple myeloma with formaldehyde
32 exposure was also shown throughout the cohort experience (see Figure 4-3 and 4-4) which adds
33 strength to this finding.
34 The epidemiologic evidence for a causal association between formaldehyde and all
35 multiple myeloma as described by the statistically significant increased risk identified in the
This document is a draft for review purposes only and does not constitute Agency policy.
4-187 DRAFT—DO NOT CITE OR QUOTE
-------
1 meta-analysis of Zhang et al. (2009) and the most recently updated analysis of the NCI cohort by
2 Beane Freeman et al. (2009) are considered to be supportive of a causal association between
3 formaldehyde exposure and mortality from multiple myeloma.
4
5 4.1.2.2.2. Brain and CNS cancer.
6 Several studies of professional groups discussed earlier investigated brain and other CNS
7 cancers among those exposed to formaldehyde on the job. Several of these studies found that
8 exposure increased risk two to three times among exposed professionals (Hall et al., 1991;
9 Stroup et al., 1986; Walrath and Fraumeni, 1984), while others found modest or no increase in
10 risk (Hayes et al., 1990; Levine et al., 1984; Walrath and Fraumeni, 1983).
11 None of the industrial cohort worker mortality studies of exposure to formaldehyde found
12 a clear relationship between formaldehyde exposure and risk of brain or CNS cancer (Pinkerton
13 et al., 2004; Coggon et al., 2003; Andjelkovich et al., 1995; Gardner et al., 1993; Stayner et al.,
14 1988; Blair et al., 1987, 1986). To date, no case-control studies of brain and CNS cancer have
15 been completed. In the Hauptmann et al. (2004) study, the authors reported that no clear
16 association was seen for cancer of the brain and CNS and exposure to formaldehyde.
17
18 4.1.2.2.3. Pancreatic and other cancers.
19 Two studies (Kernan et al., 1999; Dell and Teta, 1995) have found increases in the risk of
20 pancreatic cancer in association with possible exposure to formaldehyde. Collins et al. (2001a)
21 conducted a meta-analysis of fourteen studies (Kernan et al., 1999; Andjelkovich et al., 1995;
22 Hansen and Olsen, 1995; Gardner et al., 1993; Hall et al., 1991; Matanoski, 1991; Hayes et al.,
23 1990; Gerin et al., 1989; Stayner et al., 1988; Blair et al., 1986; Stroup et al., 1986; Levine et al.,
24 1984; Walrath and Fraumeni, 1984, 1983) and found a small increase in risk (RR =1.1 [95% CI:
25 1.0-1.3]).
26 Other sites that have been examined are stomach cancer (Coggon et al., 2003)
27 (SMR = \Al;p < 0.05), intraocular melanoma (Holly et al., 1996) (OR = 2.9 [95% CI:
28 1.2-7.0]), and thyroid cancer among women (Wong et al., 2006) (OR =8.33 [95% CI:
29 1.16-60.0]; 2 cases). However, without further substantiation, it is difficult to infer causation
30 based on these isolated results alone.
31
32 4.1.2.3. Summary: Carcinogenic Hazard in Humans
33 The weight of the epidemiologic evidence at this time supports a link between
34 formaldehyde exposure and NPC in humans. This conclusion is based on the longitudinal cohort
35 study of Hauptmann et al. (2004) as well as the case-control studies of NPC and formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-188 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure completed by Vaughan et al. (2000), West et al. (1993), Vaughan et al. (1986b) and
2 several additional case-control studies described in the text. With the exception of Hauptmann et
3 al. (2004), most of the other cohort studies found little or no increased risk of NPC from
4 exposure to formaldehyde. However, Hauptmann et al. (2004) employed different exposure
5 metrics and based their analyses on conservative internal comparisons that limited the potential
6 for the HWE to obscure true effects. The case-control studies that provide additional evidence of
7 an association between NPC and formaldehyde have more power and generally rely on better
8 diagnoses of NPC. Better ascertainment of histologic types of tumors can sometimes also be
9 obtained if the cases are taken from cancer registries. The NPC risk is also supported by
10 experimental evidence in animals in which formaldehyde induces nasal cancers (see
11 Section 4.2.2). Since the physiology of the rat nasal passage is somewhat different from that of
12 humans, it is not possible to obtain a direct site-specific correspondence between the species.
13 However, in both species, the tumors are found within the same area of the URT where
14 maximum exposure can be expected to occur.
15 Several researchers have challenged the conclusion of a relationship between
16 formaldehyde and NPC. Those critical of the link argue that, given the wide variability in results
17 across studies and competing explanations, conclusions about any link from the existing studies
18 are premature. The difficulty in attaining consensus on whether formaldehyde influences the risk
19 of NPC in humans arises from several limitations inherent in epidemiologic methods and
20 exposure assessment, as well as from the characteristics of the disease. The most prominent of
21 these limitations are the rarity of the cancer and imprecise estimates of exposure. Because NPC
22 is a very rare cancer with an incidence of less than 1 per 100,000, it is difficult to obtain precise
23 estimates of risk from cohort studies. Although case-control studies are better suited for
24 studying rare conditions, they are limited in obtaining valid and precise exposure assessments. A
25 further problem with exposure assessment is isolating formaldehyde exposure from other
26 potential chemical or particulate exposures that may influence risk of NPC. Imprecise exposure
27 assessment and the inability to isolate formaldehyde exposure from other exposures are largely
28 the bases on which Marsh and coworkers have challenged the NCI cohort study (Marsh et al.,
29 2007a, b, 2002, 1996; Marsh and Youk, 2005). Marsh and coworkers (Marsh et al., 2007a) show
30 that subjectively assessed exposure to silversmithing is tentatively associated with NPC. Given
31 that there were no prior citations of an association between silversmithing exposures and NPC in
32 the medical literature and given the many post hoc reexaminations of alternative hypotheses to
33 explain the original NCI findings, it is more likely that silversmithing is an artifactual potential
34 confounder.
This document is a draft for review purposes only and does not constitute Agency policy.
4-189 DRAFT—DO NOT CITE OR QUOTE
-------
1 It may be expected that, without new approaches for obtaining more accurate and precise
2 estimates of exposure, further follow-up of current cohorts and future epidemiologic studies of
3 formaldehyde and NPC will face the same limitations and criticisms found with existing studies.
4 These limitations notwithstanding, the epidemiologic studies reviewed here represent what may
5 be currently discernable about a formaldehyde-NPC link in humans by using rigorous
6 observational methods. As such, concluding any influence of formaldehyde must be made on the
7 weight of all human and animal evidence in the face of known and expected limitations in study
8 design and exposure assessment.
9 The results of two well-designed cohort studies found a positive association between
10 formaldehyde-exposed professionals, such as pathologists, embalmers, and funeral directors, and
11 LHP cancer, particularly ML. The largest cohort study of formaldehyde has the most extensive
12 exposure assessment (Blair et al., 1986; Stewart et al., 1986), and the cohort was followed for a
13 median duration of 35 years (Hauptmann et al., 2003). By using cumulative exposure measures
14 not previously used and by using internal comparison groups, significant increases in the risk of
15 cancer of the LHP system, particularly ML, were reported. This study demonstrated that
16 formaldehyde was a risk factor for LHP cancers, independent of other risk factors, such as
17 benzene and smoking. Hauptmann et al. (2003) found statistically significant dose trends for
18 peak exposure and AIE. Pinkerton et al. (2004) also found a significant increase in the risk of
19 ML in garment workers 20 years after their initial exposure and in workers with 10 or more years
20 of exposure. Additionally, several studies of pathologists, embalmers, and other medical
21 workers reported greater numbers of observed deaths from leukemia than expected although
22 many studies of these groups suffer from a substantial HWE based on comparisons with external
23 death rates. Two of these studies, Hayes et al. (1990) and Stroup et al. (1986), also report a
24 significantly excess risk of ML in embalmers, funeral directors, and anatomists.
25 There is a range of biological plausibility for an agent whose primary action is at the
26 POE. Acute leukemias (ALL and AML), believed to arise from transformation of stem cells in
27 the bone marrow, are less plausible. In contrast chronic lymphatic leukemia, lymphomas,
28 multiple myelomas (from plasma B cells), and unspecified cancers may involve an etiology in
29 peripheral tissues to include cells, cell aggregates, germinal centers, and lymph nodes. An
30 association of these cancers to an exogenous agent acting at the POE is biologically plausible.
31 It is the conclusion of this assessment that the weight of the epidemiologic evidence at
32 this time supports a link between formaldehyde exposure and carcinogenicity in humans.
33
This document is a draft for review purposes only and does not constitute Agency policy.
4-190 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.2. ANIMAL STUDIES
2 This section discusses the available laboratory animal data on the toxicity of inhalation,
3 oral, and dermal exposures to formaldehyde. An extensive database of laboratory animal studies
4 is available for formaldehyde, including numerous 2-year bioassays by both the inhalation and
5 oral exposure routes. Although a large portion of the literature reports studies focused on toxic
6 effects at the site of contact or portal of entry (POE), general systemic effects as well as
7 neurobehavioral effects, reproductive and developmental effects, immunologic changes, and
8 sensitization are represented in the literature as well. The organization and general content of the
9 chapter is discussed below.
10 The first subject addressed for the animal studies is the occurrence of reflex bradypnea
11 (KB) observed in rodents exposed to reactive gases, including formaldehyde (see Section 4.2.1).
12 KB is a reduction in ventilation rate, minute volume, and other physiological parameters
13 experienced by rodents exposed to an irritant/reactive gas. Although humans and nonhuman
14 primates do not exhibit the same change in respiratory rate, these studies are included in order to
15 better understand the effects on KB in interpreting rodent studies presented in the balance of the
16 chapter. Additionally, although binding to the trigeminal nerve and subsequent downstream
17 events do not result the acute signs of KB in humans, the mechanism itself may play a role in
18 understanding other adverse health effects observed in humans including more subtle changes in
19 pulmonary function, sensitization, and sensory irritation.
20 The available data for the inhalation and oral exposures confirm direct formaldehyde-
21 induced toxicity in tissues present at the POE. These observations are consistent with the
22 physicochemical characteristics, reactivity, and metabolic pathways of formaldehyde as
23 discussed in Chapter 3. Indications of cell damage, cell proliferation, and inflammatory
24 responses are similar for each route of exposure, therefore effects at the POE for inhalation and
25 oral exposures are described first (see Sections 4.2.2 and 4.2.3, respectively). Given the well-
26 established nature of these health effects and the wealth of literature for inhalation exposures,
27 complete study summaries for respiratory tract effects are provided. Studies are organized by
28 study duration—acute, subchronic and chronic—where some of the chronic bioassays were
29 designed to address carcinogenic potential.
30 Although a majority of the oral and inhalation studies focus on health effects at the
31 POE—respiratory tract and GI tract—the general systemic toxicity of formaldehyde is addressed
32 where it was integral to the study. Therefore, body weight and organ weight changes, gross
33 pathology, organ histopathology outside of the POE, blood and urine chemistry, and other
34 biochemical measures may be included in these study summaries. An overview of general
35 systemic findings is provided in Section 4.4 for all routes of exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
4-191 DRAFT—DO NOT CITE OR QUOTE
-------
1 Studies addressing immune function, neurobehavioral effects, sensitization, and
2 reproductive and developmental effects are addressed across routes of exposure. The specialized
3 nature of these studies requires discrete treatment, and inclusion of data across routes of
4 exposure allows for a synthesis of the available information, to better understand the toxic
5 potential of formaldehyde exposure on these endpoints.
6
7 4.2.1. Reflex Bradypnea
8 Reflex bradypnea (RB), which is believed to be a protective response, is often observed
9 in rodents exposed to reactive gases. It is primarily characterized by marked decreases in
10 activity, respiratory rate, body temperature, and metabolic rate. RB is not seen in humans and
11 nonhuman primates. An understanding of the RB is important to the interpretation of many of
12 the animal bioassays examining formaldehyde-induced health effects. Of chief concern is that
13 the physiological effects of RB, described below, may interfere with appropriate interpretation of
14 adverse effects noted with formaldehyde exposure. It is important to distinguish between an
15 effect directly related to RB versus formaldehyde exposure. Additionally the effects of RB may
16 mask or alter formaldehyde-induced health effects. Secondly, differential respiratory effects of
17 RB due to species and strain will result in differential inhaled doses at the same exposure level.
18 This needs to be considered both when comparing the results of animal studies and in
19 extrapolation to humans. Finally, although humans do not experience RB, the mechanism of RB
20 as a reflex response to trigeminal nerve stimulation assists in understanding human health related
21 to localized and reflex responses due to trigeminal nerve stimulation.
22 Irritant gases have been shown to decrease body temperature, heart rate, and blood
23 pressure as well as alter blood chemistry in rodents (Pauluhn, 2003, 1996; Jaeger and Gearhart,
24 1982). Because of their small size, mice can rapidly lower their body temperatures and thus their
25 metabolic rate and ventilation rate. The hypothermia that results from RB can directly affect
26 nearly all biological processes (Gordon et al., 2008). Formaldehyde exposure can dramatically
27 lower ventilation rate and reduce body temperature in mice by as much as 4°C, and it has been
28 posited that decreased oxygen supply is likely to have profound effects on organisms with
29 substantial oxygen demands (Jaeger and Gearhart, 1982). The effects of RB are reversible,
30 though it may take several minutes to several hours to return to pre-exposure conditions
31 (Pauluhn, 1996; Jaeger and Gearhart, 1982).
32 The literature on sensory irritation is broad; many studies have investigated species
33 differences, dose response relationships, tolerance, and cross-tolerance to other sensory irritants
34 (see Tables 4-8 and 4-9). This discussion focuses on the changes in respiratory rate and minute
35 volume during formaldehyde exposure. Sensory irritation is often quantified as the statistically
This document is a draft for review purposes only and does not constitute Agency policy.
4-192 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
derived exposure concentration that results in a 50% reduction in respiratory rate (RD50) in
rodents (ASTM, 2000; Kane et al., 1979). Kane and Alarie (1977) evaluated various aspects of
sensory irritation, including establishing the RD50, exploring the reproducibility of response,
investigating the effect of tracheal cannulation, and determining the potential for tolerance with
repeated exposure or pre-exposure in male Swiss-Webster mice, caused by formaldehyde and
acrolein. The RD50 was established by exposing four mice for 10 minutes at each concentration
across a range representing approximately 10 to 80% reduction in respiration and calculated by
using least squares regression. The RDso and its 95% CI for formaldehyde were calculated to be
3.1 (2.1-4.7) ppm (3.8 [2.58-5.77] mg/m3). The tracheal cannulation experiments demonstrated
that the effect on respiratory rate was caused by URT sensory irritation.
Table 4-8. Respiratory effects of formaldehyde-induced reflex bradypnea in
various strains of mice
Species/strain
Male Swiss-
Webster mice
Male Swiss-
Webster mice
Male Swiss-
Webster mice
Male Swiss OF^
mice
Male B6C3F1 mice
Male B6C3F1 mice
Male C57BL6/F1
mice
No./
group
4
8
4
6
4
4
3
Treatment"
Duration: 10 minutes.
Exposure: up to 100 ppm.
Duration: 3 hours/day for
3 days. Exposure: 0.52, 0.44,
1.16, 1.83, 3.10, 5.35, 5.60, and
11. 2 ppm.
Duration: 10 minutes (head
only). Exposure: up to 10 ppm.
Single 5-minute exposure to
four unspecified
concentrations.
Duration: 10 minutes (head
only).
Exposure: Range up to 10 ppm.
Duration: 10 minutes (head
only). Exposure: up to 15 ppm
Pretreatment: 2, 6, or 15 ppm
6 hours/day for 4 days.
Whole-body exposure for up to
2 hours.
Respiratory effects
RD50a = 3.1ppm
(95% CI: 2. 1-4.7).
RD50 = 3. 4 ppm
(95% CI: 2.4-4.7).
RD50 = 3.2ppm
(95% CI: 2. 1-4.7).
RD50 = 5.3 ppm.
RD50 = 4.9 ppm
(95% CI: 3. 9-6.4).
Naive mice: RD50 = 4.4 ppm
(95% CI: 0.9-5.0)
Pretreated mice: PJ}50 =
4.3 ppm (95% CI: 3.4-5.5).
After 1.25 hours:
Tidal volume reduced by
33%; 68% reduction in
respiratory frequency; CO2
production reduced by
50%; percent; body
temperature dropped from
37.8 to 34.7°C.
Reference
Kane and Alarie
(1977)
Kane and Alarie
(1977)
Steinhagen and
Barrow (1984)
De Ceaurriz et al.
(1981)
Steinhagen and
Barrow (1984)
Chang et al.
(1981); Barrow et
al. (1983)
Jaeger and
Gearhart (1982)
15
"Exposure concentration that results in a 50% reduction in respiratory rate.
This document is a draft for review purposes only and does not constitute Agency policy.
4-193 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-9. Respiratory effects of formaldehyde-induced reflex bradypnea in
various strains of rats
Species/strain
Male Crl-CD rats
Male Wistar rats
Male F344 rats
Male F344 rats
Male F344 rats
Male Charles
Rivers CD rats
No./
group
4
4
4
4
ND
o
J
Treatment
30-minute nose-only exposure
to a range of formaldehyde
concentrations.
Duration: 10 minutes (head
only). Exposure: up to 56 ppm.
Pretreatment:2, 6, or 15 ppm
6 hours/day for 4 days
Single 10-minute head-only
exposure to a range of
concentrations.
Pretreatment: 15 or 28 ppm
formaldehyde or 10 ppm
chlorine.
10 minute exposure to acrolein
or acetaldehyde (head only).
Pre-exposed to formaldehyde at
15 ppm for 6 hours/day for
9 days.
Whole-body exposure for up to
2 hours.
Respiratory effects
RD50=13.8ppm.
RD50 = 10.0 ppm.
Naive rats: PJ}50 =13.1 ppm
(95% CI: 10.6-17.5)
Pretreated rats: RD50 = 10.8 ppm
(95% CI: 7.6-16.9)
Baseline PJ}50 = 3 1.7 ppm.
Pre-exposure to formaldehyde-
induced tolerance at 28 ppm
(PxD50 = 20.2 ppm) but not
15 ppm.
Pre-exposure to chlorine-induced
tolerance to formaldehyde
(PJ)50 ranged from 64.5 to
115 ppm, depending on
exposure duration).
Pre-exposure to formaldehyde-
induced tolerance:
Acetaldehyde (PJ)50 = 2,991 ppm
in naive versus 10,601 ppm in
preconditioned animals)
Acrolein (RD50 = 6 ppm in naive
versus 29.6 ppm in preconditioned
animals).
After 0.7 hours:
Tidal volume reduced by 22%;
20% reduction in respiratory
frequency; CO2 production
unaffected.
Reference
Cassee et al.
(1996)
Chang et al.
(1981); Barrow et
al. (1983)
Chang and
Barrow (1984)
Babiuk et al.
(1985)
Jaeger and
Gearhart (1982)
4
5
6
9
10
11
12
13
Across the literature there is fairly good agreement on RD50 values for various strains of
mice (see Table 4-8), ranging from 3.1 ppm in male Swiss-Webster mice to 4.9 ppm in male
B6C3F1 mice. Rats are less sensitive, with RD50 values ranging from 10 ppm in male Wistar
rats to 31.7 ppm in male F344 rats. No reported RD50 for female rodents exposed to
formaldehyde exists.
Jaeger and Gearhart (1982) evaluated the effect of formaldehyde on respiratory rate, tidal
volume, minute volume, carbon dioxide (CO2) production (exhaled to air) as a reflection of total
metabolism, and core body temperature in male Charles River CD rats and male C57BL6/F1
This document is a draft for review purposes only and does not constitute Agency policy.
4-194 DRAFT—DO NOT CITE OR QUOTE
-------
1 mice. Animals (three/concentration) were exposed to 15 ppm (18.4 mg/m3) formaldehyde for up
2 to 2 hours. Mice exhibited a greater decrease in respiratory frequency and minute volume
3 compared with the rats. CC>2 production and body temperature were also affected to a greater
4 extent in the mice (see Table 4-8). The authors postulated that the decreased body temperature
5 in mice would likely lead to decreased biologic action of formaldehyde in the tissue.
6
7 4.2.1.1. Tolerance
8 Tolerance is defined as an increase in the concentration required to elicit the same degree
9 of RB response and was evaluated by Kane and Alarie (1977). In the first set of experiments,
10 mice (four/concentration) were exposed 3 hours/day for 4 days at the concentration associated
11 with either a 30 or 50% decrease in respiratory frequency (specific concentrations not given)
12 (Kane and Alarie, 1977). Naive animals served as controls for each day. The maximum
13 response increased with each additional day of exposure, and the diminution of response that was
14 typically exhibited after 60 minutes of exposure in naive animals was markedly delayed. In the
15 second set of experiments, mice were exposed to a formaldehyde concentration at one-tenth the
16 RD50 (i.e., 0.3 ppm) 3 hours/day for 3 days. On the fourth day the animals underwent a similar
17 exposure protocol to identify the concentration that resulted in an RDso, following the above
18 protocol. No change in the RD50 was demonstrated. Both of these experiments indicate no
19 change in tolerance with either type of pretreatment in Swiss-Webster mice.
20 Chang and Barrow (1984) tested whether tolerance would develop in male F344
21 (CDF[F344]Crl/Br) rats exposed to formaldehyde. Exposure to formaldehyde at 15 ppm
22 (18.4 mg/m3) for 6 hours/day, 5 days/week failed to induce tolerance. However, tolerance was
23 observed following exposure to 28 ppm (34.4 mg/m3) formaldehyde for 4 days. The
24 concentration-response curve in these animals was significantly different than that of naive
25 animals, with an increase in the RDso estimate for this exposure duration from 31.7 to 70.2 ppm.
26
27 4.2.1.2. Cross-Species Differences in Inhaled Dose
28 Formaldehyde-induced RB lowers both respiratory rate and tidal volume and thus
29 reduces the inhaled dose of formaldehyde at a given exposure concentration. Chang et al. (1983)
30 and Barrow et al. (1983) evaluated the species differences and the effective inhaled dose between
31 rats and mice, since mice seem to be more sensitive to formaldehyde-induced RB and do not
32 exhibit tolerance as shown in F344 rats. Groups (four/concentration) of male F344 rats and male
33 B6C3F1 mice were exposed to formaldehyde concentration ranges of 6.2-48 ppm
34 (7.6-59 mg/m3) or 0.78-14.0 ppm (0.96-17.2 mg/m3), respectively, for 10 minutes. Pretreated
35 animals used in the tolerance experiments were exposed to formaldehyde at 2, 6, or 15 ppm
This document is a draft for review purposes only and does not constitute Agency policy.
4-195 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
(2.45, 7.36, or 18.4 mg/m3) 6 hours/day for 4 days prior to determination of the RD50 and
concentration response across the same ranges.
A concentration-dependent decrease in respiratory rate was seen in both naive and
pretreated rats during formaldehyde exposure. Tolerance (defined as a decrease in respiratory
rate followed by a subsequent return to control values) occurred after 4 minutes of exposure and
was more pronounced at concentrations above 4 ppm. Concentration-response relationships
were very similar for naive and pretreated rats, and the RD50s were similar for both groups
(naive = 13.1 ppm [95% CI: 10.6-17.5]; pretreated = 10.8 ppm [95% CI: 7.6-16.9]). In contrast,
naive or pretreated mice did not develop tolerance during exposures. An examination of
concentration-response relationships for mice showed similar RDso values (naive = 4.4 ppm
[95% CI: 0.9-5.0] and pretreated = 4.3 ppm [95% CI: 3.4-5.5]) compared with rats, although the
slopes of the concentration-response regressions were statistically different (see Figure 4-5).
B6C3F1 Mice
F 344 Rats
• Naive (6 ppm)
° Pretreated (6 ppm)
• NaTve (15 ppm)
D Pretreated (15 ppm)
0246810 120 240 360 0246810
Time (min.) during exposure
120
240
360
Figure 4-5. Formaldehyde effects on minute volume in nai've and
formaldehyde-pretreated male B6C3F1 mice and F344 rats.
Source: Redrawn from Chang et al. (1983).
Exposure of naive or pretreated rats resulted in an increased (compensatory) tidal
volume. However, the increase in tidal volume did not compensate entirely for the decrease in
ventilation rate and was only concentration dependent in pretreated rats. Comparison of tidal
This document is a draft for review purposes only and does not constitute Agency policy.
4-196 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
volume from naive and pretreated mice exposed to formaldehyde showed a slight increase in
naive animals but a decrease in pretreated ones. The effect of formaldehyde exposure on tidal
volume was concentration dependent in both groups of mice. These results indicate that tidal
volume does not compensate entirely for the decrease in respiratory rate and that the
compensation is slightly greater in rats than in mice.
These studies (Barrow et al., 1983; Chang et al., 1983) showed thatB6C3Fl mice sustain
RB, whereas F344 rats develop tolerance more readily both during exposure and with
pretreatment. Thus, these results suggest that the rat may be the more sensitive species for the
effects of inhaled formaldehyde due in part to the difference in sensitivity between mice and rats
as evidenced by an RDso of 4.9 versus 31.7 ppm and the ability of rats to develop tolerance while
mice appear to sustain RB. Barrow et al. (1983) used the results of these experiments to estimate
an inhaled dose equivalent to the exposure concentration of 15 ppm for the strains of mice and
rats used in the chronic formaldehyde bioassays by Kerns et al. (1983) and Monticello and
Morgan (1994) described in Section 4.1.2 as follows:
Inhaled dose (ug/min-cm ) =
HCHO concentration f/ug/L) x minute volume (L/min)
Nasal cavity surface area (cm2)
(5-1)
As shown in Table 4-10, because mice were observed to be able to decrease their minute
volume by approximately 75% as compared with 45% in rats, a twofold higher inhaled dose
would be expected in rats versus mice. This difference may be relevant to the increased
incidence of SCC in the nasal cavity seen in F344 rats when compared with B6C3F1 mice.
Table 4-10. Inhaled dose of formaldehyde to nasal mucosa of F344 rats and
B6C3F1 mice exposed to 15 ppm
Parameter
HCHO concentration (ug/L)
Minute volume (L/min)
URT surface area (cm2)
Inhaled dose (ug/min/cm2)
F344 rats
18.4
0.114
13.44
0.156
B6C3F1 mice
18.4
0.012
2.89
0.076
Source: Barrow et al. (1983).
This document is a draft for review purposes only and does not constitute Agency policy.
4-197 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
4.2.1.3. Cross-Tolerance
Cross-tolerance of chemically-induced reflex responses has been examined in several
systems in order to better understand the specificity and nature of the interaction of reactive
chemicals (such as formaldehyde with chlorine) with the trigeminal nerve involved in the KB.
Development of cross-tolerance to formaldehyde following preexposure to chlorine or to
chlorine following preexposure to formaldehyde was shown to be a function of the duration of
the pretreatment in male F344 rats (Chang and Barrow, 1984) (see Table 4-11). A 7-day
recovery period resulted in only a slight loss of cross-tolerance from a 4-day pre-exposure to
either chlorine or formaldehyde (data not shown). The cross-tolerance between formaldehyde
and chlorine demonstrated in the Chang and Barrow (1984) study suggests that these chemicals
may act via a common mechanism and may involve the trigeminal nerve. In rats, cross-tolerance
was induced after chlorine exposure but not after formaldehyde exposure, which suggests that
the trigeminal nerve may have different reactive sites that are differentially activated, depending
on the stimulus.
Table 4-11. Exposure regimen for cross-tolerance study
Pre-exposure
Formaldehyde
15 ppm, 6 hours/day
Iday
4 days
10 days
Pre-exposure
Chlorine
10 ppm, 6 hours/day
Iday
4 days
10 days
Chlorine RD50
FA-pretreated
22.6 ppm
16.8 ppm
64.5 ppm
Naive
10.9 ppm
Formaldehyde RD50
Cl-p retreated
64.5 ppm
66 ppm
115 ppm
Naive
31.7
Source: Chang and Barrow (1984).
Babiuk et al. (1985) evaluated the potential for formaldehyde pretreatment to cause cross-
tolerance with various other inhaled aldehydes, including acetaldehyde and acrolein. Male F344
rats were pretreated with 15 ppm (18.4 mg/m3) formaldehyde 6 hours/day for 9 days and
challenged on the 10th day with the second aldehyde for 10 minutes at various concentrations
(four rats/concentration) to establish an RD50. Exposure to acetaldehyde and acrolein, the two
smallest molecules in the series of aldehydes tested, resulted in cross-tolerance. The RDso and
This document is a draft for review purposes only and does not constitute Agency policy.
4-198 DRAFT—DO NOT CITE OR QUOTE
-------
1 its 95% CI for acetaldehyde were estimated at 2,991 (95% CI: 2,411-3,825) ppm in the naive
2 rats, and this was increased by approximately 3.5-fold to 10,601 (95% CI: 7,902-15,442) ppm in
3 the rats pretreated with formaldehyde. With acrolein, the RD50 increased approximately fivefold,
4 from 6.0 (95% CI: 3.5-18.1) ppm to 29.6 (95% CI: 15.6-93.0) ppm. Cross-tolerance with
5 formaldehyde has only been demonstrated with acetaldehyde, acrolein, and chlorine (Babiuk et
6 al., 1985; Chang and Barrow, 1984), suggesting that it is not a generalized phenomenon.
7 Whether the phenomenon of tolerance involves modulation of specific trigeminal nerve
8 receptors or whether it results from less specific chemical injury of the nasal mucosa has not
9 been determined. For example, different mechanisms lead to stimulation of the trigeminal nerve
10 and are likely to control the decrease in respiratory rate. In particular, acetaldehyde might
11 interact with sensory nerves via an amino group (Steinhagen and Barrow, 1984; Schauenstein et
12 al., 1977), whereas the receptor-binding site for formaldehyde and acrolein is believed to be a
13 thiol group. Furthermore, different binding sites exist on the trigeminal nerve for different
14 irritants (Nielsen, 1991). Thus, Bos et al. (1992) concluded that the data on tolerance or
15 "desensitization" versus "sensitization" (as defined strictly on the basis of the respiratory apneic
16 response) may be the result of adaptation or reversible/irreversible adverse changes. The
17 mechanisms underlying sensitization or desensitization are not well characterized.
18
19 4.2.1.4. Formaldehyde Binding and Activation of Trigeminal Nerve Afferent Activity
20 Kane and Alarie (1978) evaluated the effect of 11 combinations of acrolein and
21 formaldehyde on respiratory rate in outbred specific-pathogen-free male Swiss-Webster mice.
22 Exposure concentrations ranged from 0.12-8.97 ppm (0.28-21 mg/m3) for acrolein and
23 0.37-9.73 ppm (0.45-11.9 mg/m3) for formaldehyde. The data were evaluated using a simple
24 model of competitive antagonism. Comparing the observed and predicted responses indicated no
25 apparent differences, and paired t-tests showed no statistical significance. The authors concluded
26 that acrolein and formaldehyde acted at the same receptor site and acted as competitive
27 antagonists when exposure occurred simultaneously.
28 Kulle and Cooper (1975) investigated the effects of formaldehyde on trigeminal nerve
29 afferent activity in adult male Sprague-Dawley rats. The authors isolated both the ethmoid and
30 nasopalatine branches of the trigeminal nerve and recorded afferent signaling as electrical
31 activity while reactive gases (formaldehyde, ozone, and amyl alcohol) were passed through the
32 nasal passages of the anesthetized animals. The authors reported that both branches of the
33 trigeminal nerve responded similarly to all three chemicals, and they therefore conducted the
34 balance of their experiments on the nasopalatine branch of the nerve. Nerve response was
35 calculated as the difference between exposed and control activity, and the threshold for a positive
This document is a draft for review purposes only and does not constitute Agency policy.
4-199 DRAFT—DO NOT CITE OR QUOTE
-------
1 response was arbitrarily defined as an increase of 0.1 spikes per second. The sensory threshold
2 was determined by extrapolation from the measured nerve response to a range of formaldehyde
3 concentrations (0.5-2.5 ppm) or ozone (5.0-29 ppm) for an exposure duration of 2 minutes.
4 Amyl alcohol exposure (0.3-10.0 ppm) lasted for 25 seconds. Threshold was arbitrarily defined
5 as an increase of 0.1 spikes per second. The mean thresholds were 0.25 ppm for formaldehyde,
6 5.0 ppm for ozone, and 0.30 ppm for amyl alcohol, suggesting that the trigeminal nerve is highly
7 sensitive to formaldehyde and amyl alcohol compared with ozone exposure.
8 In a second set of experiments, Kulle and Cooper (1975) investigated the effects of
9 prolonged formaldehyde-exposure on the odor response to amyl alcohol. Rats were pre-exposed
10 to a series of amyl alcohol concentrations (0.3, 0.7, 1.0, 3.3, 6.7, or 10.0 ppm [1.08, 2.52, 3.6,
11 11.9, 24, or 36 mg/m3]) then a 1-hour continuous formaldehyde exposure (0, 0.5, 1.0, 1.5, or
12 2.0 ppm [0, 0.61, 1.23, 1.84, or 2.45 mg/m3]). There was a progressive decrease in odor
13 response to amyl alcohol with increasing stimulus of formaldehyde concentration (p < 0.01,
14 analysis of variance [ANOVA]). The response to formaldehyde concentration was described by
15 a power function Y = 0.741 x X1'47, where X is the formaldehyde concentration. The effects of
16 exposure to 2.0 ppm were similar, regardless of whether it was presented immediately as a
17 separate exposure or as the final concentration of a progressively increasing series. The response
18 to amyl alcohol did not fully recover within the 1-hour extended recovery period. Thus, it
19 appeared that the afferent function depression was not due to receptor adaptation or insufficient
20 time for formaldehyde diffusion away from receptor sites.
21 In an attempt to elucidate the basis of the differential effects of various types of
22 aldehydes on sensory irritation, Tsubone and Kawata (1991) recorded the afferent activity of the
23 surgically isolated ethmoidal nerve (a branch of the trigeminal nerve) during delivery of
24 0.32-4.7 ppm (0.39-5.77 mg/m3) formaldehyde, 0.18-7.2 ppm (0.41-16.5 mg/m3) acrolein, and
25 134-2,232 ppm (241-4,021 mg/m3) acetaldehyde into the cannulated URT of male Wistar rats
26 (six/aldehyde) at a flow rate of 200 mL/minutes for 22 seconds. Only one aldehyde was used in
27 each animal and each exposure was repeated two to four times at different concentrations. The
28 activity of the nerve was recorded as the number of electrical discharges for a total period of
29 100 seconds, including preinhalation (30 second), inhalation (22 second), and postinhalation
30 (48 second) periods. Nitrogen was used as the control gas and as the vehicle to dilute the
31 aldehyde gases in order to not interfere with the gas chromatography used to analyze the
32 exposures. The vapor concentrations associated with a 50% increase in nerve activity over the
33 level of control gas were calculated as approximately 1.8, 1.2, and 908 ppm for formaldehyde,
34 acrolein, and acetaldehyde, respectively. These results are consistent with the findings of
This document is a draft for review purposes only and does not constitute Agency policy.
4-200 DRAFT—DO NOT CITE OR QUOTE
-------
1 Steinhagen and Barrow (1984) and the hypothesis that the differences in RD50 are due to
2 differences in chemical reactivity in the tissue.
3 In summary, KB is a phenomenon observed in rodents exposed to reactive gases,
4 believed to be a protective response to the irritant properties of the gas. In comparative studies,
5 mice have a more pronounced KB response to irritant gases than rats and generally respond at a
6 lower concentration than rats. Interestingly, only rats appear to develop tolerance to irritant
7 gases, while mice sustain an KB response. When formaldehyde exposure is studied in concert
8 with other reactive gases like chlorine and other aldehydes like acetaldehyde and acrolein, cross-
9 tolerance developed. However, the mechanism underlying this response is unknown. It is
10 thought that KB may occur as a result of stimulation of the trigeminal nerve. Thus, although KB
11 appears to be a phenomenon specific to rodents, the mechanism by which it occurs, trigeminal
12 nerve stimulation, may be applicable to understanding MO As in other species, such as primates
13 and humans, particularly in regard to sensitization.
14
15 4.2.2. Respiratory Tract Pathology
16 The database for evaluating the POE toxicity in the respiratory tract of inhaled
17 formaldehyde is robust, with well-designed studies that span a duration range of a few hours to
18 chronic 2-year bioassays. Toxicity testing has been performed in various species, including
19 mice, rats, hamsters, guinea pigs, dogs, and nonhuman primates. Although a few studies include
20 examination of tissues outside of the URT, the majority of studies focus on changes in cell
21 proliferation and cell pathology in the nasal mucosa. Both mice and rats are well-defined animal
22 models with standard histologic sections established to evaluate various regions of the nasal
23 passages, divided into Levels 1 to 5 and illustrated in Figure 4-6. Pathology of the nasal mucosa
24 will be discussed with reference to these sections, and the region examined will be stipulated
25 (e.g., nasoturbinates, maxilloturbinates, or ethmoid turbinates [ETs]). Additionally, pathology of
26 the respiratory epithelium will be distinguished from effects on the olfactory epithelium,
27 although the nature of the lesions is similar.
28 Direct effects of formaldehyde exposure on mucociliary clearance are presented first—as
29 this may be the first interaction of the reactive chemical with elements of the upper respiratory
30 tract. A discussion of changes in cell proliferation follows, as increased cell proliferation at low
31 levels of exposure may be a more sensitive indicator of effects on the underlying epithelium.
32 This balance of this section summarizes studies that have investigated cellular pathology in the
33 URT and in the lung. Below, full study descriptions are provided for both short-term, subchronic
34 and chronic duration studies (including, where appropriate, how cell proliferation relates to the
35 observed formaldehyde-induced pathology).
36
This document is a draft for review purposes only and does not constitute Agency policy.
4-201 DRAFT—DO NOT CITE OR QUOTE
-------
1
2 Figure 4-6. Sagittal view of the rat nose (nares oriented to the left).
3
4 Note: The figure shows the normal distribution of nasal mucosae and the section
5 levels used in contemporary histopathology (Brenneman et al., 2000; Mery et al.,
6 1994). Sections 1, 2, 4, and 5 correspond to Levels I, II, III, and IV as proposed
7 by Young (1981). S = squamous, T/R = transitional/respiratory, O = olfactory
8 mucosa.
9
10 Source: Brenneman et al. (2000).
11
12
13 4.2.2.1. Mucociliary Clearance
14 The mucociliary apparatus of the URT is the first line of defense against airborne
15 toxicants. Comprising a thick mucus layer (epiphase), hydrophase, and ciliated epithelium, the
16 mucociliary apparatus may entrain, neutralize, and remove particulates and airborne chemicals
17 from inspired air (see Figure 4-7). The mucus serves to entrain or neutralize and remove
18 exogenous agents from the nasal epithelium (e.g., particles, reactive chemicals). As reviewed by
19 Kim et al. (2003), the nasal mucus contains proteins, glycoprotein, and lipids but is primarily
20 water (95%) and is propelled along by movement of the underlying cilia. Degradation in the
21 continuity or function of the mucociliary apparatus, which provides protection to the nasal
22 epithelium, would result in higher levels of gases and particles reaching the nasal epithelium
23 itself and greater penetration of chemicals into the respiratory tract. Therefore, breakdown and
24 disruption of mucociliary function are adverse effects, since a key bodily defense to exogenous
25 agents (including infectious agents) is damaged.
26
27
This document is a draft for review purposes only and does not constitute Agency policy.
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
OM
Figure 4-7. Main components of the nasal respiratory epithelium.
Note: OM = osmiophilic membrane; EP = epiphase; HY = hypophase; Cl = cilia;
MV = microvilli; CJ = cell junction; CC = ciliated cell; NCC = nonciliated cell;
GC = goblet cell; NE = nerve; GL = gland; BV = blood vessel; ECS =
extracellular space; BM = basement membrane.
Source: Morgan et al. (1986d).
Mucus flow slows upon formaldehyde exposure, despite an increase in the ciliary
beat of the underlying epithelial cells, which propel the mucus across the nasal epithelium
(Morgan et al., 1986a, c, d; 1983). These findings are consistent with other studies since
airborne pollutants and reactive gases have been shown to decrease mucus flow rates in
several animal models (Mannix et al., 1983; Iravani, 1974; Carson et al., 1966; Dalhamn,
1956; Cralley, 1942). In addition to slowing flow, the mucus layer has been observed
breaking up as it floats on the epiphase, creating gaps in the epiphase and revealing the
hydrophase below (Morgan et al., 1986c, d). Formaldehyde reacts with glycoproteins in
This document is a draft for review purposes only and does not constitute Agency policy.
4-203 DRAFT—DO NOT CITE OR QUOTE
-------
1 the mucus of the epiphase, creating cross-links between these large molecules; this is
2 believed to increase the viscosity of the mucus.
3 In their first experiments, Morgan et al. (1983) describe progressive mucostasis (slowing
4 of mucous flow) and ciliastasis (disruption of ciliary beat) with increasing days of exposure to
5 formaldehyde in male F344 rats (15 ppm; 6 hours/day for 1, 2, 4, or 9 days). Ciliastasis occurred
6 with greater frequency and across more regions of the nasoturbinate with subsequent days of
7 exposure. After 9 days, mucostasis was recorded in all but two regions evaluated. Although the
8 severity and time course of these changes varied across regions of the nose, the process followed
9 a similar pattern: decreased flow, increased ciliary action, mucostasis, and ciliastasis. Since the
10 formaldehyde-induced deficits in mucociliary function increased with days of exposure, activity
11 did not fully recover between exposures (18 hours) (Morgan et al., 1983). Therefore, the
12 severity and extent of adverse effects are dependent on both the concentration of exposure and
13 duration (in this case, days of repeated exposures).
14 In subsequent studies, Morgan et al. (1986c) examined the exposure-response
15 relationship of formaldehyde effects on mucociliary function and functional recovery 18 hours
16 after exposure ceased. Exposure regimens similar to the above experiment included additional
17 exposure concentrations (0.5, 2, and 6 ppm) and an additional time point of 15 days duration.
18 Exposure at 2 and 6 ppm resulted in the same progression of effects on mucus flow and ciliary
19 beat. Considering both severity and extent of effects a clear exposure-response relationship was
20 demonstrated. Additionally, within each exposure group, effects progressed both in severity and
21 extent by duration of exposure to formaldehyde (from 1 to 4, 9, and 15 days of exposure)
22 (Morgan et al., 1986c).
23 Flow and ciliary beat were not reduced, but rather increased, in epithelium from rats
24 exposed to 0.5 ppm formaldehyde. Mucus flow in 2 of 10 areas assessed was clearly increased
25 (275 and 200% of controls) after 4 days of exposure to 0.5 ppm formaldehyde. Two other
26 epithelial regions showed a similar trend (150% of controls), but this change was not statistically
27 significant. Interestingly, measurements made in corresponding areas after 9 days of exposure
28 did not show an increase, and measurements in one region were reduced to 37% of control.
29 Although it is not known whether the observed increase in mucus flow rate is a subtle indication
30 of an adaptive response to a low level irritant, the increase appears to be transient. It is not
31 known if flow rate would continue to decrease below control levels for repeated exposures at
32 0.5 ppm for longer than 9 days.
33 The regions affected at 15 ppm generally included the lateral aspects of the nasoturbinate
34 and both the dorsal and medial aspects of the maxilloturbinate. In general there was an anterior
35 to posterior effect with increasing concentration and time. Additionally, impaired mucociliary
This document is a draft for review purposes only and does not constitute Agency policy.
4-204 DRAFT—DO NOT CITE OR QUOTE
-------
1 function was more extensive with greater concentration and length of exposure. Nasal lesions
2 were seen on the nasal epithelium and correlated with those areas where some inhibition of
3 ciliary function was measured. Areas without mucus flow but that still retained ciliary function
4 did not develop epithelial lesions. Morgan et al. (1986c) reported "coagulated mucus," viewed
5 as a "continuous membrane" over the epithelium after 6 hours of exposure to 15 ppm
6 formaldehyde. Minor cell damage and infiltrating neutrophils and monocytes were also seen in
7 these areas. The coagulated mucus was not seen in similarly exposed rats that were allowed
8 18 hours of recovery before sacrifice. However, ciliated cells were damaged, and there was a
9 greater presence of neutrophils and macrophages (MPs) after this recovery period. The authors
10 noted that, as the exposure continued, these areas exhibited increased signs of inflammation and
11 epithelial damage, eventually resulting in "severe degenerative changes."
12 Morgan et al. (1986a) refined their study design to implement a nose-only exposure to
13 formaldehyde in order to better examine the progression of changes in mucociliary function
14 during short-term exposure, allowing examination of mucus flow immediately following
15 exposure. Three F344 rats/group were exposed to 15 ppm (18.4 mg/m3) formaldehyde for 10,
16 20, 45, or 90 minutes or 6 hours. Two groups of rats were exposed to 2 ppm to determine a no
17 effect level for 90 minutes or 6 hours. The extent and severity of mucostasis and ciliastasis seen
18 after a 6-hour 15 ppm (18.4 mg/m3) formaldehyde exposure and a 1-hour recovery period were
19 similar to the earlier study (Morgan et al., 1986a), indicating that similar exposure conditions
20 were reached with this nose-only apparatus. Ciliastasis and mucostasis were both less severe and
21 less extensive in a time-dependent manner and at the earlier time points of 10, 20, 45, and
22 90 minutes. Significant recovery was seen in mucociliary function by allowing a 1-hour
23 recovery between exposure and sacrifice. Regions of both the nasal septum and lateral wall,
24 which exhibited no mucus flow when examined immediately after a 6-hour exposure, had
25 measurable flow after the 1-hour recovery period. Similar recovery was seen at all durations of
26 exposure. No decreases in mucociliary function were seen after exposure for either 90 minutes
27 or 6 hours at 2 ppm formaldehyde. However, given evidence of recovery (Morgan et al., 1986a)
28 and the time taken to dissect and view the tissues ex vivo may have obscured more subtle effects.
29 To assess more immediate effects on mucociliary apparatus, Morgan et al. (1984a) have
30 examined formaldehyde effects on the mucociliary apparatus of isolated frog palates. This
31 system allowed observation of mucociliary function during exposure. Unexposed frog palates
32 were covered by a continuous sheet of mucus of variable thickness, which was observed to flow
33 in streams across the palette, exhibiting a wave-like form in some areas of the epiphase. The
34 authors reported particle movement in a lower, less viscous layer that was consistent with a less
35 viscous underlying hydrophase, similar to that described in rat mucosa. The basal mucus flow
This document is a draft for review purposes only and does not constitute Agency policy.
4-205 DRAFT—DO NOT CITE OR QUOTE
-------
1 rate was 0-4 mm/minute, with localized ciliary activity. Short periods of increased mucus flow
2 were associated with seemingly spontaneous increases in ciliary beat.
3 Formaldehyde exposure resulted in an initial increase in ciliary beat and mucus flow rate
4 in all palates exposed at 1.37, 4.36, and 9.58 ppm formaldehyde (but not 0.23 ppm). With
5 increasing formaldehyde concentration and time of exposure, mucostasis was evident as mucus
6 became stiff and eventually rigid. Ciliary beat continued after mucostasis was reached until
7 palates were exposed to 4.36 and 9.48 ppm formaldehyde, when ciliastasis was reached. The
8 time course to peak mucus flow rate, mucostasis, and ciliastasis was concentration dependent,
9 with mucostasis reached in less than 3 minutes at 9.48 ppm. In contrast, increased mucus flow
10 peaked at 8 minutes in palettes exposed at 1.52 ppm formaldehyde, which, though declining,
11 remained above basal levels after 25 minutes with no mucostasis or ciliastasis noted at this level.
12 F16-Neyret et al. (2001) demonstrated reduced mucociliary clearance and decreased
13 frequency of ciliary beats by using a similar isolated frog palette mucociliary apparatus.
14 However the palates were exposed by formaldehyde in the Ringer's solution in which the palates
15 were placed (0, 1.25, 2.5, or 5 ppm). Also, mucus was removed from the palettes and did not
16 come into direct contact with the formaldehyde. Despite these differences, formaldehyde caused
17 mucociliary clearance to decrease in a time- and concentration-dependent manner; mucostasis
18 occurred after 60 minutes of exposure to 5 ppm formaldehyde (see Figure 4-8). Ciliary beat was
19 decreased in a time-dependent manner at 2.5 and 5 ppm exposure but increased at 1.25 ppm
20 formaldehyde (see Figure 4-8). Reduced mucociliary clearance at 2.5 and 5 ppm was consistent
21 with the reduced ciliary beat. However, clearance decreased at 1.25 ppm formaldehyde, where
22 there was an apparent increase in ciliary beat. The authors suggest this may be a result of
23 disrupting the harmonic movement of the cilia, impairing effective mucociliary clearance. Based
24 on study results, the authors hypothesize that changes in ciliary beat, including excitation at
25 lower exposures, are likely to be a direct effect of formaldehyde on epithelial cells or other
26 cellular components of the mucosa.
27 In summary, numerous studies have identified impaired mucociliary clearance activity
28 associated with formaldehyde exposure (see Table 4-12). Although low-dose and short-term
29 exposures first increase ciliary beat, impaired mucus flow, slowed ciliary beat, and eventual
30 mucostasis and ciliastasis have been demonstrated in both in vivo and in vitro exposure systems.
31 These effects are both concentration and duration dependent and can be seen in as few as
32 15 minutes from exposure. Repeated inhalation exposures in rats indicate the effect does not
33 fully recovery in an 18-hour period between exposures, contributing to greater impairment over
34 extended periods of exposure.
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-206 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
£ 125-
5 100-
8 75-
S
-•-5.0 ppm
-•-2.5 ppm -.
A 1.25 ppm £
g 75H
CT
t 50
I
t 25
:=
O
15
30 45
Time (min)
60
15
30 45
Time (min)
-•-5.0 ppm
-•-2.5 ppm
-A-1.25 ppm
-O- Control
60
Figure 4-8. Decreased mucus clearance and ciliary beat in isolated frog palates
exposed to formaldehyde after 3 days in culture.
Source: F16-Neyret et al. (2001).
This document is a draft for review purposes only and does not constitute Agency policy.
4-207 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-12. Summary of formaldehyde effects on mucociliary function in the upper respiratory tract
Species
Male F344
rats
Male F344
rats
Male F344
rats
Isolated frog
palates
Isolated frog
palates
Na
10
6
3 per
group
Not
stated
4
Treatment
15 ppm formaldehyde
6 hours/day for 1, 2, 4,
or 9 days
0 0 5 2 6 or 15 ppm
formaldehyde
6 hours/day for 1,4,9,
or 15 days
15 ppm formaldehyde
for 10, 20, 45, or
90 minutes or 6 hours
0.23, 1.37, 4.36, or
9.58 ppm
formaldehyde
0 1 25 25 and 5 ppm
formaldehyde every 15
minutes for 60 minutes
Measure of
mucociliary
function
Mucus flow and
ciliary beat
Mucus flow and
ciliary beat and
histopathologic
analysis
Mucus flow and
ciliary beat
Mucus flow rates
and histopathology
Mucociliary
clearance and ciliary
beat
Summary of results by location
Mucostasis in regions 2, 3, 4, 5, and 8 for all rats after a single dose.
Mucostasis in all but two regions evaluated by day 9. Ciliastasis followed
mucostasis.
Flow or ciliary beat were increased at 0 5 ppm
AfterlAfter 1 day, slowed or halted mucociliary flow at 15 ppm after 6
hours.
After 9 days, slowed or halted mucociliary flow decreased or completely
stopped in all nasal regions evaluated.
Regions affected included lateral aspect of the nasoturbinate and dorsal and
medial aspects of maxilloturbinate.
Ciliastasis and mucostasis increased in a time- and concentration-dependent
manner, with maximal response at 6 hours.
Significant recovery was observed when a 1-hour recovery period occurred
between exposure and sacrifice.
Ciliary beat and mucus flow increased from baseline at 1.37, 4.36, and
9.58 ppm.
Over time, mucus became rigid, and Ciliastasis occurred
Ciliary beat decreased in a time-dependent manner at 2 5 and 5 0 ppm but
was increased at 1.25 ppm.
Mucostasis occurred after 60 minutes at 5 ppm.
Reference
Morgan et al.
(1983)
Morgan et al
(1986c)
Morgan et al.
(1986a)
Morgan et al.
(1984a)
F16-Neyret et
al. (2001)
& s
II
s ,
to
o
oo
N = number of animals in study.
fe
H
O
o
-------
1 Morgan et al. (1983) suggested that the initial stimulation of ciliary activity may be a
2 defensive response to the irritant gas, possibly indicating some penetration of formaldehyde to
3 the underlying epithelial cells. Later effects of mucostasis may be a result of cross-linking of
4 mucus glycoproteins by formaldehyde, creating a rigid mucus that is not able to flow even with a
5 rigorous ciliary beat. It is unknown if the eventual cessation of ciliary beat is a result of
6 compound-related effects on ciliated epithelium as formaldehyde diffuses through the mucus or
7 an indirect effect associated with mucostasis. However, in vitro experiments by F16-Neyret et al.
8 (2001) indicate that formaldehyde in solution, supporting isolated frog palates without mucus,
9 resulted in the same sequence of effects, including increased ciliary beat at the lowest exposure.
10 These data suggest a role of formaldehyde beyond its ability to form protein cross-links in
11 mucociliary proteins.
12
13 4.2.2.2. Cell Proliferation
14 Formaldehyde-induced cell proliferation has been demonstrated under range of exposure
15 conditions in vivo and in vitro as well (see Chapter 3). Formaldehyde-induced mitogenesis may
16 be a primary effect (as demonstrated in the in vitro work) or secondary to adaptive responses and
17 tissue remodeling (Swenberg et al., 1983). This section provides a comprehensive discussion of
18 formaldehyde effects on cell proliferation in the epithelial tissues in the respiratory tract. The
19 majority of the work discussed investigates cell proliferation with in vivo labeling of
20 proliferating cells, although additional methods, such as flow-cytometry, have been employed in
21 some instances.
22 Swenberg et al. (1986) conducted a series of experiments in rodents to assess cell
23 proliferation in the nasal mucosa after formaldehyde inhalation. Radiolabeled thymidine
24 [3H]-thymidine was injected intraperitoneally (IP.) into male F344 rats and B6C3F1 mice after
25 formaldehyde exposure to assess the extent of in vivo incorporation into proliferating cells. Two
26 hours later, animals were sacrificed and the nasal passages were fixed, embedded, and sectioned
27 to examine the nasal mucosa. Slides were exposed for 12 weeks and developed to identify cells
28 that incorporated the radiolabeled thymidine. The percentage of labeled cells, as indicated by the
29 presence of five or more grains over the nucleus, was determined by visual count. A total of
30 4,000 or 1,500 cells were counted per section for rats and mice, respectively.
31 The first set of studies reported by Swenberg et al. (1986) compared the dose response of
32 rats and mice. Animals were exposed to 0, 0.5, 2, 6, or 15 ppm (0, 0.61, 2.45, 7.36, or
33 18.4 mg/m3) formaldehyde 6 hours/day for 3 days. Tritiated thymidine for cell labeling was
34 injected 2 hours after the end of exposure. No change in the percentage of labeled cells was seen
35 after 0.5 or 2 ppm formaldehyde exposure. However, the nasal passages of rats exposed at 6 and
This document is a draft for review purposes only and does not constitute Agency policy.
4-209 DRAFT—DO NOT CITE OR QUOTE
-------
1 15 ppm showed 10- to 20-fold increases over controls in LI at level 2. A similar cell
2 proliferation response was seen in mice treated with 15 ppm formaldehyde, although no increase
3 over control was seen in mice exposed to 6 ppm formaldehyde. These findings are consistent
4 with other data that indicate rats are more sensitive to formaldehyde exposure than mice. This
5 may be due to differences in the reflex apneic response between the two species. As discussed in
6 Section 4.2.1.1, mice maintain decreases in minute volume in response to formaldehyde, which
7 results in a lower overall effective internal dose to the mice.
8 Comparing cell proliferation rates after 2 versus 18 hours of exposure, Swenberg et al.
9 (1986) found that the longer exposure duration gave twice the cell proliferation rates after
10 repeated exposures. Therefore, these researchers conducted a second dose-response study to
11 examine cell proliferation 18 hours after exposure instead of the shorter exposure duration. The
12 dose-response study varied dose as well as duration of treatment. Rats were exposed 6 hours/day
13 to either 0.5, 2, or 6 ppm (0.61, 2.45, or 7.36 mg/m3) formaldehyde over periods of 1, 3, or
14 9 days. Formaldehyde exposure at 0.5, 2, or 6 ppm for 1 day increased cell proliferation in the
15 nasal epithelium. However, these increases were transient, and cell proliferation was not
16 increased after 3 or 9 days of exposure to 0.5 ppm or 2 ppm formaldehyde. Although still
17 elevated after a 3-day exposure to 6 ppm formaldehyde, cell proliferation returned to control
18 values after 9 days of exposure to 6 ppm formaldehyde (Swenberg et al., 1986). Therefore,
19 although concentration is a major determinant of cell proliferation, duration of exposure also
20 influenced formaldehyde-induced cell proliferation in the nasal epithelium.
21 Swenberg et al. (1986) directly tested the effects of cumulative exposure versus
22 concentration for both mice and rats. Animals were treated with one of three regimens, resulting
23 in the same C x t product: 3 ppm x 12 hours, 6 ppm x 6 hours, or 12 ppm x 3 hours, each
24 exposure resulting in 36 ppm-hours. The animals were exposed once a day for either 3 or 9 days.
25 Tritiated thymidine was injected 18 hours after exposure to label of proliferating cells. Tissue
26 sections from levels 1 and 2 of the nasal passages were examined in each case, and the
27 percentage of cells labeled was reported as the percentage of proliferating cells (see Figure 4-9).
28 Cell proliferation at level 1 in the nasal cavity was much greater than at level 2 for all
29 C x t combinations of formaldehyde exposure in both mice and rats (see Figure 4-9). The
30 authors noted that level 1 is more anterior and lacks significant defense from the mucociliary
31 apparatus, which may account for the observed greater sensitivity to formaldehyde. At all C x t
32 exposure products, 3 days of exposure resulted in greater cell proliferation than 9 days of
33 exposure. This was true for both species and for both examined levels of the nasal cavity. The
34 decrease in cell proliferation by day 9 is consistent with data on rats labeled 18 hours
35 postexposure (Swenberg et al., 1986).
This document is a draft for review purposes only and does not constitute Agency policy.
4-210 DRAFT—DO NOT CITE OR QUOTE
-------
MICE
RATS
(a)
10 1
5'
MICE
RATS
CONTROL
3x12
6x6
12x3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
MOUSE
RAT
MICE
RATS
(d)
Figure 4-9. Effect of formaldehyde exposure on cell proliferation of the
respiratory mucosa of rats and mice.
Note: a and b are data following 3 days of exposure; c and dare for 9 days of
exposure, a and c are from level 1 (most anterior); b and d are from level 2.
[3H]-thymidine was administered 18 hours after the last exposure.
Source: Swenberg et al. (1986).
When comparing C x t exposures for a single species and location, the findings are more
complex. Cell proliferation in level 2 of the nasal passages appeared to be more dependent on
concentration than on duration or cumulative exposure, with the strongest response seen for
12 ppm formaldehyde in combination with the shortest exposure period, 3 hours (see Figure 4-9).
This pattern was observed in both rats and mice after 3 days of exposure and in rats after 9 days
of exposure. No increases in cell proliferation at level 2 were seen for any C x t combination in
mice after 9 days. In contrast, increases in cell proliferation at level 1 of the nasal passages were
not strictly concentration dependent. After a 3-day exposure, no clear differences were seen
among different C x t treatments for either mice or rats, suggesting cumulative exposure may be
This document is a draft for review purposes only and does not constitute Agency policy.
4-211 DRAFT—DO NOT CITE OR QUOTE
-------
1 the important metric. Therefore, it may be concluded that cell proliferation for level 1 of the
2 nasal passages, where there is less protection of the epithelium, is influenced by concentration,
3 time, and duration of exposure. Cell proliferation at level 2 appeared to be more dependent on
4 concentration than time of exposure (Swenberg et al., 1986).
5 Cassee and Feron (1994) reported a qualitative increase in histochemical staining for
6 proliferating cell nuclear antigen (PCNA) in the respiratory epithelium of the nasoturbinates,
7 maxilloturbinates, septum, and lateral wall at levels 2 and 3 of rat nasal passages after repeated
8 exposures to 3.5 ppm (4.29 mg/m3) formaldehyde 22 hours/day for 3 days. While no increases
9 were seen in olfactory epithelium, frank necrosis, squamous metaplasia, and hyperplasia of both
10 ciliated and nonciliated epithelium were noted at these section levels.
11 Quantitative cell proliferation studies have been conducted by several researchers in the
12 same laboratory (Reuzel et al., 1990; Wilmer et al., 1989; Zwart et al., 1988; Wilmer et al., 1987;
13 Woutersen et al., 1987) (see Summary Table 4-17). These studies build off of those of Swenberg
14 et al. (1986), who labeled proliferating cells with [3H]-thymidine in assessing cell proliferation
15 within the nasal mucosa. The studies, all performed in male albino Wistar rats and using a
16 similar experimental design, provide the basis for comparing different exposure levels and dose
17 regimens across studies. Wilmer et al. (1987) demonstrate a concentration-dependent increase in
18 cell proliferation after 3 days of repeated 8-hour exposures at 5, 10, or 20 ppm (6.13, 12.3, or
19 24.6 mg/m3) formaldehyde, regardless of continuous versus interrupted exposure conditions
20 (2.83, 8.87, and 19.8 versus 0.86% proliferation in controls). Similar trends were seen when the
21 repeated continuous exposures were extended for 4 weeks, but cell proliferation was not
22 maintained at the same levels. As observed by Swenberg et al. (1986), these results suggest that
23 duration of repeated exposures may be an important determinant of cell proliferation rates.
24 Woutersen et al. (1987) reported that the majority of the dose-dependent increases in cell
25 proliferation seen at section level 3 after 3 days of repeated 6-hour exposures to 10 and 20 ppm
26 (12.3 and 24.6 mg/m3) formaldehyde occurred in areas of the epithelium showing "clear
27 squamous metaplasia and hyperplasia." Cell proliferation rates in metaplastic epithelium of
28 29.5 and 33.2% were much higher than the 1.4 to 2.8% proliferation in the visibly unaffected
29 respiratory epithelium from rats exposed at 10 ppm formaldehyde. Although there was a slight
30 trend towards increased cell proliferation in the visibly unaffected epithelium of exposed animals
31 compared with unexposed controls, the majority of increased cell proliferation resulting from
32 exposure to 10 and 20 ppm formaldehyde was attributed to the metaplastic epithelium.
33 Similarly, dose-dependent increases in cell proliferation seen at level 3 after 3 days of
34 repeated 6-hour exposures at 0.3, 1, and 3 ppm (0.37, 1.23, and 3.68 mg/m3) formaldehyde
35 (p < 0.001) corresponded to focal basal cell hyperplasia and loss of cilia (Woutersen et al., 1987).
This document is a draft for review purposes only and does not constitute Agency policy.
4-212 DRAFT—DO NOT CITE OR QUOTE
-------
1 No necrosis or focal erosion was noted at these levels of formaldehyde exposure. Cell
2 proliferation was not sustained at this location, and no lesions were noted after 13 weeks of
3 repeated 6-hour exposures. The authors hypothesized that defensive mechanisms, such as the
4 mucociliary apparatus, may have provided greater protection of the mucosa at level 3. Swenberg
5 et al. (1986) drew a similar conclusion when evaluating extended exposures, suggesting that
6 more posterior sections had a greater adaptive ability than those anterior sections with little
7 mucociliary function. Both Woutersen et al. (1987) and Swenberg et al. (1986) reported
8 sustained cell proliferation and development of lesions in the more anterior cross section.
9 Repeated exposures to 3 ppm formaldehyde (6 hours/day) resulted in significant increases in cell
10 proliferation in the epithelial cells at level 2, with accompanying disarrangement, focal
11 hyperplasia, and squamous metaplasia (Woutersen et al., 1987). Although no cell death was
12 observed at level 2 when viewed by light microscopy, "strongly indented and disarranged nuclei"
13 were seen by electron microscopy, which may be consistent with apoptosis (Woutersen et al.,
14 1987). However, later work in the same laboratory indicated no increased cell proliferation at
15 levels 2 or 3 in male Wistar rats exposed to formaldehyde at 1 or 2 ppm (1.23 and 2.45 mg/m3)
16 (8-hour repeated exposures for 3 days or 13 weeks) and only minimal response in rats exposed at
17 4 ppm formaldehyde (interrupted 8-hour exposures for 3 days or 13 weeks) (Wilmer et al.,
18 1989).
19 Reuzel et al. (1990) published the only report in which formaldehyde effects on cell
20 proliferation were studied for longer daily exposure durations: 22 hours/day versus
21 6-8 hours/day. Male Wistar rats were exposed to formaldehyde, ozone, or the combination of
22 the two 22 hours/day for 3 consecutive days. The concentrations of formaldehyde were 0.3, 1.0,
23 or 3.0 ppm (0.37, 1.23, or 3.68 mg/m3). Rats were injected with [3H]-thymidine 2 hours rather
24 than 18 hours after the last exposure. Cell proliferation was quantified by enumerating the
25 percentage of labeled cells in fixed and stained tissue sections. Cell proliferation on the
26 nasoturbinates, maxilloturbinates, lateral wall, and septum at levels 2 and 3 were quantified and
27 reported separately. Cell proliferation was increased at all locations in level 2 at 3 ppm
28 formaldehyde exposure (p < 0.05) but not at 0.3 or 1 ppm exposures (see Summary Table 4-17).
29 Whereas proliferation of cells in the nasoturbinate, maxilloturbinate, and septum was nearly
30 undetectable in control animals, 4, 5, and 3% proliferation was reported after repeated 22-hour
31 exposures to 3 ppm formaldehyde. Basal proliferation in the lateral wall was greater than in
32 other areas, approximately 1% increasing to 6% after exposure to 3 ppm formaldehyde.
33 Although basal levels of cell proliferation were slightly higher in all areas of level 3,
34 formaldehyde had no significant effects on cell proliferation in the level 3 areas evaluated. There
35 was a slight trend for increases at 3 ppm, but all proliferation rates were below 1%. Exposure to
This document is a draft for review purposes only and does not constitute Agency policy.
4-213 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
3 ppm formaldehyde also damaged the respiratory epithelium at levels 2 and 3, where cell
disarrangement and hyperplastic and metaplastic lesions were reported.
Roemer et al. (1993) investigated the effects of formaldehyde exposure on cell
proliferation in the trachea and lung in addition to nasal mucosa. Male Sprague-Dawley rats
were exposed head only to 2, 6, or 20 ppm (2.45, 7.36, or 24.5 mg/m3) formaldehyde 6 hours/day
for either 1 or 3 days. Proliferating cells were labeled with 5-bromodeoxyuridine (BrdU), the
label injected 16-22 hours after formaldehyde exposure ended. Free lung cells were harvested
by tracheal lavage, and the majority of isolated cells were MPs (>97%). Epithelial cells were
isolated from the nasal and tracheal mucosa by dissection, physical disaggregation, and enzyme
treatment to release epithelial cells. All cells were fixed and stained with fluorescent dyes to
detect BrdU and total DNA. Flow cytometry was used to determine the percentage of BrdU-
labeled cells as a measure of cell proliferation. Cells undergoing unscheduled DNA synthesis
(e.g., DNA repair) were excluded by cell cycle analysis.
The proportion of BrdU-labeled cells from the nose and trachea increased two- to
threefold above control values after a single 6-hour exposure to formaldehyde (see Table 4-13).
The lowest effective dose for increased cell proliferation was 2 ppm for nose and tracheal cell
proliferation (p < 0.05). However, increased proliferation in the nasal mucosa at the lowest dose
was transient, returning to control levels after a 3-day exposure. Cell proliferation remained
increased in the nasal mucosa after exposure to 6 or 10 ppm (7.36 or 12.3 mg/m3) formaldehyde
for 3 days. In contrast, proliferation of tracheal cells appeared to be reduced as a result of a
3-day exposure to 2 or 6 ppm formaldehyde. A similar trend was seen in free lung cells, but the
differences were not statistically significant.
Table 4-13. Cell proliferation in nasal mucosa, trachea, and free lung cells
isolated from male Wistar rats after inhalation exposures to formaldehyde
lDaya
Nose
Trachea
Lungd
3 Days
Nose
Trachea
Lung
Control
1.3b
1.2
1.8
Control
1.3
1.2
1.8
2 ppm
2.4C
3.1C
2.6
2 ppm
1.4
0.3C
2.2
6 ppm
3.7C
2.1°
3.3
6 ppm
2.5C
0.6C
2.4
20 ppm
2.7
2.8
3.1
20 ppm
2.3C
2.5C
5.1
"Exposures were 6 hours/day.
Proliferation is measured as the percent of BrdU-labeled cells.
Statistically different from controls (p < 0.05).
dThe majority of free lung cells were MPs (97%).
Source: Roemer et al. (1993).
This document is a draft for review purposes only and does not constitute Agency policy.
4-214 DRAFT—DO NOT CITE OR QUOTE
-------
1 The flow cytometry employed by Roemer et al. (1993) allowed for subtle changes in
2 proliferation rates to be measured with good discrimination. However, the method of cell
3 isolation did not allow examination of proliferation rates in discrete regions of the mucosa,
4 which may have attenuated the magnitude of the response. Additionally, proliferation rates
5 represent a mix of cell types that were not separated in this analysis, making the findings difficult
6 to interpret. This may be especially noteworthy in the free lung cells that were reportedly
7 primarily MPs.
8 Monticello et al. (1990) investigated whether changes in cell proliferation rate correlated
9 with areas of cell injury or with areas that developed tumors due to formaldehyde exposures by
10 using a unique metric of cell proliferation. They hypothesized that treatment-related effects on
11 cell populations could influence the apparent cell proliferation measured as LI, even though no
12 proliferative effect had occurred. For example, cell death could give an apparent increased
13 proliferation as a LI (% cells proliferating) by reducing the total number of cells present. This
14 would be especially true for a stratified epithelium, where the number of basal cells in active
15 proliferation may not change but cells above the basal layer might die or slough off, thereby
16 reducing the overall number of population of cells counted. The unit length labeling index
17 (ULLI) metric was developed to normalize proliferation rates against length of basal membrane
18 rather than cell population. However, application of a ULLI to the pseudostratified epithelium of
19 the nasal mucosa introduced additional complexities. First, undamaged mucosa has a single
20 layer of epithelial cells that have the capability for cell proliferation. Second, cells only become
21 layered in response to cell damage as a protective measure. Therefore, the total cells present and
22 the linear cell density should be considered, as well as the number and density of proliferating
23 cells, in developing an understanding of the proliferative response of these tissues to toxic insult.
24 Monticello et al. (1990) directly compared the apparent effects of formaldehyde exposure
25 on cell proliferation when quantified as an LI or as a ULLI. Male F344 rats were divided into
26 groups (n = 6) and exposed to 0, 2, 6, or 15 ppm (0, 2.45, 7.36, or 18.4 mg/m3) formaldehyde
27 6 hours/day, 5 days/week for 12 weeks. Rats were administered [3H]-thymidine continuously for
28 the last 5 days of exposure by surgically implanted osmotic pumps. After sacrifice, nasal
29 passages were fixed, and sections from standard level 3 were prepared for examination. Cell
30 proliferation was quantified at the midseptum and the lateral meatus at this level. Basement
31 membrane length, total number of cells present, and number of labeled proliferating cells were
32 recorded for each location. Each of these areas also was scored for the presence of nasal lesions.
33 The formaldehyde-related lesions included epithelial hyperplasia, squamous metaplasia,
34 and acute inflammation. These lesions were most severe in animals exposed to 15 ppm, mild at
35 6 ppm, but absent at 2 ppm. Cell proliferation, measured either as LI or ULLI, was increased in
This document is a draft for review purposes only and does not constitute Agency policy.
4-215 DRAFT—DO NOT CITE OR QUOTE
-------
1 the level 3 septum and lateral meatus after 13 weeks of exposure to 15 ppm formaldehyde but
2 not to 6 or 2 ppm (see Table 4-14). There was a slight increase in both cell number and labeled
3 cells in the lateral meatus of rats exposed to 6 ppm formaldehyde, but both measures of
4 proliferation were unchanged from controls. The increased proliferation in the lateral meatus at
5 15 ppm was entirely due to an increased number of labeled cells. Total cells were unchanged at
6 15 ppm; therefore, both LIs demonstrated a similar increase over control. In addition to
7 increased labeled cells in the septum at 15 ppm, total cells were increased from 470 to 640
8 (p< 0.05). Where the total cells and linear cell density were increased, the ULLI was
9 proportionally increased over the LI. These observations are consistent with the development of
10 squamous metaplasia and hyperplasia seen at 15 ppm. However, while both LI and ULLI
11 showed an eightfold increase in cell proliferation in the lateral meatus, they gave different results
12 in the septum where cell number was increased by formaldehyde treatment. LI increased 19-fold
13 and ULLI 25-fold with repeated exposures to 15 ppm formaldehyde. Although these data are
14 based on only 5-6 animals/group, and only in an extended study, the results suggest that the
15 ULLI and LI may not be proportional under all conditions studied. In similar experiments the LI
16 and ULLI provided different indices of proliferation in the olfactory epithelium after methyl
17 bromide exposure (Monticello et al., 1990). Methyl bromide exposure decreased cell
18 number/mm of basement membrane in a time-dependent manner, and the LI and ULLI were not
19 proportional across these changes. At day 3 there was an increase in labeled cells but a decrease
20 in total cells; therefore, the LI was increased greater than 20-fold, where the ULLI was only
21 increased eightfold. The authors endeavored to explain why the ULLI and LI yielded different
22 findings. Where ULLI is a more time-efficient method of assessing cell proliferation, the authors
23 suggested that representative areas should be quantified by LI to better understand the nature of
24 increased ULLI.
25 Monticello et al. (1990) reported similar results in a contemporary abstract; although
26 treatment groups were slightly different than in the above experiments, the findings were similar.
27 Rats were exposed to 0, 0.7, 2, 6, 10, or 15 ppm (0, 0.86, 2.45, 7.36, 12.3, or 18.4 mg/m3)
28 formaldehyde 6 hours/day for 4 days, 6 weeks, or 3 months. ULLIs were determined in the
29 septum and lateral meatus (methods not detailed). It is not stated whether [3H]-thymidine
30 labeling was carried out by injection or continuous infusion. Significant increases in cell
31 proliferation were reported after repeated exposures to 6, 10, and 15 ppm for 4 days and 6 weeks.
32 After 3 months of exposure, cell proliferation was still increased in rats exposed to 10 and
33 15 ppm formaldehyde. The authors noted that, although increased cell proliferation was seen at
34 earlier time points, sustained increased cell proliferation was only seen at 10 and 15 ppm, which
35 they considered the clearly carcinogenic doses.
This document is a draft for review purposes only and does not constitute Agency policy.
4-216 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
Table 4-14. The effect of repeated formaldehyde inhalation exposures for
3 months on cell count, basal membrane length, proliferation cells, and two
measures of cell proliferation, LI and ULLI, in male F344 rats
Lateral meatus
Total cells
BM length (mm)b
Cells/mm BM
Labeled cells
LI
ULLI
Septum
Total cells
BM length (mm)
Cells/mm BM
Labeled cells
LI
ULLI
Formaldehyde exposure level (6 hours/day, 5 days/week for 3 months)
0 ppm
1,800 ± 100
12.7 ±0.6
150±5
130 ± 10
7.2%c
10.2 cells/mmd
470 ± 20
2.9±0.1
160 ± 10
20 ± 1
4.3%
6.9 cells/mm
2 ppm
1,800 ± 150
11.9±0.5
150 ± 10
130 ±20
7.2%
10.9 cells/mm
460 ± 30
2.7±0.1
170 ± 10
40 ± 10
8.7%
14.8 cells/mm
6 ppm
2,300a± 1700
13.4 ±0.3
170 ± 10
210±30
9.1%
15.7 cells/mm
470 ± 20
2.9±0.1
160 ±3
10±2
2.1%
3.45 cells/mm
15 ppm
1,900 ± 160
11.6±0.7
150 ±5
1,400 ± 130
73.7%
120.7 cells/mm
640a ± 20
2.9 ±0.1
220a± 10
250 ±50
39%
86.2 cells/mm
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
"Different from control, p < 0.05.
bBM is basal membrane length in mm.
Calculated from group averages: LI = (labeled cells)/total cells.
dCalculated from group averages: ULLI = (labeled cells)/BM length.
Source: Monticello et al. (1990).
Monticello et al. (1991) applied the ULLI measurements in evaluating formaldehyde
effects on cell proliferation after short-term and subchronic repeated exposures. Six male F344
rats/group were exposed to 0, 0.7, 2, 6, 10, or 15 ppm (0, 0.86, 2.45, 7.36, 12.3, or 18.4 mg/m3)
formaldehyde 6 hours/day for 1, 4, or 9 days and for 6 weeks, using a 5 days/week regimen.
Rats were injected with [3H]-thymidine 18 hours postexposure to label proliferating cells. All
animals were sacrificed 2 hours later. Nasal passages were fixed, and sections from levels 2 and
3 were prepared for examination. Cell proliferation was quantified for three locations in level 2
(specifically, the lateral meatus, midseptum, and medial aspect of the maxilloturbinate) and for
two regions of level 3 (the lateral wall and midventral septum). Each of these areas also was
scored for the presence of nasal lesions.
This document is a draft for review purposes only and does not constitute Agency policy.
4-217 DRAFT—DO NOT CITE OR QUOTE
-------
1 As discussed above, proliferating cells were visually identified by the number of grains
2 over the nucleus, 10 grains indicating a proliferating cell. Cell proliferation was quantified as the
3 number of proliferating cells per length of basement membrane (cells/mm) and reported as a
4 ULLI. The report does not indicate the length of membrane viewed for each section as an
5 indication of how representative the counts are for each region. Lesions associated with
6 formaldehyde exposure may change the density of cells/mm of basement membrane (Monticello
7 et al., 1990). Areas of disarranged cells, erosion, metaplasia, or layering of epithelial cells may
8 exhibit different cell profiles. These processes would alter cell density, and therefore the ULLI,
9 independent of differential proliferation rates. As such, it is not expected to be proportional to
10 cell proliferation rates across conditions that have the potential to change cell density
11 (Monticello etal., 1990).
12 No formaldehyde-induced epithelial lesions or increases in the ULLI were seen in rats
13 exposed to 0.7 or 2.0 ppm formaldehyde, regardless of duration (see Table 4-15).
14 Formaldehyde-induced lesions were present in all regions of the nasal epithelium after exposures
15 to 10 and 15 ppm formaldehyde, regardless of duration (Monticello et al., 1991). Incidence and
16 severity of the lesions increased with concentration and duration of treatment and were
17 correlated to areas with increased cell proliferation. Rats exposed to 6 ppm formaldehyde
18 developed lesions in the level 2 nasal passages, where the ULLI was clearly elevated, but not in
19 the deeper level 3 passages. For example, no formaldehyde-related lesions were seen at the
20 lateral meatus and septum of level 3 at 1, 4, and 9 days of repeated exposure at 6 ppm, although
21 cell proliferation was increased. This transient increase in ULLI returned to near-control levels
22 after 6 weeks of repeated exposure (see Table 4-15). Monticello et al. (1991) suggested that cell
23 proliferation is a more sensitive indicator of cellular response and not necessarily dependent on
24 cellular necrosis.
25 The sustained cell proliferation at the lateral meatus and midseptum in rats exposed to 10
26 and 15 ppm formaldehyde, locations where SCCs are known to arise, supports a role for
27 compensatory cell proliferation in tumor development. However, Monticello et al. (1991) noted
28 that regional differences in sustained cell proliferation do not always correspond to the
29 occurrence of nasal tumors, primarily SCCs, in formaldehyde-exposed rats. Where sustained
30 cell proliferation has been demonstrated in the medial maxilloturbinate (MMT) at level 2
31 (Monticello et al., 1991), SCCs have not been found to originate in this area at similar exposures
32 (Monticello et al., 1996; Woutersen et al., 1989). Monticello et al. (1991) suggested that the
33 findings of Bermudez and Allen (1984), indicating that the epithelial cells of the maxilloturbinate
34 are more resistant to the genotoxic effects of DEN, support the possibility that differences in
35 regional tissue susceptibility may contribute to site specificity of formaldehyde-related SCCs.
This document is a draft for review purposes only and does not constitute Agency policy.
4-218 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-15. Formaldehyde-induced changes in cell proliferation (ULLI) in
the nasal passages of male F344 rats exposed 6 hours/day
Location3
Level 2: lateral meatus
1 day
4 days
9 days
6 weeks
Level 2: midseptum
1 day
4 days
9 days
6 weeks
Level 2: medial
maxilloturbinate
1 day
4 days
9 days
6 weeks
Level 3 : lateral meatus
1 day
4 days
9 days
6 weeks
Level 3 : midseptum
1 day
4 days
9 days
6 weeks
Exposure concentration
0 ppm
2.16
1.46
1.44
0.91
1.08
1.03
1.09
0.41
2.49
1.36
1.38
1.02
1.83
1.10
1.36
0.98
3.02
2.81
1.68
2.18
0.7 ppm
1.31
1.37
1.20
0.88
1.01
0.97
0.80
0.24
1.75
1.54
0.80
1.21
1.72
1.27
1.40
0.91
1.74
3.09
1.06
1.54
2 ppm
2.36
1.72
1.73
1.36
1.69
0.67
0.97
0.68
2.81
1.09
1.48
1.11
2.46
1.09
1.74
0.86
2.39
1.43
1.43
2.57
6 ppmb
16.9b
30.5b
23.5b
14.4b
3.85
10.0b
10.9b
2.10
18.15b
25.03b
22.54b
16.32b
7.53b"c
8.77b',c
7.35b' c
2.08
4.20
9.22b',c
9.50Y
2.58
10 ppmb
11. 2b
20.9b
28.6b
23.9b
17.9b
26.1b
19.6b
21.4b
5.9
20.3b
21.0b
26.1b
14.5b
20.0b
30.6b
24.2b
24.4b
18.7b
28.6b
14.0b
15 ppmb
12.7b
25.8b
24.6b
28.7b
16.7b
29.1b
29.1b
25.9b
5.3
19.4b
28.7b
25.1b
16.4b
30.8b
40.4b
34.8b
19.3b
34.4b
32.5b
27.5b
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
aULLI is expressed as the number of labeled cells/mm of basement membrane.
blndicates significantly different from control, p < 0.05.
Indicates a location where epithelial lesions were not seen by light microscopy.
Source: Monticello et al. (1991).
Monticello et al. (1996) further explored the correlation between measures of cell
proliferation and tumor site by modifying the ULLI to take into consideration the total number of
cells in a region that may be subject to increased cell proliferation. The population weighted
ULLI (PWULLI) is the product of the expected number of cells on a three-dimensional surface
in the nasal mucosa and the ULLI of a cross section of that surface. For this series of
experiments, six male F344 rats/group were exposed to 0, 0.7, 2, 6, 10, or 15 ppm (0, 0.86, 2.45,
7.36, 12.3, or 18.4 mg/m3) formaldehyde for up to 24 months with interim sacrifices at 3, 6, 12,
This document is a draft for review purposes only and does not constitute Agency policy.
4-219 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
and 18 months. Before each interim sacrifice [3H]-thymidine was continuously injected for the
last 5 days of exposure through a surgically implanted pump. Nasal passages were prepared, and
six standard sections were taken and developed as above for [3H]-thymidine-labeled cells.
Stained tissue sections were viewed in order to map all nasal tumors. A ULLI was determined
for each region (details not provided). The total cell population of each nasal region was
estimated from control animals sacrificed at 3 months (see Table 4-16). Cell profiles were
counted across 0.5 mm of basement membrane length at two locations for each region (site not
specified). Total cells per region were estimated from these counts and the modeled surface area
expected in each region (Fluid Dynamics Analysis Package version 7.0). It is unclear if one or
more rats were used to quantify cell population. Cell counts and variability were not reported.
Table 4-16. Cell population and surface area estimates in untreated male
F344 rats and regional site location of squamous cell carcinomas in
formaldehyde-exposed rats for correlation to cell proliferation rates
Nasal region
Anterior lateral meatus
Anterior midseptum
Anterior dorsal septum
Anterior medial maxilloturbinate
Posterior lateral meatus
Posterior midseptum
Maxillary sinus
Region not specified0
Total cells
(number)3
976,000
184,000
128,000
104,000
508,000
190,000
884,000
-
Area
(mm2)b
59.5
10.5
3.84
7.63
38.1
10.8
38
-
Cell density
(cell/mm2)
16,400
17,500
33,300
13,600
13,300
17,600
23,300
-
SCC incidence0
lOppm
12
0
0
0
2
0
0
6
15 ppm
17
1
o
5
4
9
1
0
25
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
aTotal cell number determined in unexposed rats as a product of representative cell counts and expected surface area
of the region.
bModeled surface area of the defined region by FDIP version 7.0.
°The number of animals bearing a tumor located in the region. Animals were exposed 6 hours/day for 24 months
prior to sacrifice.
Source: Monticello et al. (1996).
ULLIs were quantified by region of the nasal passages in order to correlate with regional
localization of tumors. For example, the anterior midseptum included cells from the midseptum
from approximately standard section levels 2 to 3. An anterior to posterior pattern of
formaldehyde effects, especially differences in cell proliferation rates, has been well established.
As such, cell proliferation rates would be expected to vary across the nasal regions used in this
This document is a draft for review purposes only and does not constitute Agency policy.
4-220 DRAFT—DO NOT CITE OR QUOTE
-------
1 analysis. Areas considered to possibly be preneoplastic were not quantified for this work.
2 Monticello et al. (1996) reported increased ULLIs in the ALM and the MMT of rats exposed to
3 10 or 15 ppm at all time points (3, 6, 12, and 18 months) but provided no indication of variability
4 or a statistical analysis, making it difficult to determine where true differences may exist. Some
5 caution should be used in interpreting the ULLI counts assigned for each region.
6 The PWULLI was calculated by multiplying the reported ULLIs by the calculated cell
7 populations by region. SCC incidence by region had a greater correlation to the calculated
8 PWULLI than the ULLI, R2 = 0.88 versus R2 = 0.46, respectively. The authors noted that the
9 relative lack of correlation with the ULLI was influenced by findings at the maxilloturbinate
10 where cell proliferation was high but SCC incidence was low. Other tumor types were not
11 included in the analysis (polypoid adenomas, adenocarcinomas, and rhabdomyosarcomas).
12 Additionally, 54 of the SCC tumors could not be accurately localized and were excluded from
13 the analysis, resulting in exclusion of 30 and 39% of animals with SCCs in the 10 and 15 ppm
14 treatment groups, respectively. The authors cautioned that the absence of these data might have
15 skewed the regional analysis of tumor location. Although the purpose of weighting the ULLIs
16 by total population of cells available in each region is to better represent the chance of a tumor
17 arising in each region, the cancer incidence was represented by the number of animals, not the
18 number of tumors, per region. Based on the exclusion of location data (up to 40% of the
19 animals), lack of variability and significance reported for the ULLI for cell counts, and SCC
20 incidence considered by animal rather than by tumor, the significance of a greater correlation by
21 PWULLI versus ULLI is of questionable value.
22 Monticello et al. (1989) also assessed formaldehyde-induced cell proliferation and
23 regional site location of lesions in the respiratory tract of rhesus monkeys (see Section 4.2.2.2,
24 Figures 4-11 and 4-12 for a full study description). LIs from the histoautoradiograms indicated
25 increased cell proliferation in transitory, respiratory, and olfactory epithelial cells after the 6-
26 week formaldehyde exposure. Similar trends were seen after only 1 week but were statistically
27 significant only in the respiratory epithelium. Although increased proliferation in the trachea and
28 carina was statistically significant after 1 week of exposure, the greater increases seen after
29 6 weeks of exposure were not statistically significant. A small sample size (n = 3) and high
30 variability may have contributed to the lack of statistical significance. The authors noted that
31 increased cell proliferation was seen in locations with minimal histologic changes, indicating
32 proliferation may be a more sensitive predictor of adverse health effects of formaldehyde
33 exposure (see Figures 4-11 and 4-12).
This document is a draft for review purposes only and does not constitute Agency policy.
4-221 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-17. Summary of formaldehyde effects on cell proliferation in the upper respiratory tract
Species
Male F344
rats;
male
B6C3F1
mice
Male F344
rats
Male F344
rats;
male
B6C3F1
mice
Male F344
rats
Male
BC3F1
mice
Male
albino
Wistar rats
Na
NRd
NR
NR
4-5
4-5
5d
Treatment11
0.5, 2, 6, or 15 ppm
6 hours/day for
3 days
0.5, 2, or 6 ppm
6 hours/day
1, 3, or 9 days
3 ppm for 12 hours,
6 ppm for 6 hours, or
12 ppm for 3 hours
3 or 9 days
15 ppm
6 hours/day
1 or 5 days
15 ppm
6 hours/day
1 or 5 days
3.5 ppm
8 hours, twice a day
for 3 days
Measure of cell
proliferation
LI: percent labeled cells
on tissue sections
(3H-thymidine I.P.d
2 hours postexposure)
LI: percent labeled cells
on tissue sections (3H-
thymidine I.P. 18 hours
postexposure)
LI: percent labeled cells
on tissue sections (3H-
thymidine I.P. 18 hours
postexposure for 2 hours)
LI: percent labeled cells
on tissue sections (3H-
thymidine I.P. 18 hours
postexposure for 2 hours)
LI: percent labeled cells
on tissue sections (3H-
thymidine I.P. 18 hours
postexposure for 2 hours)
Qualitative staining for
PCNA on tissue sections
Summary of results by location0
Level 2: Rats exhibited greater increased cell proliferation than mice.
No increase seen in rats or mice at 0.5 or 2.0 ppm.
No increase seen in mice at 6 ppm, but rats had 20-fold increase in
proliferation.
10- to 20-fold increase seen in both rats and mice at 15 ppm.
Level 2: Transient increase in cell proliferation on day 1 at 0.5 and 2.0 ppm.
Increase in cell proliferation on days 1,3, and 9 by 6 ppm.
Level 1 :
3 days: Greater increased proliferation in rats than mice. Increases
similar for various concentrations yielding the same C x t product.
9 days: Mice exhibited duration-dependent increases in proliferation,
inverse to concentration for constant C x t.
Level 2
3 days: Concentration-dependent increase in cell proliferation.
9 days: Concentration-dependent increase in cell proliferation in rats; no
increase in mice.
Level 2: Increase in cell proliferation in respiratory epithelium,
nasoturbinates, maxilloturbinates, and lateral wall.
1 day: 5. 5 lf versus 0.43% in controls
5 days: 10.1%f
Level 2: Increase in cell proliferation in respiratory epithelium,
nasoturbinates, maxilloturbinates, and lateral wall.
1 day: 2. 14f versus 0.27% in controls
5 day: 3.42%f
Levels 2 and 3 : Increase in cell proliferation in respiratory epithelium,
nasoturbinates, maxilloturbinates, septum, and lateral wall.
Reference
Swenberg et
al. (1986)
Swenberg et
al. (1986)
Swenberg et
al. (1986)
Chang et al.
(1983)e
Chang et al.
(1983)e
Cassee and
Feron(1994)e
& S n
5'
s ,
to
to
to
fe
H
o
o
-------
Table 4-17. Summary of formaldehyde effects on cell proliferation in the upper respiratory tract (continued)
& s
II
s ,
Species
Na
Treatment
Measure of cell
proliferation
Summary of results by location0
Reference
Male
albino
Wistar rats
0, 5, or 10 ppm
8 hours/day
continuously for
3 days or 4 weeks, or
0, 10, or 20 ppm
8 hours/day
intermittent8
for 3 days or 4 weeks
LI: percent labeled cells
on tissue sections (3H-
thymidine IP. 18 hours
postexposure for 2 hours)
Section level not stipulated in report.
3 days: 0.86% in controls
2.83%f at 5 ppm continuous
8.87%f at 10 ppm continuous
9.80%f at 10 ppm interrupted
19.8%f at 20 ppm interrupted
Wilmer et al.
(1987)e
4 weeks: 0.68% in controls
1.33% at 5 ppm continuous
8.85%h at 10 ppm continuous
3.41%f at 10 ppm interrupted
13.9%f at 20 ppm interrupted
Male
albino
Wistar rats
0, 1, 10, or 20 ppm
6 hours/day for
3 days
LI: percent labeled cells
(18 hour postexposure ex
vivo 3H-thymidine labeled
excised mucosa)
Level 3
Metaplastic epithelium: increased proliferation
31.4% at 10 ppm, 37.6% at 20 ppm
Visibly unaffected respiratory epithelium
1.6% in controls
2.6% at 10 ppm, 2.8% at 20 ppm
Woutersen et
al. (1987)e
Male
albino
Wistar rats
0, 1, or 2 ppm
8 hours/day
continuously for
3 days or 4 weeks,
or 0, 2, or 4 ppm
8 hours/day
intermittent8
for 3 days or 4 weeks
LI: percent labeled cells
on tissue sections (3H-
thymidine I.P. 18 hours
postexposure for 2 hours)
Level 2
3 days: No change from controls
0.60% in controls
0.34% at 1 ppm continuous
0.61% at 2 ppm continuous
0.29% at 2 ppm interrupted
0.58% at 4 ppm interrupted
4 weeks: no change from controls
1.03% in controls
0.81% at 1 ppm continuous
0.91% at 2 ppm continuous
1.16% at 2 ppm interrupted
2.86% at 4 ppm interrupted
Wilmer et al.
(1989)e
to
to
fe
H
o
O
Male and
female
albino
Wistar rats
0, 0.3, 1, 3 ppm
6 hours/day,
5 days/week
for 3 days or
13 weeks.
LI: percent labeled cells
on tissue sections (3H-
thymidine I.P. 18 hours
postexposure for 2 hours)
Level 2: Increased cell proliferation at days 3 and 13 weeks (p < 0.001).
Level 3: Transient dose-dependent increase at 1 and 3 ppm;
only seen at day 3
(p< 0.001).
Note: Results pooled by sex. Data shown graphically on log-normal scale.
Zwart et al.
(1988)e
-------
Table 4-17. Summary of formaldehyde effects on cell proliferation in the upper respiratory tract (continued)
Species
Male
Wistar rats
Male
Sprague-
Dawley
rats
Male F344
rats
Na
5
3-5
6
Treatment11
0,0.3, 1, or 3 ppm
22 hours/day for
3 days
0, 2, 6, or 20 ppm
6 hours/day for
1 or 3 days
0.7, 2, 6, 10, or
15 ppm
6 hours/day,
5 days/week for 1, 4,
or 9 days or 6 weeks
Measure of cell
proliferation
LI: percent labeled cells
on tissue sections (2 hours
postexposure ex vivo 3H-
thymidine-labeled excised
mucosa)
LI: percent labeled cells
by flow cytometry
(5-bromodeoxyuridine IP.
18 hours postexposure for
2 hours)
ULLI (unit length LI) (3H-
thymidine I. P. 18 hours
postexposure for 2 hours)
Summary of results by location0
Level 2: 3 ppm increased cell proliferation in nasoturbinates,
maxilloturbinates, septum, and lateral wall (p < 0.05).
Level 3 : No significant increases in cell proliferation.
Respiratory and olfactory epithelial cells.
1 day: 1.3% in controls
2.4%at2ppmf 3 days:
3 .7% at 6 ppmf 1 .4% at 2 ppm
2.7% at 20 ppmf 2.5% at 6 ppmf
Trachea! epithelial cells 2 . 3 % at 20 ppmf
1 day: 1.2% in controls
3.1%at2ppmf 3 days:
2.1% at 6 ppm 0.3% at 2 ppmf
2.8% at 20 ppmf 0.6% at 6 ppmf
Free lung cells (>97% MPs): 2.5% at 20 ppmf
no significant change.
Level 3
No increases in cell proliferation at 0.7 or 2 ppm.
Level 4
ULLI increases in locations without lesions at 6 ppm.
Increases in ULLI at all locations at 10 and 15 ppm.
Reference
Reuzel et al.
(1990)
Roemer et al.
(1993)
Monticello et
al. (1991)
& s
II
s ,
to
to
fe
H
o
o
-------
Table 4-17. Summary of formaldehyde effects on cell proliferation in the upper respiratory tract (continued)
Species
Male
rhesus
monkeys
Na
o
J
Treatment11
6ppm
6 hours/day for
1 or 6 weeks
Measure of cell
proliferation
LI: percent labeled cells
on tissue sections (3H-
thymidine IP. 18 hours
postexposure for 2 hours)
Summary of results by location0
Nasal passages: Duration-dependent increase in cell proliferation at all levels
(B-E) in transitional, respiratory, and olfactory epithelium.
Increased cell proliferation in areas with minimal lesions.
Larynx: trend for increased proliferation
Trachea: increased cell proliferation
1 week : 1.14 versus 0.55% in controls
6 weeks: 3.73%
Carina of trachea: increased cell proliferation.
1 week: 1.34 versus 0.43% in controls
6 weeks: 3.60%
Respiratory bronchioles: no increase in proliferation.
Reference
Monticello et
al. (1989)
& s
II
s ,
to
to
fe
H
aN = number of animals per treatment group.
bTreatment is given as the concentration of formaldehyde, duration of exposure each day, and length of the experiment in days and weeks.
'Standard section levels of the nasal passages as shown in Figure 4-6 are given for experiments in rats or mice.
dNR = not reported; I.P. = intraperitoneally.
"Study is described in full in Section 4.2.1.2.2.4.
TJifferent from control, p < 0.05.
Intermittent exposures were 30 minutes per hour for 8 hours.
hData from one animal only.
O
o
-------
1 4.2.2.3. Short-Term Studies
2 Inhalation of formaldehyde for a few hours has been shown to result in damage of the
3 nasal mucosa, depending on the exposure concentration. Bhalla et al. (1991) observed changes
4 in cell morphology in male Sprague-Dawley rat nasal epithelia after a single 4-hour exposure to
5 10 ppm (12.3 mg/m3) formaldehyde. Three exposed rats were sacrificed 1 hour and 24 hours
6 after exposure, with two control rats at each time point. Noses were fixed, decalcified, and sliced
7 along the midsagittal plane through the nasal septum. The exposed turbinates were examined by
8 scanning electron microscopy. Transverse sections through the hard palate, at the level of the
9 incisive papillae, were prepared for light microscopy from similarly exposed rats (n = 10). The
10 authors provided detailed descriptions of cell epithelial organization in untreated rat turbinates
11 and changes observed in formaldehyde-treated rats, as set forth below. No statistical analysis
12 was provided.
13 Scanning electron microscope examination of nasoturbinates showed increased mucus,
14 erythrocyte infiltration, swelling of microvillus cells, and some cell separation in formaldehyde-
15 treated rats. Nasoturbinates examined 1 day after exposure showed greater effects, including cell
16 damage, matted cilia, and blebbing of cell membranes. Damage to microvillus cells of the
17 maxilloturbinate included deformed cilia, cell swelling and rupture, and lack of typical microvilli
18 on the cell margins. As in the nasoturbinates, damage was more marked 24 hours after exposure.
19 The epithelium of the ETs exhibited less cell damage than in the nasal and maxillary regions,
20 with the slight lesions noted in the upper (ET1) portion and little to no damage noted on the mid
21 and lower (ET2 and ET3) regions. Examination of transverse tissue sections revealed swollen
22 goblet cells and stretched epithelial cells that formed an epithelial lining approximately 40%
23 taller than the lining seen in control rats. There was also a patchy loss of ciliated cells in the
24 respiratory epithelium, where columnar cells were present.
25 Buckley et al. (1984) investigated the respiratory tract lesions associated with several
26 sensory irritants. As part of this investigation, male Swiss-Webster mice were exposed to
27 3.13 ppm (3.85 mg/m3) formaldehyde 6 hours/day for 5 days. A total of nine chemicals were
28 tested in parallel. The report indicates there were 24-34 mice in each group, although not
29 detailed for each chemical. One-half of the treatment group and unexposed controls were
30 sacrificed immediately after the last exposure. The remaining exposed mice were sacrificed
31 72 hours later. The head, trachea, and lungs were fixed and heads decalcified. Five sections
32 were taken of each nose at levels equivalent to standard levels 2-6 (see Figure 4-6) and were
33 examined by light microscopy. Details on lung and trachea sections were not given.
34 Formaldehyde induced lesions in the respiratory epithelium of exposed mice, including
35 inflammation, exfoliation, erosion, ulceration, necrosis, and squamous metaplasia. The section
This document is a draft for review purposes only and does not constitute Agency policy.
4-226 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
level for these effects was not given. No effects were reported in the squamous epithelium,
olfactory epithelium, trachea, or lungs of formaldehyde-exposed mice.
Monteiro-Riviere and Popp (1986) evaluated damage to the respiratory epithelium due to
acute formaldehyde exposures. Male F344 (CDF [F344]/CrlBr) rats (three to five per group)
were exposed at 0.5, 2.0, 6.0, or 15 ppm (0.62, 2.5, 7.4, or 18.5 mg/m3) formaldehyde
6 hours/day for either 1, 2, or 4 days. Rats were sacrificed either immediately after exposure or
18 hours later (see Table 4-18). After fixation and decalcification, blocks of tissue were
collected from transverse sections of the skull. The first block of tissue, 1 jim thick, was taken
just posterior of the incisor teeth. The second block was taken halfway between the first block
and the incisive papillae. The dorsal nasal conchae, lateral wall, and ventral nasal conchae were
microdissected, postfixed, and viewed by transmission electron microscopy (Monteiro-Riviere
and Popp, 1986).
Table 4-18. Concentration regimens for ultrastructural evaluation of male
CDF rat nasoturbinates
Formaldehyde3'11
0.5 ppm (3)
2.0 ppm (3)
6 ppm
(5 each group)
15 ppm
(5 each group)
Duration
6 hours for 1 day
6 hours for 4 days
6 hours for 1 day
6 hours for 4 days
6 hours for 1 day
6 hours for 1 day
6 hours for 2 days
6 hours for 4 days
6 hours for 1 day
6 hours for 2 days
Time of sacrifice
18 hours later
18 hours later
Immediately
18 hours later
18 hours later
Observations
No lesions.
Altered ciliary configuration.
No lesions.
Altered ciliary configuration.
Focal lesions on dorsal and nasal conchae and
lateral wall.
Severity of lesions increased with exposure
duration.
Focal lesions on dorsal and nasal conchae and
lateral wall.
Severity of lesions increased with exposure
duration.
Severity of lesions increased with concentration.
17
18
19
20
21
22
23
24
25
26
27
aNumber of exposed rats is shown in parentheses.
bFive control rats were examined for each experiment.
Source: Monteiro-Riviere and Popp (1986).
No lesions were observed at either 0.5 or 2.0 ppm formaldehyde for either 1 day or
4 days, evaluated 18 hours after exposure. However, an unusual altered ciliary configuration,
including blebbing of the cell membrane, was observed in almost all formaldehyde-treated rats,
whereas it was only "occasionally noted" in control rats. Focal lesions in the dorsal and ventral
This document is a draft for review purposes only and does not constitute Agency policy.
4-227 DRAFT—DO NOT CITE OR QUOTE
-------
1 conchae and lateral wall were seen in rats exposed at 6 and 15 ppm for 1 day and sacrificed
2 immediately after exposure. These lesions included cytoplasmic and autophagic vacuoles, loss
3 of microvilli, and hypertrophy. Lesions increased in severity with both exposure concentration
4 and duration. Neutrophil infiltration and intercellular edema were seen after 1 day at 6 and
5 15 ppm. Nonkeratinized squamous metaplasia was noted after 4 days at 6 and 15 ppm in treated
6 rats. Cell death and sloughing were noted after only 2 days of exposure at 6 ppm formaldehyde.
7 As described above, Cassee and Feron (1994) examined the effects of intermittent
8 exposure to formaldehyde (3.5 ppm [4.3 mg/m3]), ozone (0.44 ppm [0.86 mg/m3]), or a
9 combination of the two on changes to the rat nasal epithelium. Exposure occurred through six
10 consecutive 12-hour cycles in which rats were exposed for 8 hours and then not exposed for a
11 further 4 hours. Rats were weighed before the first and after the last exposure periods and
12 sacrificed immediately after the last exposure. To collect tissue for biochemical analysis, skulls
13 were split sagittally and the respiratory epithelium collected. Tissues from six rats were pooled
14 and homogenized to enable the measurement of glutathione (GSH) and the activities of the
15 following enzymes: glutathione S-transferase (GST), glutathione peroxidase (GPX), glucose-6-
16 phosphate dehydrogenase (G6PDH), glutathione reductase (GR), alcohol dehydrogenase (ADH),
17 and formaldehyde dehydrogenase (FALDH). The remaining heads were fixed, decalcified, and
18 sectioned (standard cross sections [see Figure 4-6]).
19 All groups, including controls, lost weight during the course of treatment. Rats exposed
20 to formaldehyde, ozone, or both lost more weight than controls (p < 0.05, p < 0.01, and
21 p< 0.001, respectively). Formaldehyde treatment alone increased GPX from 48.6 to
22 64.0 |imole/minute-mg protein (p < 0.05) (see Table 4-19). Formaldehyde exposure, in
23 conjunction with ozone, decreased GST from 490 to 389 |imole/minute-mg protein (p < 0.05).
24 No other enzyme activities or tissue GSH levels were affected by formaldehyde exposure.
25 Formaldehyde-exposed rats exhibited lesions in the nasal epithelium at levels 2 and 3 of
26 the nose, with effects slightly more severe in level 2. Lesions observed include necrosis,
27 hyperplasia accompanied by squamous metaplasia, and rhinitis. Exposure to formaldehyde in
28 the presence of ozone resulted in more severe squamous metaplasia (statistics not given). These
29 findings are similar to those of Monteiro-Riviere and Popp (1986), indicating that single or
30 repeated exposures can result in cell damage and death. Cell death and increased cell
31 proliferation were seen here after 3 days of repeated exposures to 3.5 ppm formaldehyde. While
32 no increases were seen in olfactory epithelium, frank necrosis, squamous metaplasia, and
33 hyperplasia of both ciliated and nonciliated epithelium were noted at level 2 and 3.
34
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-228 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-19. Enzymatic activities in nasal respiratory epithelium of male
Wistar rats exposed to formaldehyde, ozone, or both
Enzyme
ADH
GST
GPX
G6PDH
GR
FALDH
Controls"
2.66 (0.99)
490 (32)
48.6 (4.3)
58.9 (7)
275 (16)
0.77 (0.03)
Formaldehyde (3.5 ppm)
3.53(0.13)
494 (24)
64.0 (7.9)c
60.8 (4.7)
288.2 (16)
0.68 (0.04)
Ozone (0.4 ppm)
3.40 (0.33)
514(4)
55.6 (2.0)
65.8(1.0)
279 (17)
0.68 (0.07)
Bothb
2.42 (0.61)
389 (28)c
54.5 (0.3)
45.5 (6.8)
236 (14)
0.80 (0.08)
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
aValues shown are the means and SDs of three measurements of a pooled sample. Units are
|imole/minute/mg of cytosolic protein.
'Hats were exposed intermittently, 12-hour cycles of 8 hours exposed and 4 hours unexposed, for 3 days.
Different from control, p < 0.05.
Source: Cassee and Feron (1994).
Javdan and Taher (2000) exposed male and female albino Wistar rats (five/group) at 0, 2,
or 5 ppm (0, 2.5, or 6.2 mg/m3) formaldehyde 8 hours/day for either 3 or 30 days. Transverse
tissue sections at the base of the incisive teeth and the first palatine folds were examined by light
microscopy. Lesions reported after 3 days of exposure to 2 ppm formaldehyde included chorion
congestion, cell disarrangement, squamous hyperplasia, atypical mitosis, and epithelial
hyperplasia. Similar lesions were seen after 30 days but were more severe. Effects at 5 ppm
formaldehyde included goblet cell proliferation, olfactory epithelial hyperplasia, calcified
regions, and an abscess on the chorion. These lesions were more severe after 30 days of
exposure.
Kamata et al. (1996a, b) conducted several high-dose studies by inhalation in rats.
Specifically they exposed male F344 rats to 0, 128.4, or 294.5 ppm (0, 158, or 362 mg/m3)
formaldehyde for 6 hours (Kamata et al., 1996a). In a subsequent study in the same laboratory,
male F344 rats were exposed to either 0, 15, or 145 ppm (0, 18.5, or 178 mg/m3) formaldehyde
nose only for 6 hours (Kamata et al., 1996b). Congestion was noted in the nasal cavities of
formaldehyde-exposed rats and was more severe at 145.6 ppm (Kamata et al., 1996b). Rats
exposed to 15 ppm formaldehyde had lesions in the nasal turbinate and trachea (not detailed)
(Kamata et al., 1996b). A slight hypersecretion of mucus was noted in the tracheal epithelium in
the absence of histopathologic changes. Rats exposed to 145.6 ppm had more dramatic lesions
that penetrated more deeply into the respiratory tract. Hyperkeratosis of the squamous
epithelium was found at level 1 of the nasal cavity. Hypersecretion, desquamation, and irregular
This document is a draft for review purposes only and does not constitute Agency policy.
4-229 DRAFT—DO NOT CITE OR QUOTE
-------
1 mucosal epithelium were seen in levels 2, 3, 4, and 5 of the nasal cavity, with more severe
2 changes noted in the nasal septum. Increased secretion and desquamation of mucosal cells
3 occurred in the trachea, and a slight hyperplasia of the alveolar wall was noted in rats exposed to
4 145.6 ppm formaldehyde (Kamata et al., 1996b)
5 Hester et al. (2003) carried out a transcriptional analysis of the nasal epithelium of male
6 F344 rats 24 hours after nasal instillation of 40 jiL of 400 mM formaldehyde. Immediately after
7 sacrifice, cells were harvested from the nasal cavity for RNA extraction. The authors found
8 several phase I and II enzymes, indicative of oxidative stress, to be elevated. They also reported
9 the greatest increase in inflammatory genes, such as iNOS and neuropeptides. In an effort to
10 phenotypically link any gene changes to pathology, Hester et al. (2003) noted that this exposure
11 scenario has been demonstrated to induce regenerative hyperplasia with minimal cytotoxicity. In
12 this regard, they observed no significant change in nine genes involved in three apoptotic
13 pathways.
14 In an expansion of their earlier study, Hester et al. (2005) carried out a transcriptional
15 analysis of the nasal epithelium of male F344 rats that had been exposed to formaldehyde by
16 nasal instillation for a single exposure, 5 days of exposures, or 28 days of exposure. In addition,
17 this study also attempted to characterize the comparative toxicity of glutaraldehyde with
18 structurally similar formaldehyde (van Birgelen et al., 2000). Thus, four animals per group were
19 instilled with 40 jiL of deionized water (control group), 40 jiL of 400 mM formaldehyde, or
20 40 jiL of 20 mM glutaraldehyde. Phenotypically, both aldehydes induced similar
21 histopathologic changes.
22 Both aldehydes induced similar changes in DNA repair and apoptotic pathways initially,
23 but the patterns of gene changes were different after about 5 days of exposure. Eight genes were
24 differentially expressed between formaldehyde and glutaraldehyde that indicated different
25 pathways for DNA repair, including recombination, base excision repair, and nucleotide excision
26 repair. Within this group, replication protein 70 and DNA excision repair ERCC1 showed a
27 twofold induction by formaldehyde compared with glutaraldehyde. Since both of these genes
28 and their products function by recognizing and removing damaged DNA bases, Hester et al.
29 (2005) hypothesized that formaldehyde-exposed cells may remove damaged bases more
30 efficiently than glutaraldehyde-exposed cells
31 In addition to nasal pathology, several researchers specifically investigated
32 formaldehyde-induced effects in the trachea, bronchi, and pulmonary tissues of the deep
33 respiratory tract in a variety of species (Lino dos Santos Franco et al., 2006; Kamata et al.,
34 1996a, b; Schreibner et al., 1979; lonescu et al., 1978).
This document is a draft for review purposes only and does not constitute Agency policy.
4-230 DRAFT—DO NOT CITE OR QUOTE
-------
1 lonescu et al. (1978) described progressive damage in pulmonary tissue of adult male
2 rabbits exposed to an aerosol of 3% formaldehyde solution 3 hours/day for up to 50 days
3 (method of aerosol generation or particle size were not provided). An equivalent air
4 concentration was not reported and cannot be derived from the information given. Animals were
5 sacrificed at several time points (3, 7, 15, 20, 30, and 50 days), and fragments of the caudal lobes
6 of both lungs were taken to examine bronchi (intrapulmonary and distal) and lung parenchyma.
7 Enzymatic activity was characterized in frozen sections for p-galactosidase, adenosine
8 triphosphatase (ATPase), adenosine monophosphatase (AMPase), lactate dehydrogenase (LDH),
9 malate dehydrogenase, succinate dehydrogenase (SDH), acid phosphatase, Tween-60 esterase,
10 naphthol-AS-D-acetate esterase, proline oxidase, hydroxyproline epimerase, leucyl
11 aminopeptidase, and p-glucuronidase. A portion of the lung was fixed and sectioned and viewed
12 by light microscopy to determine changes in cell populations and tissue pathology.
13 In addition, biochemical analysis revealed that enzymatic activity of P-galactosidase,
14 ATPase, AMPase, LDH, malate dehydrogenase, and SDH were all unchanged by formaldehyde
15 exposure across the course of treatment (lonescu et al., 1978). The activities of several enzymes
16 were increased through the course of exposure, including acid phosphatase, Tween-60 esterase,
17 naphthol-AS-D-acetate esterase, proline oxidase, and hydroxyproline epimerase. Although no
18 details were reported, the authors described the changes as progressive, with the increase in
19 proline oxidase and hydroxyproline epimerase seen only in the second half of the treatment
20 course. The activities of two enzymes, leucyl aminopeptidase and P-glucuronidase, were
21 observed to decrease rapidly (time frame not provided) (lonescu et al., 1978)
22 Histologic changes in the lung tissue were noted after only 3 days of exposure and were
23 generally progressive throughout the course of treatment. Early changes in the bronchial
24 epithelium included increased mucus secretion, hyperplasia, and hypertrophy of epithelial cells.
25 Lymphocyte infiltration was noted in many areas, and a limited thickening of the alveolar walls
26 was reported after 3 days of exposure. Epithelial cell lesions, thickening of the alveolar, and
27 infiltration of lymphocytes increased as exposure continued. Mucus cells increased as much as
28 40% after 40 days of treatment. After 40 days of treatment, lonescu et al. (1978) observed
29 "destructive and fibrotic lesions" and provided a detailed description of progressive lesions.
30 Schreiber et al. (1979) also examined histologic changes in lung tissue after high
31 formaldehyde exposures. Syrian golden hamsters (34, sex not stated) were exposed to 250 ppm
32 (308 mg/m3) formaldehyde 1 hour/day for 1, 2, 5, or 15 days. Five hamsters in each treatment
33 group were sacrificed 2 days after exposure was ended. Three hamsters in each group were
34 sacrificed 1, 2, or 6 weeks after exposure ended to determine if formaldehyde-induced changes
35 regressed over time. Tracheal washing was carried out to collect cytologic samples in each
This document is a draft for review purposes only and does not constitute Agency policy.
4-231 DRAFT—DO NOT CITE OR QUOTE
-------
1 animal prior to sacrifice. Samples were fixed, stained, and examined by light microscopy.
2 Lungs and tracheae were removed en bloc and fixed, and 20, 1 jim thick cross sections were
3 taken (location not detailed). The remaining respiratory tissue was sectioned at 200 jim
4 intervals. Sections were stained and viewed by light microscopy.
5 Abnormal epithelial cells were found in tracheal washings from formaldehyde-exposed
6 hamsters. Schreibner et al. (1979) described cells with lobulated nuclei and a coarse chromatin
7 pattern, especially in cells showing signs of degeneration (e.g., vacuolization of nuclei and
8 cytoplasm) (Schreiber et al., 1979). Cell number and damage were not quantified, and there was
9 no discussion of the effects of exposure duration on treatment, if any, on these observations.
10 Tracheal washing was normal 2 and 6 weeks after the end of exposure, indicating that the
11 cytological changes were reversible (Schreiber et al., 1979).
12 Formaldehyde exposure caused multifocal lesions in the mucociliary epithelium in the
13 trachea and larger bronchi. Dysplastic and poorly differentiated squamous metaplastic foci
14 replaced ciliated epithelium (Schreiber et al., 1979). Abnormal nuclear membranes, tonofibrils
15 around the nuclei, the appearance of nucleoli, and heterochromatin condensation were distinct in
16 the formaldehyde-treated hamsters. These changes, observed 2 days after formaldehyde
17 exposure, were reversible over time and not seen 2 and 6 weeks later.
18 Because of the similarity of form and physiology of rhesus monkey URTs to the human
19 respiratory tract, the effects of short-term formaldehyde exposure were evaluated in both nasal
20 and lung tissue in these monkeys by Monticello et al. (1989). Male rhesus monkeys (nine/group)
21 were exposed to 6 ppm formaldehyde (7.4 mg/m3) 6 hours/day for 5 days/week for either 1 or
22 6 weeks. Control animals were exposed to the same regimen of filtered air for 1 week. Monkeys
23 were weighed during the course of exposure and observed for clinical signs of irritation or
24 sickness. Monkeys were intravenously injected with [3H]-thymidine 18 hours after the last
25 formaldehyde treatment to evaluate induced cell proliferation. Sections of the nasal passages,
26 trachea, larynx, lung carina, and duodenum were processed for histoautoradiography. Tissues
27 fixed and sectioned for examination by light microscopy included nose, adrenal, sternum (bone
28 marrow), duodenum, esophagus, eyes, gallbladder, heart, kidney, liver, lymph nodes, pancreas,
29 stomach, spleen, and tongue. The nose was cut into a series of transverse sections, 3 jim thick,
30 and sections from five levels were examined (see Figure 4-10). Lung lobes were trimmed
31 midsagittally and sectioned with care to include airway bifurcations. Sections of the nasal
32 passages, trachea (cross section), larynx (cross section), lung carina (frontal section), and
33 duodenum were also processed for histoautoradiography.
34
This document is a draft for review purposes only and does not constitute Agency policy.
4-232 DRAFT—DO NOT CITE OR QUOTE
-------
A B C D E
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Figure 4-10. Diagram of nasal passages, showing section levels chosen for
morphometry and autoradiography in male rhesus monkeys exposed to
formaldehyde.
Source: Redrawn from Monticello et al. (1989).
There were no significant changes in body weight over the course of the experiment.
Oronasal breathing was noted in the first 15 minutes of formaldehyde exposure (Monticello et
al., 1989). Monkeys did experience eye irritation (mild lacrimation and conjunctival hyperemia)
during exposure.
Formaldehyde-related lesions were reported in the nasal passages, tracheas, and in the
larynx of treated animals (see Figure 4-11) (Monticello et al., 1989). Nasal epithelium from
treated animals exhibited many of the histologic lesions described in rodent studies, including
loss of goblet cells, loss of cilia, epithelial hyperplasia, squamous metaplasia, and neutrophilic
inflammatory response in the respiratory epithelium. The lesions were more severe after
6 weeks of exposure and were present over a greater percentage of the epithelium compared with
the 1-week exposure group (p < 0.05) (see Figure 4-11).
This document is a draft for review purposes only and does not constitute Agency policy.
4-233 DRAFT—DO NOT CITE OR QUOTE
-------
25r
20
x
o>
•o 15
D)
| 10
(D
0
A)
•
i
B C D E
LEVEL
CONTROL
1 WKEXP
6 WKEXP
B)
25r
20
15
10
0
B
C D
LEVEL
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Figure 4-11. Formaldehyde-induced cell proliferation in male rhesus
monkeys exposed to formaldehyde
Note: Animals were exposed to 6 ppm formaldehyde 6 hours/day, 5 days/week
for 1 or 6 weeks. Bar graph depicting mean labeling indices for the respiratory
epithelium at levels B-E. A: One-week exposure group. B: Six-week exposure
group. * Statistically different from controls (p < 0.05). Statistically different
from 1-week exposure group (p < 0.05).
Source: Redrawn from Monticello et al. (1989).
There was a distinct anterior to posterior gradient in both 1-week and 6-week treatment
groups in which the anterior regions had a higher percentage of impacted epithelium (Monticello
et al., 1989). However, the longer duration exposure produced significantly more lesions in the
larynx and trachea compared with those observed after only 1 week of exposure (p < 0.05). No
formaldehyde-related lesions were reported for the epithelium of the maxillary sinus, a structure
not present in rodents. Labeling indices (LIs) from the histoautoradiograms indicated increased
cell proliferation in transitory, respiratory, and olfactory epithelial cells after the 6-week
This document is a draft for review purposes only and does not constitute Agency policy.
4-234 DRAFT—DO NOT CITE OR QUOTE
-------
2
O
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
formaldehyde exposure (see Figure 4-12) (Monticello et al., 1989). Similar trends were seen
after only 1 week but were statistically significant only in the respiratory epithelium. Although
increased proliferation in the trachea and carina was statistically significant after 1 week of
exposure, the greater increases seen after 6 weeks of exposure, compared with controls, were not
statistically significant. A small sample size (n = 3) and high variability may have contributed to
the lack of statistical significance. Monticello et al. (1989) noted that increased cell proliferation
was seen in locations with minimal histologic changes, indicating proliferation may be a more
sensitive predictor of adverse health effects of formaldehyde exposure.
WK EXP B)
WK EXP 60
% SURFACE AREA WITH LESION
60
50
40
30
20
10
0
-
T
xxxxxxxxxx
-N\\\\V
2 Hri:
B C D E L/T
LEVEL
ou
^o
40
on
OU
20
^ n
I U
n
•
•
C
D
ffl En
H
M 8
9 n
SH
B C D E L/T
LEVEL
Figure 4-12. Formaldehyde-induced lesions in male rhesus monkeys exposed
to formaldehyde.
Note: Animals were exposed to 6 ppm formaldehyde 6 hours/day, 5 days/week
for 1 or 6 weeks. Bar graph showing levels B-E of the nasal passages and the
larynx/trachea (L/T), depicting percent surface area with formaldehyde-induced
lesions. Morphometry of level A was excluded due to the similarity of normal
features of transitional epithelium to formaldehyde-induced lesions in the
respiratory epithelium. A: One-week exposure group. B: Six-week exposure
group.
* Statistically different from controls (p < 0.05).
| Statistically different from 1-week exposure group (p < 0.05).
Source: Redrawn from Monticello et al. (1989).
There are two reports in the literature assessing changes in pulmonary tissues after acute
formaldehyde exposures (Kamata et al., 1996a, b). Kamata et al. (1996a) exposed male F344
This document is a draft for review purposes only and does not constitute Agency policy.
4-23 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
rats to 0, 128.4, or 294.5 ppm (0, 158, or 362 mg/m3) formaldehyde for 6 hours. Lung lavage
samples were collected and the fluid analyzed for the lipids, free cholesterol, phosphatidyl
ethanolamine, phosphatidyl choline, sphingomyelin, and triglyceride.
The bronchoalveolar lavage (BAL) was analyzed for triglycerides, cholesterol, and
phosphatidyl choline. As in the first experiment (Kamata et al., 1996a), triglyceride
concentration was reduced in the lavage of treated animals, in this case, to 16% of controls in
lavage in those rats exposed to 145.6 ppm formaldehyde (see Table 4-20). Cholesterol
concentration was unchanged and phosphatidyl choline was increased to 220% of that of control
rats as a result of exposure to 145.6 ppm formaldehyde. However, BAL lipids were unchanged
in 15 ppm exposed rats. Triglycerides were reduced in unwashed lung tissue from
formaldehyde-treated rats in a concentration-dependent manner and free fatty acids were reduced
in rats exposed to 145.6 ppm formaldehyde. Neither triglyceride nor sphingomyelin was
detected in lung lavage fluid from the high treatment group.
Table 4-20. Lipid analysis of lung tissue and lung lavage from male F344
rats exposed to 0,15, or 145.6 ppm formaldehyde for 6 hours
Control3
15 ppma
145 ppma
Lung tissue
Free fatty acids (mg/g lung)
Triglyceride (mg/g lung)
Cholesterol (mg/g lung)
Phosphatidyl ethanolamine (mg/g lung)
Phosphatidyl choline (mg/g lung)
Sphingomyelin (mg/g lung)
3.30 (0.7)
1.55 (0.23)
1.72(0.10)
7.41(1.81)
11.0(1.49)
3.44 (0.75)
3.11(1.23)
0.74(0.14)c
1.41 (0.25)
7.46 (2.28)
9.65 (3.21)
3.13(1.28)
1.41 (0.63)b
0.62(0.17)c
1.16(0.55)
5.49(1.78)
7.53 (3.52)
2.51 (0.95)
Lung lavage
Triglyceride (mg/lung)
Cholesterol (mg/lung)
Phosphatidyl choline (mg/lung)
0.31(0.10)
0.04 (0.01)
0.66 (0.23)
0.24 (0.09)
0.04 (0.01)
0.84 (0.35)
0.05 (0.02)c
0.04 (0.01)
1.45 (0.3 l)c
aSD given in parentheses.
bSignificant difference from controls (p < 0.05).
Significant difference from controls (p < 0.01).
Source: Kamata et al. (1996b).
Concentration-dependent decreases were seen in nonprotein sulfhydryl (SH) groups and
lipooxygenase in nasal mucosa homogenate and nonprotein SH groups in lung tissue
This document is a draft for review purposes only and does not constitute Agency policy.
4-236 DRAFT—DO NOT CITE OR QUOTE
-------
1 homogenate (see Table 4-21). Increases in both lipooxygenase and LDH activities were found in
2 lung tissue homogenate from formaldehyde-exposed rats.
3
This document is a draft for review purposes only and does not constitute Agency policy.
4-237 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
Table 4-21. Formaldehyde effects on biochemical parameters in nasal
mucosa and lung tissue homogenates from male F344 rats exposed to 0,15,
or 145.6 ppm formaldehyde for 6 hours
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Control3
15 ppma
145 ppma
Nasal mucosab
Nonprotein SH groups (uM/g tissue)0
Lipid peroxides (uM/g tissue)
Glucose-6-dehydrogenase (U/g tissue)d
1.64 (0.50)
118(23)
1.96(0.10)
1.29 (0.28)
71 (16)f
1.87 (0.07)
0.73 (0.21)f
59 (18)f
2.07(0.13)
Lunge
Nonprotein SH groups (uM/g tissue)0
Lipid peroxides (uM/g tissue)
Glutathione reductase (U/g tissue)d
Lactate dehydrogenase (U/g tissue)d
1.83 (0.18)
72(8)
0.42 (0.25)
77.37 (9.28)
1.70(0.11)
95 (15)f
0.25 (0.05)
88.69 (7.66)
1.29 (0.28)f
93 (8)g
0.22 (0.05)
93.62 (4.99)f
aSD given in parentheses.
b5 or 10% nasal mucosa homogenates.
°nmol malonaldehyde/g tissue.
dUnits per gram tissue.
e20% lung homogenates.
Significant difference from controls (p < 0.01).
Significant difference from controls (p < 0.05).
Source: Kamata et al. (1996b).
Lino dos Santos Franco et al. (2006) studied the effects of inhaled formaldehyde on lung
injury and changes in airway reactivity in rats. The extent of local and systemic inflammation
was assessed by changes in leukocyte counts in BAL fluid, blood, bone marrow, and spleen.
Changes of reactivity of isolated tracheae and intrapulmonary bronchi in response to
methacholine were monitored in response to formaldehyde exposure. The authors exposed male
Wistar rats to formaldehyde generated from a 1% solution of formalin. However, they provided
insufficient information for the exposure concentration to be determined. Groups of six animals
were exposed to formaldehyde for either 0, 30, 60, or 90 minutes on 4 consecutive days. All
experiments were carried out 24 hours after the final exposure.
The authors reported a significantly increased number of leukocytes in the BAL fluid of
animals exposed to formaldehyde via inhalation. The effect reached a maximum for the longer
exposure duration (90 minutes). Compared with controls, rats exposed to formaldehyde
90 minutes/day for 4 days also displayed an increase in the number of total blood leucocytes
(1.4 ± 0.06 x io4 versus 0.8 ± 0.01 x 104 cells/mm3). These values are means ± standard error of
This document is a draft for review purposes only and does not constitute Agency policy.
4-23 8 DRAFT—DO NOT CITE OR QUOTE
-------
1 the mean (SEM) for six animals/group. The effect appeared to reflect changes in the
2 mononuclear cell population (1.1 ± 0.02 x 104 versus 0.6 ± 0.003 x 104 cells/mm3) rather than
3 peripheral blood neutrophils (0.2 ± 0.003 x 104 cells/mm3 in test animals and controls). There
4 was also an apparently compound-related increase in the total cell count in the spleen
5 (112.7 ± 4.4 x 106 versus 94.2 ± 5.5 x 106 cells). However, a change in the number of cells
6 eluted from bone marrow did not reach statistical significance (54.6 ± 1.3 x 106 versus
7 45.0 ± 4.3 x io6 cells). Lino dos Santos Franco et al. (2006) provided data on dose-dependent
8 changes in methacholine-induced contractions in isolated tracheae and bronchi obtained from
9 formaldehyde-exposed and control rats. Although the maximal contractile response induced by
10 methacholine in tracheae of formaldehyde-treated rats was unchanged compared with controls,
11 contractions in isolated bronchi were significantly weaker than those observed in controls.
12 The authors examined the effect of formaldehyde inhalation on rat lung mast cells.
13 Degranulation and significant neutrophil infiltration were features of the response to
14 formaldehyde (see Table 4-22).
15
16 Table 4-22. Mast cell degranulation and neutrophil infiltration in the lung of
17 rats exposed to formaldehyde via inhalation
18
Treatment group
Controls
Formaldehyde-exposed
Mast cell degranulation
(cells/mm2)3
Ob
2.0±0.4C
Neutrophil infiltration
(cells/mm2)3
0.3 ±0.2
5.2±1.7C
19
20 aValues are means ± SEM; n = 6.
21 b4.2 ± 0.6 cells/mm2 intact mast cells were found in the lungs of controls.
22 °No statistical analysis was provided by the authors for these changes.
23
24 Source: Lino dos Santos Franco et al. (2006).
25
26
27 Selected pharmacological agents were used to explore the mechanism by which exposure
28 to formaldehyde might have brought about the observed lung infiltration and bronchial
29 hyporesponsiveness. Lino dos Santos Franco et al. (2006) provided data showing that separate
30 pretreatment of the animals with compound 48/80, sodium cromoglycate (SCO), and
31 indomethacin reduced the formaldehyde effect on neutrophil release into BAL but had no effect
32 on mononuclear cell counts. Compound 48/80 and SCO also reversed the formaldehyde-induced
33 reduction in bronchial response to methacholine, but indomethacin had the opposite effect
34 (causing an additional decrease in bronchial responsiveness). In broad terms, these findings
35 were thought to implicate mast cells as a possible mediator of the toxicological effects of
This document is a draft for review purposes only and does not constitute Agency policy.
4-239 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde. Histologically, a significantly increased number of degranulated mast cells were
2 evident in the pulmonary tissue of rats that were exposed to formaldehyde.
3 Lino dos Santos Franco et al. (2006) also examined the regulatory role of NO on
4 formaldehyde-induced bronchial activity. Nitrites generated by cultured cells of BAL from
5 formaldehyde-treated rats increased about threefold compared with those from controls.
6 However, pretreatment with the NO synthase inhibitor, 7V-nitro-L-arginine methyl ester,
7 prevented the formaldehyde-induced bronchial hyporesponsiveness to methacholine but had no
8 effect on pulmonary leukocyte recruitment. These data implicate the existence of distinct
9 mechanisms for the induction of lung inflammation versus bronchial hyporeactivity. Further
10 support for this concept came from an experiment in which rats were pretreated with capsaicin to
11 examine the involvement of sensory fibers in lung inflammation and the bronchial
12 hyporesponsiveness induced by formaldehyde inhalation. Although the treatment did not
13 influence formaldehyde-induced bronchial hyporesponsiveness to methacholine, the number of
14 leukocytes recovered in the BAL fluid were reduced compared with those of rats exposed to
15 formaldehyde alone.
16
17 Extrapulmonary effects
18 Kamata et al. (1996a) exposed male F344 rats to 0, 128.4, or 294.5 ppm (0, 158, or 362
19 mg/m3) formaldehyde for 6 hours. In addition, blood samples were monitored for hematology
20 and clinical chemistry parameters, including red blood cell (RBC) count, hemoglobin (Fib),
21 packed cell volume (PCV), mean corpuscular volume (MCV), mean corpuscular hemoglobin
22 concentration (MCHC), white blood cell (WBC) count, and plasma levels of total protein (TP),
23 albumin (ALB), blood urea nitrogen (BUN), glucose, phospholipids, triglycerides, total
24 cholesterol, cholinesterase, and LDH. Male rats exposed to 294.5 ppm formaldehyde had
25 increased RBC count, Hb, hematocrit (HCT), MCV, and serum glucose (p < 0.05) compared
26 with controls (Kamata et al., 1996a). There were concentration-related decreases in serum
27 measures of TP, ALB, and phospholipids (p < 0.05). BUN was decreased in rats exposed to
28 128.4 ppm but increased in the higher treatment group (p < 0.05). Phospholipid analysis of the
29 lung surfactant indicated a decrease in the production in formaldehyde-treated animals (p <
30 0.05). Total free cholesterol, phosphatidyl ethanolamine, and phosphatidyl choline were reduced
31 to 60, 55, and 38% of controls for rats treated with 294.5 ppm formaldehyde (p < 0.05).
32 Sphingomyelin was reduced to 32% of controls in the low treatment group (p < 0.05).
33 In a subsequent study in the same laboratory (Kamata et al., 1996b), male F344 rats were
34 exposed to either 0, 15, or 145 ppm (0, 18.5, or 178 mg/m3) formaldehyde nose only for 6 hours
35 (Kamata et al., 1996b). Fifteen animals were treated at each level and separated into subgroups
This document is a draft for review purposes only and does not constitute Agency policy.
4-240 DRAFT—DO NOT CITE OR QUOTE
-------
1 of five animals each for tissue collection and the determination of other endpoints. Blood
2 samples were collected from one subgroup to determine such hematological and clinical
3 chemistry parameters as RBC count, Hb, PCV, MCV, MCHC, WBC count, and plasma levels of
4 TP, ALB, BUN, glucose, phospholipids, triglycerides, total cholesterol, LDH, alkaline
5 phosphatase (ALP), aspartate aminotransferase (AST), alanine aminotransferase (ALT), and
6 G6PDH. BAL was collected from five animals of each group and analyzed for phospholipids.
7 Lung homogenate from five animals in each treatment group was analyzed for nonprotein SH
8 groups, lipid peroxides, and total lipids. The 20,000 x g supernatant of the lung homogenate was
9 assayed for the activities of GR, G6PDH, and LDH. Similarly, nonprotein SH groups and lipid
10 peroxidase were measured in homogenates of excised nasal mucosa. At autopsy, organs (brain,
11 heart, lung, liver, kidney, spleen, and testis) were weighed and tracheae and nasal turbinates
12 examined. After fixation and decalcification, five sections across the nose were taken,
13 corresponding to standard sections 1-5 (see Figure 4-6).
14 Several blood parameters were affected after these acute exposures. The WBC count was
15 slightly increased, from 4.7 x 103 cells/mm3 in control rats to 5.1 x io3 cells/mm3 and 6.1 x io3
16 cells/mm3 at 15 and 145.6 ppm formaldehyde, respectively (Kamata et al., 1996b). Serum levels
17 of AST and LDH decreased in an apparent concentration-dependent manner (AST 68 and 54%
18 of controls and LDH 48 and 28% of controls, respectively). Serum levels of G6PDH and ALT
19 were decreased similarly across exposure groups at 45 and 78% of controls, respectively.
20 A synopsis of respiratory pathology findings following short-term exposure to
21 formaldehyde is presented in Table 4-23.
22
23 4.2.2.4. Subchronic Studies
24 In a study by Maronpot et al. (1986), female and male B6C3F1 mice (10/group) were
25 exposed at 0, 2, 4, 10, 20, or 40 ppm (0, 2.46, 4.92, 12.3, 24.6, or 49.2 mg/m3) formaldehyde 6
26 hours/day, 5 days/week for 13 weeks. Clinical observations were made daily, and mice were
27 weighed weekly. At autopsy, tissue sections from each organ system (approximately 50 tissues
28 per mouse) were fixed, stained, and examined by light microscopy. Noses were fixed,
29 decalcified, and transversely trimmed at three levels: the incisor teeth, midway between the
30 incisor teeth and first molar teeth, and the second molar teeth (corresponding to sections 2, 3, and
31 4 in Figure 4-6).
32 Although control mice gained weight, mice exposed to 40 ppm formaldehyde lost weight
33 during the 13-week exposures. Expressed by the authors as a percent of weight gain in controls,
34 the weight losses were -235% in males and -168.6% in females. Early mortality for both male
35 and female mice exposed to 40 ppm was 80%. Although gross and histochemical effects in
36 excised pieces from each organ system were evaluated, endometrial hypoplasia in mice treated
This document is a draft for review purposes only and does not constitute Agency policy.
4-241 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-23. Summary of respiratory tract pathology from inhalation exposures to formaldehyde—short-term studies
Species/strain
No./
group
Treatment
Respiratory effects
LOAEL/NOAEL
Reference
Nasal path ology
Male Sprague-
Dawley rats
Male Swiss-
Webster mice
Male F344 rats
Male Wistar rats
Male F344 rats
Male and female
Wistar rats
Male F344 rats
Male rhesus
monkeys
3
24-34
3-5
20
5
5/sex
15
9
Single 4-hour exposure to
10 ppm formaldehyde.
0 or 3.13 ppm formaldehyde
6 hours/day for 5 days.
0, 0.5, 2, 6, or 15 ppm
6 hours/day for 1, 2, or 4 days.
0 or 3.5 ppm formaldehyde
through six consecutive 12-hour
cycles in which rats were
exposed for 8 hours; 10 were
unexposed for 4 hours.
0, 6, or 15 ppm
[14C] -formaldehyde 6 hours/day
for a single day (naive group).
A pretreated group was exposed
to 6 or 15 ppm formaldehyde
6 hours/day for 4 days prior to
[14C] -formaldehyde exposure.
0, 2, or 5 ppm formaldehyde
8 hours/day for 3 or 30 days.
0, 15, or 145.6 ppm
formaldehyde for a single 6-hour
exposure.
0 or 6 ppm formaldehyde
6 hours/day for 1 or 6 weeks.
[3H]-thymidine was injected
prior to sacrifice.
Marked histopathologic changes to the nasoturbinates,
maxilloturbinates, ethmoidal turbinates, and goblet and
microvillus cells.
Histopathologic lesions to the respiratory epithelium,
including inflammation, exfoliation, erosion,
ulceration, necrosis, and squamous metaplasia.
Histopathologic lesions to the nasal conchae, lateral
wall, and ventral nasal conchae.
The activity of GPX was increased in respiratory
epithelium homogenates. The nasal respiratory
epithelium showed frank necrosis.
Cellular necrosis to the nasal epithelium. 10.05%
cellular proliferation.
Cell disarrangement, squamous hyperplasia, atypical
mitosis, and epithelial hyperplasia.
Histopathologic lesions in the nasal turbinates and
trachea
Histopathologic lesions, including loss of goblet cells,
loss of cilia, epithelial hyperplasia, squamous
metaplasia, and neutrophilic inflammation.
LOAEL = 10 ppm.
LOAEL = 3.13ppm.
NOAEL = 2 ppm for
focal lesions. Some
changes in ciliary
configuration were
evident at all exposures.
LOAEL = 3. 5 ppm.
LOAEL = 6 ppm.
NOAEL = 2 ppm.
LOAEL = 15 ppm.
LOAEL = 6 ppm.
Bhalla et al.
(1991)
Buckley et al.
(1984)
Monteiro-Riviere
and Popp( 1986)
Cassee and Feron
(1994)
Chang et al.
(1983)
Javdan and Taher
(2000)
Kamata et al.
(1996b)
Monticello et al.
(1989)
J^.
to
to
fe
H
o
o
-------
ts
^ 8
1!
f^ S
3 ,
Table 4-23. Summary of respiratory tract pathology from inhalation exposures to formaldehyde—short-term studies
(continued)
Species/strain
No./
group
Treatment
Respiratory effects
LOAEL/NOAEL
Reference
Tracheal and lung pathology
Syrian golden
hamsters (sex
unstated)
Male rabbits
(strain unstated)
Male F344 rats
Male F344 rats
Male Wistar rats
5
ND
6
15
6
0 or 250 ppm 1 hour/day for 1,
2, 5, or 15 days.
Aerosol generated from a 3%
formaldehyde solution
3 hours/day for up to 50 days
(air concentration unknown).
0, 128.4, or 294.5 ppm for a
single 6-hour exposure.
0, 15, or 145.6 ppm
formaldehyde for a single 6-hour
exposure.
Aerosol generated from a 1%
formalin solution 0, 30, 60, or
90 minutes/day on 4 consecutive
days (air concentration
unknown).
Abnormal cells in tracheal lavage, an effect that was
reversed on cessation of treatment.
Necrosis of the bronchi and lung parenchyma.
Increased activities of acid phosphatase, Tween-60
esterase, naphthol-AS-D-acetate esterase, proline
oxidase, and hydroxyproline epimerase. Reduced
activities of leucyl aminopeptidase and (3-
glucuronidase. Adverse histopathologic changes.
Phospholipid content was reduced in lung surfactant,
for example, sphingomyelin to 43% of controls in the
low-concentration group.
Biochemical changes in lung homogenates. Altered
lipid content of BAL in high concentration rats.
Increased leukocyte count in bronchoalveolar fluid.
Degranulation of mast cells and increased neutrophil
infiltration.
LOAEL = 250 ppm.
ND.
LOAEL = 128.4 ppm.
LOAEL = 15 ppm.
ND.
Schreiber (1979)
lonescu et al.
(1978)
Kamata et al.
(1996a)
Kamata et al.
(1996b)
Lino dos Santos
Franco et al.
(2006)
Extrapulmonary effects
Male F344 rats
Male F344 rats
6
15
0, 128.4, or 294.5 ppm for a
single 6-hour exposure.
0, 15, or 145.6 ppm
formaldehyde for a single 6-hour
exposure.
LOAEL = 128.4 ppm.
LOAEL = 15 ppm.
Kamata et al.
(1996a)
Kamata et al.
(1996b)
-i^
to
fe
H
ND = not determined; LOAEL = lowest-observed-adverse-effect level; NOAEL = no-observed-adverse-effect level.
O
O
-------
1 with 40 ppm was the only effect noted outside the respiratory system. The authors considered
2 this effect secondary to the observed respiratory tract lesions and frank toxicity at 40 ppm
3 formaldehyde.
4 While no statistical comparison was provided, respiratory tract lesions showed an
5 increased incidence with concentration, as well as an increased distribution throughout the
6 respiratory tract (see Table 4-24). No lesions were seen in the nasal cavity, larynx, trachea, or
7 lung of control mice or mice treated with 2 ppm formaldehyde. Minimal squamous metaplasia in
8 the nasal cavity was noted in 1 of 10 male mice treated with 4 ppm formaldehyde, but none were
9 observed in the female mice. However, squamous metaplasia was observed in all mice in the
10 higher treatment groups (10, 20, and 40 ppm). Lesions became more severe and penetrated more
11 deeply into the respiratory tract as exposure concentration increased. Where lesions were present
12 in the nasal cavities of all mice exposed to 10 ppm, similar lesions were reported in the larynx
13 and trachea of some animals exposed to 20 ppm and all animals exposed to 40 ppm
14 formaldehyde. Mice exposed to 40 ppm formaldehyde exhibited lesions as deep as the lung,
15 including squamous metaplasia, submucosal fibrosis inflammation, and epithelial hyperplasia.
16 The findings of Maronpot et al. (1986) indicated a no-observed-adverse-effect level
17 (NOAEL) of 4 ppm and a LOAEL of 10 ppm in mice, based on squamous metaplasia in the
18 nasal epithelium. Although a LOAEL of 10 ppm was observed, there was 80% mortality for
19 both sexes at 40 ppm, indicating a very narrow range between the first observed adverse health
20 effects and frank effect concentrations in mice for this 13-week treatment.
21 In a study by Woutersen et al. (1987), male and female albino SPF Wistar rats (10/group)
22 were exposed to 0, 1, 10, or 20 ppm (0, 1.23, 12.3, or 24.6 mg/m3) formaldehyde 6 hours/day,
23 5 days/week for 13 weeks. Rats were checked daily and weighed weekly. Three longitudinal
24 sections of lungs, trachea, and larynx and six standard cross sections of the nose were taken for
25 microscopic examination. Two rats per exposure group were similarly treated for 3 days and
26 sacrificed 18 hours later, and nasoturbinates were dissected to measure cell proliferation.
27 Woutersen et al. (1987) noted that the majority of the dose-dependent increases in cell
28 proliferation seen at section level 3 after 3 days of repeated 6-hour exposures to 10 and 20 ppm
29 (12.3 and 24.6 mg/m3) formaldehyde occurred in areas of the epithelium showing "clear
30 squamous metaplasia and hyperplasia." Cell proliferation rates in metaplastic epithelium of
31 29.5 and 33.2% were much higher than the 1.4 to 2.8% proliferation in the visibly unaffected
32 respiratory epithelium from rats exposed at 10 ppm formaldehyde. Although there was a slight
33 trend towards increased cell proliferation in the visibly unaffected epithelium of exposed animals
34 compared with unexposed controls, the majority of increased cell proliferation resulting from
35 exposure to 10 and 20 ppm formaldehyde was attributed to the metaplastic epithelium
36 (Woutersen et al., 1987).
37
This document is a draft for review purposes only and does not constitute Agency policy.
4-4.1.2.2.3-2B*AFT—DO NOT CITE OR QUOTE
-------
it
& s
II
Table 4-24. Location and incidence of respiratory tract lesions in B6C3F1 mice exposed to
formaldehyde
Location of respiratory
tract lesions
Nasal cavity
Squamous metaplasia
Seropurulent inflammation
Larynx
Squamous metaplasia
Trachea
Squamous metaplasia
Epithelial hyperplasia
Seropurulent inflammation
Submucosal fibrosis
Lung (Bronchus)
Squamous metaplasia
Inflammation
Submucosal fibrosis
Control
Male
a
—
-
-
Female
-
—
-
-
2 ppm
Male
-
—
-
NDb
ND
ND
Female
-
—
-
ND
ND
ND
4 ppm
Male
1/10
—
-
ND
ND
ND
Female
-
—
-
ND
ND
ND
10 ppm
Male
10/10
4/10
—
-
-
Female
10/10
—
1/10
-
20 ppm
Male
10/10
10/10
6/9
3/10
4/10
-
Female
10/10
8/10
3/9
5/10
2/10
-
40 ppm
Male
10/10
10/10
10/10
10/10
2/10
8/10
9/10
4/10
3/10
2/10
Female
10/10
10/10
7/8
10/10
5/10
5/10
3/10
2/10
5'
s ,
J^.
K>
fe
H
"Dash indicates no lesions recorded in that treatment group.
bND = no data.
Source: Maronpot et al. (1986).
O
o
-------
1 Statistically significant increases were seen in focal respiratory epithelial hyperplasia and
2 keratinization in both male and female rats at the highest treatment level (20 ppm) (see
3 Table 4-25). Male rats also had statistically significant increases in observed respiratory
4 epithelial squamous metaplasia, focal olfactory epithelial thinning, and rhinitis. Both male and
5 female rats treated with 10 ppm formaldehyde showed statistically significant increases in
6 squamous metaplasia, hyperplasia, and keratinization of the respiratory epithelium (Woutersen et
7 al., 1987).
8 Disarrangement of the respiratory epithelium was only significantly increased in female
9 rats, but this change was observed at both the 10 and 20 ppm treatment levels. Although some
10 lesions were observed in animals treated with 1 ppm formaldehyde, their incidences were not
11 statistically significant and the findings were equivocal.
12 Feron et al. (1988) examined recovery of formaldehyde-induced nasal lesions after
13 subchronic exposures. Male albino SPF Wistar rats (50-55/group) were exposed to 0, 10, or
14 20 ppm (0, 12.3, or 24.6 mg/m3) formaldehyde 6 hours/day, 5 days/week for either 4, 8, or
15 13 weeks. All groups were observed for a total of 130 weeks, including treatment and recovery.
16 Rats were weighed weekly for the first 13 weeks and monthly thereafter. Rats (five/group) were
17 sacrificed immediately after the end of exposure (4, 8, or 13 weeks). The balance of the rats
18 were sacrificed after 130 weeks, inclusive of exposure time. At sacrifice, noses were fixed and
19 sectioned by using standard section levels.
20 Formaldehyde exposure (20 ppm) was associated with reduced body weight throughout
21 the exposure period (4, 8, or 13 weeks). However, body weight in these groups matched that of
22 controls after 8, 40, and 100 weeks, respectively. Rats exposed to 10 ppm for 8 or 12 weeks had
23 slightly decreased body weight (further details not given).
24 Nonneoplastic lesions were reported in the nasal mucosa of rats exposed to either 10 or
25 20 ppm formaldehyde and examined immediately after exposure was discontinued (4, 8, or
26 13 weeks). Lesions increased in severity with both exposure duration and concentration (details
27 of severity and incidence were not provided). Rhinitis, hyperplasia, and squamous metaplasia of
28 the respiratory epithelium were seen in rats from both dose groups, but changes in olfactory
29 epithelia were only seen in rats exposed to 20 ppm, where cell disruption, thinning of the
30 epithelium, and simple cuboidal or squamous metaplasia were also reported. Changes in the
31 dorsomedial region, at the junction of the respiratory and olfactory epithelium, were similar to
32 those seen in the olfactory epithelium of rats exposed to 20 ppm formaldehyde. A similar
33 concentration- and duration-dependent increase in histopathologic changes in nasal epithelium
34 was observed after the full 130 weeks, which included 126, 122, or 117 weeks of recovery for
35 the three duration groups, 4, 8, and 13 weeks, respectively (see Table 4-25).
This document is a draft for review purposes only and does not constitute Agency policy.
4-246 DRAFT—DO NOT CITE OR QUOTE
-------
4
Table 4-25. Formaldehyde effects (incidence and severity) on histopathologic
changes in the noses and larynxes of male and female albino SPF Wistar rats
exposed to formaldehyde 6 hours/day for 13 weeks
Respiratory epithelium
Diffuse squamous
metaplasia
Focal squamous
metaplasia
Focal hyperplasia
Focal disarrangement
Focal keratinization
Severity
Slight
Moderate
Severe
Very slight
Slight
Moderate
Very slight
Slight
Moderate
Very slight
Slight
Moderate
Very slight
Slight
Moderate
Concentration of formaldehyde (ppm)
0 1 10 20
Males
a
5b
5b
- 1 - -
1 6b
- - 4 -
- - 1 1
6b r
- - i -
- - i -
- - 3 -
- - 1 -
- 2 6b 1
3 6b
- - - 1
0 1 10 20
Females
o
— ~~ ~~ J
_ _ _ 4
o
— ~~ ~~ J
- - 1 -
r
- - 2 -
- - 2 1
1 6b 6b
- - 2 1
1 6b 6b
6b 6b
- - 2 4
Olfactory epithelium
Focal thinning
Focal squamous
metaplasia
Focal keratinization
Rhinitis
Slight
Moderate
Severe
Slight
Moderate
Very slight
Slight
- - - 2
- - - 1
5b
_ _ _ 4
_ _ _ 4
- - - 1
- - - 2
- 2 5b 10C
- - - 2
- - - 2
- - - 2
o
— — — J
- - - 1
- - 3 2
Larynx
Squamous metaplasia
Keratinization
Very slight
Slight
Moderate
Slight
- - - 3
- - - 1
- - - 1
- - - 2
NEd NE
- NE NE -
- NE NE -
- NE NE -
5
6
7
aDash indicates no lesions reported.
''Different from control, p < 0.05.
"Different from control,/? < 0.01.
dNE = not evaluated.
Source: Woutersen et al. (1987).
This document is a draft for review purposes only and does not constitute Agency policy.
4-247 DRAFT—DO NOT CITE OR QUOTE
-------
1 Feron et al. (1988) did not provide a direct comparison among lesions reported at the
2 interim sacrifice and terminal sacrifice after the extended recovery period. However, similar
3 lesions were reported after the recovery period, including focal hyperplasia and stratified
4 squamous metaplasia of the respiratory epithelium, stratified cuboidal or squamous metaplasia in
5 the dorsomedial area, and replacement of olfactory epithelium. The incidence and severity of
6 these lesions in rats exposed to 20 ppm formaldehyde were statistically different from control
7 animals, regardless of exposure duration (see Table 4-26).
9
10
11
12
13
Table 4-26. Formaldehyde-induced nonneoplastic histopathologic changes in
male albino SPF Wistar rats exposed to 0,10, or 20 ppm formaldehyde
(6 hours/day, 5 days/week) and examined at the end of 130 weeks inclusive of
exposure
Formaldehyde, ppm
Total noses examined
4 Weeks
0 10 20
44 44 45
8 Weeks
0 10 20
45 44 43
13 Weeks
0 10 20
45 44 44
Respiratory epithelium focal hyperplasia
Very slight
Slight
Moderate
000
0 3 8b
0 0 1
0 1 3
2 2 12b
0 1 0
0 5a 2
1 6 14b
004
Respiratory epithelium focal stratified squamous metaplasia
Very slight
Slight
Moderate
Severe
Respiratory /olfactory epithelium stratified cuboidal
or squamous metaplasia
Rhinitis
3 6 14b
4 2 19b
023
000
004
7 7 18a
8 16 17a
2 1 20b
002
000
0 0 17b
4 6 22a
2 10a 2
3 18b 26b
1 5 14b
0 0 1
0 2 23b
8 11 23b
Olfactory epithelium replacement by respiratory epithelium and regeneration
Very slight
Slight
Moderate
Severe
000
1 0 6
0 0 1
000
002
0 0 14b
003
0 0 1
0 0 1
0 0 12b
0 0 12b
0 0 1
14
15
16
17
18
19
aSignificantly different from control, p < 0.05.
bSignificantly different from control, p < 0.01.
Source: Feron et al. (1988).
This document is a draft for review purposes only and does not constitute Agency policy.
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
Although a slight increase in changes to the olfactory epithelium and dorsomedial area
was seen in rats treated with 20 ppm formaldehyde for only 4 weeks, these differences were
significant and more severe in the 8- and 13-week treatment groups. Replacement of olfactory
epithelium by respiratory epithelium was described as slight after 8 weeks of exposure and slight
to moderate after 13 weeks of exposure in the 20 ppm treatment groups. Therefore,
formaldehyde-induced lesions were not resolved after a considerable nonexposure recovery
period of up to 126 weeks (Feron et al., 1988).
Feron et al., (1988) derived a correlation between the development of nonneoplastic
changes in nasal epithelium and the development of nasal tumors as a result of these subchronic
formaldehyde exposures. Two SCCs were reported in rats exposed to 10 ppm formaldehyde but
were not considered to be formaldehyde related because of their locations (nasolacrimal duct,
incisor tooth). Six tumors were observed in the 20 ppm, 13-week exposure group (see
Table 4-27) of which three of the tumors were SCCs similar to those observed as a result of
chronic formaldehyde exposure. Two polypoid adenomas also were reported in rats exposed to
20 ppm formaldehyde. Feron et al. (1988) concluded that subchronic exposures to 20 ppm
formaldehyde could result in an increase in nasal tumors, an effect that followed observation of
cellular proliferation.
Table 4-27. Formaldehyde-induced nasal tumors in male albino SPF Wistar
rats exposed to formaldehyde (6 hours/day, 5 days/week for 13 weeks) and
examined at the end of 130 weeks inclusive of exposure
Tumor type
No. of rats exposed for 4 weeks
Polypoid adenoma
sec
No. of rats exposed for 8 weeks
Polypoid adenoma
sec
No. of rats exposed for 13 weeks
sec
Cystic squamous cell carcinoma
Carcinoma in situ
Ameloblastoma
0 ppm
44
0
0
45
0
2
45
0
0
0
0
10 ppm
44
0
0
44
0
1
44
1
0
0
0
20 ppm
45
1"
1
43
1"
1
44
3a
1
la
1
aTumor considered to be associated with formaldehyde exposure.
Source: Feron etal. (1988).
This document is a draft for review purposes only and does not constitute Agency policy.
4-249 DRAFT—DO NOT CITE OR QUOTE
-------
1 A companion study from the same laboratory examined the effects of lower concentration
2 formaldehyde exposures (Zwart et al., 1988). Male and female albino Wistar rats (50/group)
3 were exposed to 0, 0.3, 1, or 3.0 ppm (0, 0.37, 1.2, or 3.7 mg/m3) formaldehyde 6 hours/day,
4 5 days/week for 13 weeks. Body weight, general condition, and behavior were recorded weekly.
5 No effects of formaldehyde exposure on body weight changes were noted, and growth was
6 considered comparable among different exposure groups and controls. Rats were sacrificed
7 during week 14, and noses were fixed and sectioned (exact time after exposure ended not given).
8 Six standard cross sections were examined for each animal by light microscopy, anterior to
9 posterior. Noses were fixed and decalcified, and six standard cross sections were taken and
10 developed.
11 No formaldehyde-related lesions were reported in the respiratory epithelium at section
12 level 3 after 13 weeks of formaldehyde exposure (0.1 ppm, 1 ppm, or 3 ppm). Signs of
13 inflammation (rhinitis, sinusitis, mononuclear cell infiltrates) were observed in
14 formaldehyde-treated rats, but there was no concentration-response relationship (data not
15 provided). Formaldehyde-related pathology in the anterior part of level 2 epithelium was
16 reported in 37/50 males and 21/50 female rats exposed to 3.0 ppm for 13 weeks. Both
17 keratinized and unkeratinized squamous metaplasia were present, and disarranged cells and
18 hyperplastic respiratory epithelium were found in the transitional zone between squamous and
19 pseudostratified epithelium at level 2. Foci of keratinized squamous epithelium, glandularization
20 of goblet cells, and deciliated epithelium were observed by electron microscopy in anterior
21 sections of level 2 of rats exposed to 3 ppm formaldehyde. Epithelial cells with irregularly
22 shaped and strongly indented nuclei were described at level 2 in animals exposed to 0.3 and
23 1 ppm formaldehyde and were considered to be disarranged as well at 3 ppm formaldehyde
24 exposures.
25 Although early cell proliferation at level 3 corresponded to basal cell hyperplasia at
26 3 days, neither effect persisted for the course of the exposure. The authors speculate that this is
27 an indication of an adaptive response, perhaps through increased function of the mucociliary
28 apparatus present at level 3. In contrast, the early changes at section level 2 were less dramatic
29 but persisted through 13 weeks, including clear formaldehyde-related pathology.
30 Concentration times time (C x t) issues have been investigated for histopathology as well
31 as for cellular proliferation, outlined above. Specifically, Wilmer et al. (1989, 1987) compared
32 the effects of 8-hour continuous and 8-hour intermittent formaldehyde exposure in two studies.
33 Fifty male albino Wistar rats (10/group) were exposed to different exposure regimens to achieve
34 similar compound-related C x t products. A C x t product of 40 ppm-hours (49.2 mg/m3-hours)
35 was attained by an 8-hour exposure to 5 ppm (6.2 mg/m3) or a 4-hour exposure to 10 ppm (12.3
This document is a draft for review purposes only and does not constitute Agency policy.
4-250 DRAFT—DO NOT CITE OR QUOTE
-------
1 mg/m3) (Wilmer et al., 1987). Similarly, an 80 ppm-hours (98.4 mg/m3-hours) C x t product was
2 attained from continuous 10 ppm exposure or intermittent 20 ppm (24.6 mg/m3) exposure. Rats
3 were exposed to one of these regimens 8 hours/day for either 3 days (two/group) or 4 weeks
4 (eight/group). Eighteen hours after exposure ended, rats were injected with [3H]-thymidine and
5 sacrificed 2 hours later. Noses were fixed and decalcified, and six standard cross sections were
6 taken and developed.
7 Thinning and disarrangement of the respiratory epithelium, squamous metaplasia, basal
8 cell hyperplasia, and rhinitis were seen in formaldehyde-treated rats. Lesions were most severe
9 in group 4 (20 ppm intermittent). Groups 2 and 3 had similar lesions (10 ppm intermittent and
10 continuous). Rats in group 1 had mild lesions. Formaldehyde concentration was the major
11 determinate in severity of nasal lesions. Formaldehyde effects were less severe in group 1 than
12 in group 3, even though the C x t product was the same, indicating concentration rather than
13 duration or cumulative exposure correlates to severity. Epithelial lesions in group 3 rats were
14 similar among rats exposed to 10 ppm, regardless of duration (groups 2 and 3).
15 In a follow-up study, Wilmer et al. (1989) assessed both cellular proliferation and
16 histologic lesions in Wistar rats exposed to formaldehyde in groups that differed by
17 concentration and time. Group A served as a control group (0 ppm). Group B was exposed to
18 1 ppm for 8 hours, group C to 2 ppm for 8 hours, group D to 2 ppm for 4 hours (30 minutes for
19 8 hours), and group E to 4 ppm for 4 hours (30 minutes for 8 hours). The experimental design
20 and cellular proliferation results are illustrated in Table 4-28. Intermittent exposures at 2 and 4
21 ppm resulted in formaldehyde-related histopathologic lesions similar to those reported by Zwart
22 et al. (1988). Disarrangement and squamous metaplasia in respiratory epithelium were observed
23 at 4 ppm (see Table 4-28). Disarrangement, nest-like infolds, goblet cell hyperplasia, and rhinitis
24 were observed at 2 ppm. Rats exposed continuously for 8 hours at 2 ppm formaldehyde had
25 fewer lesions than rats intermittently exposed to 2 ppm and were not statistically different from
26 controls. Although lesions were noted in rats given the continuous 1 ppm, 8-hour treatment,
27 their incidence was not significantly different from the controls (see Table 4-28). It should be
28 noted that the control rats in this study were reported to have a higher frequency of lesions than
29 controls in two previous studies from this laboratory employing the same techniques (Zwart et
30 al., 1988; Woutersen et al., 1987). For example, lesions noted in the respiratory epithelium of 25
31 control rats included 13 disarrangements, 13 basal cell hyperplasia, and 5 each of goblet cell
32 hyperplasia, nest-like infolds, and squamous metaplasia. This is in contrast to the data of
33 Woutersen et al. (1987), who reported no lesions in the respiratory epithelium of 20 control rats
34 (male and female). Although Zwart et al. (1988) discussed inflammatory lesions in control rats,
35 no mention was made of the other scored lesions in control animals. Overall, Wilmer et al.
This document is a draft for review purposes only and does not constitute Agency policy.
4-251 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-28. Formaldehyde effects on nasal epithelium for various
concentration-by-time products in male albino Wistar rats
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Respiratory epithelium
at cross section level 2
Disarrangement
Focal
Diffuse
Necrosis
Focal
Diffuse
Basal cell hyperplasia
Focal
Diffuse
Squamous metaplasia
Focal
Keratinization
Nest-like infolds
Focal
Diffuse
Goblet cell hyperplasia
Focal
Diffuse
Rhinitis
Exposure regimen (number of animals)
A (25)
0 ppm
12
1
4
0
9
4
5
0
5
0
0
5
3
B(22)
1 ppm
8-Hour
continuous
4
1
3
0
4
0
0
0
4
3
1
2
2
C(24)
2 ppm
8-Hour
continuous
8
0
0
0
6
0
1
1
11
1
1
8
3
D(23)
2 ppm
8-Hour
intermittent
3a
15b
2
2
11
4
7
0
14C
0
2
13b
16C
E(25)
4 ppm
8-Hour
intermittent
8
11°
3
2
10
11
16C
3
7
1
1
10
8
ap < 0.05, compared with group A.
bp < 0.001, compared with group A.
°p O.01, compared with group A.
Source: Winner etal. (1989).
(1989) reported clear adverse effects at 2 ppm formaldehyde, resulting from intermittent
exposure for 8 hours/day, 5 days/week for 13 weeks. The indication of no effects at 1 ppm and 2
ppm continuous exposure should be considered with some caution, given the unusual incidence
of lesions in the control animals.
The results reported by Wilmer et al. (1989, 1987) indicate a greater influence of
concentration, rather than exposure regimen (continuous versus intermittent) on formaldehyde
toxicity. However, these studies were conducted as repeated 8-hour exposure regimens over a
course of days or weeks. Therefore both regimens allowed for a 16-hour recovery time before
This document is a draft for review purposes only and does not constitute Agency policy.
4-252 DRAFT—DO NOT CITE OR QUOTE
-------
1 the next reexposure and do not represent a true continuous exposure. This research group has
2 speculated that defensive adaptation of the nasal mucosa may include the function of the
3 mucociliary apparatus (Feron et al., 1989). Morgan et al. (1986a) have shown formaldehyde
4 effects on mucus flow and ciliary beat in F344 rats to result from hourly exposures to 15 ppm
5 formaldehyde. However, effects seen in repeated 8-hour exposures may not correspond to those
6 effects resulting from shorter duration exposures to higher formaldehyde concentrations.
7 Rusch et al. (1983a, b) performed a comparative study of formaldehyde effects on the
8 nasal epithelium in F344 rats, Syrian golden hamsters, and cynomolgus monkeys. Groups of
9 animals were exposed at 0, 0.2, 1, or 3 ppm (0, 0.25, 1.2, or 3.7 mg/m3) formaldehyde
10 22 hours/day, 7 days/week for 26 weeks. Six male monkeys, 10 male and 10 female hamsters,
11 and 20 male and 20 female rats were exposed at each exposure level. The experiment was run in
12 two trials, each with its own control group: trial 1 at 0.2 or 1 ppm and trial 2 at 3 ppm. Animals
13 were weighed weekly and physically assessed (details not given). At sacrifice, organ weights
14 were recorded for the kidney, adrenals, heart, and liver. Tissue sections of the lung (4), trachea,
15 and nasal turbinates (4) of each animal were examined by light microscopy (section locations not
16 given). Additionally, sections were examined by electron microscopy for rats in the control and
17 1 ppm treatment groups (five rats per group).
18 Body weights of both male and female rats in the 3 ppm treatment group were depressed
19 by 20% between week 2 and the end of the 26-week exposure. Absolute liver weights were
20 decreased in these animals as well (26% lower in males and 12% lower in females,/* < 0.05).
21 This decrease in liver weight remained significant for male rats when normalized for body
22 weight (a ratio of 2.9 in treated versus 3.16 in controls) but not for female rats. No significant
23 body weight or organ weight changes were seen in hamsters or monkeys. Increased incidences
24 of congestion (36/156), hoarseness (32/156), and nasal discharge (62/156) were observed in
25 monkeys in the 3.0 ppm treatment group versus no hoarseness or congestion and only five
26 observations of nasal discharge in 156 observations for control monkeys. Increased nasal
27 congestion was noted in the two lower treatment groups of monkeys: 30/156 and
28 45/156 observations, respectively, versus 9/156 observations in nasal discharge in the controls.
29 The authors reported an increase in nasal discharge and lacrimation in treated hamsters but no
30 increases in symptoms in rats. However, observations of adverse symptoms in the control rats
31 were greater than 10% on some measures.
32 Rhinitis increased in rats in the 3 ppm treatment group, and the incidence in controls was
33 notable (see Table 4-29). All groups of monkeys showed some rhinitis, and no treatment effects
34 were observed in either monkeys or hamsters. Monkeys and rats in the high treatment group
35 (3 ppm) had a greater incidence of lesions in the nasoturbinate epithelium (see Table 4-30).
This document is a draft for review purposes only and does not constitute Agency policy.
4-253 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
10
11
12
13
14
15
16
17
Rusch et al. (1983a, b) noted that most lesions were mild to moderate but were "somewhat more
severe" in the high treatment group. Hamsters did not exhibit a similar increase, with few
lesions noted in the nasal epithelium. Overall, these studies show a clear increase in adverse
health effects at 3 ppm for rats and monkeys, with no adverse effects seen in hamsters at this
treatment level or rats and monkeys at the lower concentrations (0.2 ppm and 1 ppm).
Table 4-29. Rhinitis observed in formaldehyde-treated animals; data pooled for
male and female animals
Trial 1:
I, Control
II, 0.2 ppm
III, 1 ppm
Trial 2:
IV, Control
V, 3 ppm
F344 rats
17/38
14/39
14/38
12/40
25/39
Cynomolgus monkeys
4/6
4/6
5/6
2/6
4/6
Syrian golden hamsters
0/14
0/4
0/11
0/9
2/16
Source: Rusch et al. (1983a, b).
Table 4-30. Epithelial lesions found in the middle region of nasoturbinates of
formaldehyde-treated and control animals; data pooled for males and
females
Trial 1:
I, Control
II, 0.2 ppm
III, 1 ppm
Trial 2:
IV, Control
V, 3 ppm
F344 rats
Basal cell
hyperplasia
0/38
0/38
0/36
4/39
25/37
Squamous metaplasia/
hyperplasia
2/38
1/38
3/36
3/39
23/37
Cynomolgus monkeys
Squamous metaplasia/
hyperplasia
0/6
0/6
1/6
0/6
6/6
Syrian golden hamsters
Nasal epithelium
No lesions noted
No lesions noted
18
19
20
21
22
23
24
25
Source: Rusch et al. (1983a, b).
Andersen et al. (2008) examined the effect of formaldehyde exposure at several
concentrations and durations. This study comprised histopathology and cell proliferation data, as
well as genomic analyses at Level II of the nasal cavity. Toxicogenomics analysis was
performed only at Level II because this was the region where the most severe lesions have been
This document is a draft for review purposes only and does not constitute Agency policy.
4-254 DRAFT—DO NOT CITE OR QUOTE
-------
1 reported in chronic bioassays (Andersen et al., 2008; Monticello et al., 1991; Kerns et al., 1983).
2 More specifically, Andersen et al. (2008) stated that the histopathologic and cell proliferation
3 effects at Levels II and III (with similar tissue structure) (Monticello et al., 1991) provided
4 phenotypic anchoring for the genetic analysis. Table 4-31 summarizes many of the broad
5 phenotypic findings.
6 The primary conclusions of this study with regard to the histopathology and cell
7 proliferation are as follows:
8
9 • The presence of inflammatory cell infiltrates in the nasal epithelial tissue of F344 rats is
10 highly variable and provides no coherent pattern with dose or duration at levels below
11 6 ppm.
12 • Hyperplasia was observed following exposure to >2 ppm.
13 • Metaplasia was observed at 6 ppm on day 5, but not before or after.
14 • Cell proliferation (as measured by labeling indices) was significantly elevated in
15 Levels I-III at 6 ppm on day 5 and Level I on day 15, leading to the conclusion that
16 significant changes in cell proliferation may not occur at exposures to <2 ppm.
17 • A significant decrease in cell density was observed at Level I in animals exposed to
18 6 ppm formaldehyde for 15 days, which was posited to be related to tissue remodeling in
19 response to this concentration.
20
21 Based on their analysis of the microarray data, Andersen et al. (2008) concluded that no
22 genes were significantly altered by exposure to 0.7 ppm from 1 to 15 days. Exposure to 2 ppm
23 primarily resulted in gene changes at 5 days of exposure, but not thereafter. One gene was
24 significantly increased on day 1, but the authors did not identify that gene. At 6 and 15 ppm,
25 42 and 745 genes were altered at day 1, respectively. After 5 days, gene changes were only
26 observed at 6 ppm (15 ppm was not examined after day 1). These findings support conclusions
27 reached by their laboratory in an earlier analysis. Thus, the primary conclusion in the Andersen
28 et al. (2008) study is that genomic changes, including those suggestive of mutagenic effects, did
29 not temporally precede or occur at lower doses than phenotypic changes in the tissue. The
30 implications of this finding will be examined later in Section 4.5.
31 Studies have also investigated the ability for formaldehyde to induce pathology in the
32 trachea, bronchi, and lung tissue. These studies have reported tracheal tissue changes, lung
33 inflammation, necrosis, changes to the biochemistry of BAL fluid and lung surfactant in a variety
This document is a draft for review purposes only and does not constitute Agency policy.
4-255 DRAFT—DO NOT CITE OR QUOTE
-------
it
& s
I-
Table 4-31. Cellular and molecular changes in nasal tissues of F344 rats exposed to formaldehyde
"X. PPm
Response\
I
H
M
PI
P2
P3
CD
G
Dl
0
0
0
0
-
0.7
1
0
0
0
2
6
0
0
1
6
8
0
0
42
15
-
-
-
745
D1R
0
4
0
0
-
0.7
2
1
0
0
2
1
3
0
0
6
7
8
0
0
15
-
-
-
-
D5
0
1
0
0
39±9
-
-
321±30
-
0.7
1
0
0
37±15
-
-
336±64
0
2
5
3
0
65±40
-
-
377±141
15
6
8
8
7
155±89a
a
a
400±61
28
15
-
-
-
-
D6
0
5
0
0
-
0.7
2
0
0
0
2
4
1
0
0
6
7
8
0
9
15
-
-
-
-
D6R
0
6
0
0
-
0.7
1
0
0
-
2
3
2
0
-
6
7
8
0
-
15
-
-
-
-
D15
0
3
0
0
79±55
-
-
362±61
-
0.7
1
0
0
56±37
-
-
340±57
0
2
0
2
0
51±44
-
-
321±37
0
6
5
7
0
119±38a
-
-
293±53b
54
15
-
-
-
5'
3 ,
J^.
K>
fe
H
D = day; R = recovery.
I = infiltrations (number out of 8 total animals); H = hyperplasia (number/8); M = metaplasia (number/8).
P1-P3 = proliferation at levels I-III (ULLI).
CD = cell density (cells/mm) at Level I.
G = genes significantly altered at Level II of nasal epithelial tissue.
"Significantly elevated ULLI and LI at Level I on day 5 or significantly elevated 1LI1 at Level I on day 15; indexa without numerical value indicates significant
increases in ULLI in all subregions of Levels II and III at day 5.
bStatistically significant difference from control (p < 0.05).
Source: Andersen et al. (2008).
O
o
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
of species. Ozen et al. (2003a) noted changes in zinc concentration in the lung tissue following
exposure for formaldehyde. Dallas et al. (1989) and Dinsdale et al. (1993) observed changes in
P450 enzyme activity in the lung associated with formaldehyde exposure.
Ozen et al. (2003a) measured zinc, copper, and iron content in lung tissue from
formaldehyde-exposed Wistar rats. Adult male rats were exposed to 0, 5, or 15 ppm (0, 6.2, or
18.5 mg/m3) formaldehyde 8 hours/day, 5 days/week for either 4 or 13 weeks. Rats were
checked daily and weighed weekly. At sacrifice, rats were autopsied and examined for gross
pathological changes. Lung tissue was homogenized and analyzed for zinc, copper, and iron.
Body weight gain was depressed in all treatment groups in a concentration-dependent
manner (p < 0.001) (see Table 4-32). Formaldehyde-exposed rats consumed less food and water
than controls and showed unsteady breathing, increased nose cleaning, excessive licking,
frequent sneezing, and nasal mucosa hemorrhages. Significant decreases were seen in the zinc
content of lungs after either 5 or 10 ppm formaldehyde exposure (see Table 4-33). Copper
content was unchanged from controls in all treatment regimens, whereas iron content was
increased after 4 weeks of 5 ppm exposure and after 13 weeks of either 5 or 10 ppm
formaldehyde exposure (Ozen et al., 2003a).
Table 4-32. Percent body weight gain and concentrations of iron, zinc, and
copper in cerebral cortex of male Wistar rats exposed to formaldehyde via
inhalation for 4 and 13 weeks
Exposure (mg/m3)
Weight gain (%)a
Zinc (mg/kg)a
Copper (mg/kg)a
Iron (mg/kg)a
4-week data
0
6.1
12.2
20.11 ±2.87
7.27±1.49e
5.24±1.52e
120 ±6.03
130±7.26C
185±10.36e
4.60 ± 0.42
5.60±0.50b
5.80±0.60d
25.07 ±2.83
23.00 ±2.32
22.14 ±1.95b
13-week data
0
6.1
12.2
60.53 ±7.84
38.41 ±2.53e
25.87 ±1.32e
123 ±6.22
155±7.94e
163±6.03e
4.67 ±0.38
5.41±0.56C
6.10±0.73e
24.92 ±2.84
22.00 ±2.41
21.00 ±1.96b
22
23
24
25
26
27
28
29
30
31
aValues are means ± SDs (n = 7).
Statistical significance of differences versus controls, as calculated by the authors:
V < 0.05.
°p < 0.02.
dp < 0.002.
>< 0.001.
Source: Ozenetal. (2003b).
This document is a draft for review purposes only and does not constitute Agency policy.
4-257 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-33. Zinc, copper, and iron content of lung tissue from formaldehyde-
treated male Wistar rats
Concentration
Oppm
5 ppm
lOppm
Oppm
5 ppm
10 ppm
Duration"
Control
4 weeks
4 weeks
Control
13 weeks
13 weeks
Zincb'c
20.7 (1.6)
16.1(1.3)d
13.8(1.2)e
20.0 (1.6)
15.3 (1.4)e
13.0(l.l)e
Copper"'0
0.39 (0.05)
0.32 (0.07)
0.36 (0.04)
0.39 (0.05)
0.37 (0.04)
0.39 (0.05)
Ironb'c
12.5 (0.8)
12.9 (1.0)
17.5 (1.3)e
12.7 (0.4)
17.9(l.l)e
22.4 (1.4)e
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
aRats were exposed 8 hours/day, 5 days/week for the number of weeks indicated.
bConcentrations are expressed as moles/mg of tissue, wet basis.
°Values are means (» = 7); SDs shown in parentheses.
dp < 0.005, compared with controls, as calculated by authors.
ep < 0.001, compared with controls, as calculated by the authors.
Source: Ozenetal. (2003a).
There are two reports of lung cytochrome P450 levels after formaldehyde exposure. The
first report by Dallas et al. (1989) describes concentration- and duration-dependent changes in P450
levels. Male Sprague-Dawley rats were exposed at 0, 0.5, 3.0, or 15 ppm (0, 0.62, 3.7, or 18.5
mg/m3) formaldehyde 6 hours/day, 5 days/week for 1 day, 4 days, 12 weeks, or 24 weeks. There
were six rats in each exposure group, but the experiment was run in two parts, with three rats in
each subgroup. Rats were sacrificed after 1 day, 4 days, 12 weeks, or 24 weeks of exposure, and
liver microsomes were prepared. TP and P450 content were determined on each sample.
Average P450 levels in control groups ranged from 17-76 pmol P450/mg protein.
However, no P450 was detected in lung from formaldehyde-treated animals after 1 day of
exposure, with a method detection limit of approximately 10 pmol P450/mg protein. In contrast,
P450 levels were elevated significantly above controls in a concentration-dependent manner after
4 days of formaldehyde exposure (see Table 4-34). Although P450 levels remained elevated in
some experimental groups after 12 and 24 weeks of exposure, results were variable and less
dramatic.
A later study by Dinsdale et al. (1993) attempted to confirm the increase in P450 levels
reported by Dallas et al. (1989). In their first experiment, Dinsdale et al. (1993) treated male
Sprague-Dawley rats at approximately 10 ppm (12.3 mg/m3) formaldehyde 6 hours/day for
4 days. The formaldehyde vapor was generated from formalin by a concentric jet atomizer. For
the second experiment, Dinsdale et al. (1993) similarly exposed rats to formaldehyde, but the gas
This document is a draft for review purposes only and does not constitute Agency policy.
4-258 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-34. Total lung cytochrome P450 measurements of control and
formaldehyde-treated male Sprague-Dawley rats
Formaldehyde
Oppm
0.5 ppm
3.0 ppm
15 ppm
1 Daya'b
Expt 1
17(6)
ND
ND
ND
Expt. 2
44(13)
ND
ND
ND
4 Days
Expt. 1
39(11)
103 (52)
357 (10)e
362 (38)e
Expt. 2
23(3)
137 (14)e
278 (100)e
334 (4)e
12 Weeks
Expt. 1
29 (10)
87(ll)d
91 (10)d
130 (2)e
Expt. 2
19 (23)
35(7)
67 (34)
56(6)
24 Weeks
Expt. 1
76 (49)
172 (12)c
92 (103)
151 (9)
Expt. 2
18(11)
38(9)
30(15)
48 (7)c
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
aRats were exposed 6 hours/day, 5 days/week for the duration shown.
bCytochrome P450 expressed as pmol P450/mg of protein. Values are means (SDs) (n = 3).
'Different from control, p < 0.05.
Different from control, p < 0.01.
"Different from control, p < 0.001, as calculated by the authors.
ND = not detected above the limit of detection, approximately 10 pmol/mg protein.
Source: Dallas et al. (1989).
was generated by the thermal depolymerization of paraformaldehyde as was done by Dallas et al.
(1989). The concentration of P450 and activity of several P450 isozymes were measured in lung
microsomes (pentoxyresorufm O-dealkylase, benzyloxyresorufin O-dealkylase, ethoxyresorufin
O-dealkylase, and 2-aminofluorene 7V-hydroxylation). ALP and y-glutamyl transpeptidase
activity were measured in BAL fluid collected from each animal. No changes were seen in BAL
enzyme activity or the activity of lung microsomes for the P450 substrates tested. Cytochrome
P450 levels were unchanged in experiment 1, where formaldehyde was generated from formalin.
Cytochrome P450 levels were increased in experiment 2 with formaldehyde generated from
paraformaldehyde (see Table 4-35).
Table 4-35. Cytochrome P450 levels in formaldehyde-treated rats
Group
Control
10 ppm formaldehyde
Experiment 1
(formalin)3
Experiment 2
(paraformaldehyde)11
(nmol/mg protein)
82 ±30
73 ±27
85 ±5
125±23b
aValues are means ± SDs (n = 3-5).
bDifferent from controls,/? < 0.05.
Source: Dinsdale et al. (1993).
This document is a draft for review purposes only and does not constitute Agency policy.
4-259 DRAFT—DO NOT CITE OR QUOTE
-------
1 Extrapulmonary toxicity
2 Several studies have investigated toxicity in organs other than those associated with the
3 respiratory tract. An earlier cross-species study examined changes in lung tissue resulting from
4 continuous exposure (Coon et al., 1970). Animals were exposed to 3.7 ppm (4.6 mg/m3)
5 formaldehyde for 90 days. Five species of animals were studied: male and female Sprague-
6 Dawley and Long-Evans derived rats (15), male and female Princeton-derived guinea pigs (15),
7 male New Zealand albino rabbits (3), male squirrel monkeys (Saimiri sciureus) (3), and purebred
8 male beagle dogs (2). Blood samples were taken for Hb concentration, HCT, leukocyte counts,
9 and serum levels of BUN, AST, ALT, ALP, and LDH. Sections of heart, lung, liver, kidney, and
10 spleen were fixed and examined from each species (details of method not provided). Brain,
11 spinal cord, and adrenal tissue also were examined in monkeys and dogs as well as thyroid from
12 dogs. Liver and kidney sections were stained for reduced nicotinamide adenine dinucleotide,
13 lactate, isocitrate, and p-hydroxybutyrate. Tissue sections of the nasal mucosa were not
14 examined in this study.
15 Hematological parameters were unaffected by formaldehyde treatment. The lung tissue
16 of all species exhibited interstitial inflammation after 90 days of formaldehyde exposure
17 (detailed description not provided). Formaldehyde-treated rats and guinea pigs also had focal
18 chronic inflammation in heart and kidney tissue sections. However, the authors were uncertain
19 whether the observed changes to heart and kidney were due to formaldehyde exposure.
20 As mentioned above, Woutersen et al. (1987) exposed male and female albino SPF
21 Wistar rats (10/group) to 0, 1, 10, or 20 ppm (0, 1.23, 12.3, or 24.6 mg/m3) formaldehyde
22 6 hours/day, 5 days/week for 13 weeks. Rats were checked daily and weighed weekly. During
23 week 13, blood samples were taken for Hb, PCV, RBC count, and a differential count of
24 leukocytes. Urine samples were also analyzed. At sacrifice, blood samples were analyzed for
25 ALB, creatinine, glucose, TP, BUN, and the enzyme activities (AST, ALT, and ALP). GSH and
26 protein content were determined in liver homogenates. Organs were examined and weighed:
27 adrenals, brain, heart, kidneys, liver, lungs, ovaries, pituitary, spleen, testes, thymus, and thyroid.
28 No gross pathological changes were seen upon autopsy, but body weights decreased in
29 both male and female rats at the 20 ppm treatment level. Of the organs weighed, 6 of 11 had
30 significantly increased relative rates in male rats exposed to 20 ppm formaldehyde. Relative
31 brain weight was increased in female rats at the same treatment level (Woutersen et al., 1987).
32 Clinical chemistry parameters of liver and kidney function and hematological parameters
33 were also measured after the 13-week treatment by Woutersen et al. (1987). Compared with
34 those of controls, activities of AST, ALT, and ALP were significantly elevated in plasma from
35 the 20 ppm treated male rats (by 124, 132, and 126%, respectively;/* < 0.05). Total plasma
This document is a draft for review purposes only and does not constitute Agency policy.
4-260 DRAFT—DO NOT CITE OR QUOTE
-------
1 protein was reduced to 95% of controls in the same animals. Although there was an observed
2 increase in BUN in male rats treated with 1 ppm, this was not considered a treatment effect.
3 Furthermore, no statistically significant differences were seen for these parameters in female rats
4 at any concentration level (Woutersen et al., 1987).
5 Sul et al. (2007) exposed Sprague-Dawley rats to 0, 5, and 10 ppm formaldehyde for
6 6 hours/day (5 days/week) for 2 weeks and collected lung samples for tissue damage and
7 genomic analysis. According to their results, 21 genes were altered in a dose-dependent manner
8 by microarray analysis; 2 were up regulated and 19 were down regulated in the lung tissue of
9 animals exposed to formaldehyde. However, six of the nine genes further analyzed by PCR did
10 not show dose dependency (authors did not comment). Although the authors briefly describe the
11 functions and potential implications for changes in the expression of some of the altered genes,
12 there is no discussion of the relationship between these altered genes (i.e., there is no pathway
13 analysis).
14 In 2006, Im et al. (2006) published a proteomic analysis using the same exposure
15 protocols (possibly using the same animals as in the Sul et al. [2007] study, although neither
16 study makes reference to the other). Im et al. (2006) examined DNA damage in lymphocytes
17 and liver tissues, as well as protein and lipid oxidation in plasma and liver samples. Similar to
18 changes reported in the lung (discussed elsewhere), using two-dimensional electrophoresis and
19 matrix-assisted laser desorption ionization time-of-flight mass spectrometry, the authors also
20 reported dose-dependent changes in the levels of 32 proteins in plasma (19 up, 13 down). None
21 of the changes in plasma proteins correspond to the changes in lung reported by Sul et al. (2007).
22 Again, no pathway analysis was provided. Interestingly, Im and colleagues (2006) also
23 demonstrated a dose-dependent increase in plasma IL-4 and dose-dependent decrease in IFNy,
24 perhaps indicative of Th-2-mediated inflammatory response. An overview of formaldehyde
25 exposure-related pathology in the respiratory system of laboratory animals is presented in
26 Table 4-36.
27
28 4.2.2.5. Chronic Inhalation Bioassays
29 The respiratory pathology observed in chronic bioassays is consistent with the subchronic
30 studies. As exposure concentration and duration of exposure are increased, the pathology
31 becomes more severe and penetrates more deeply into the respiratory tract. These effects are
32 progressive over time. Tumors are reported in several bioassays, primarily SCCs. Experimental
33 results regarding both the severity of respiratory tract pathology as well as the tumor incidence
34 vary by species strain and experimental design. As discussed above rodents experience KB, and
35 species differences in respiratory and physiological depression would result in differences in
This document is a draft for review purposes only and does not constitute Agency policy.
4-261 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-36. Summary of respiratory tract pathology from inhalation exposures to formaldehyde, subchronic studies
Species/strain
No./
group
Treatment"
Respiratory effects
LOAEL/NOAEL
Reference
Nasal path ology
B6C3F1 mice
(male and
female)
SPF Wistar Rats
(male and
female)
SPR Wister rats
(male)
Wistar rats (male
and female)
Wistar rats (male)
Wistar rats (male)
F344 rats (male
and female),
Syrian golden
hamsters (male
and female),
cynomolgus
monkeys
10
10
50-55
50
10
10
20 rats,
10
hamsters,
6 monkeys
0, 2, 4, 10, 20, or 40 ppm
formaldehyde 6 hours/day,
5 days/week for 13 weeks
0, 1, 10, or 20 ppm
formaldehyde 6 hours/day, 5
days/week for 13 weeks
0, 10, or 20 ppm formaldehyde
for 6 hours/day, 5 days/week
for 4, 8, or 13 weeks
0,0.3, 1, or 3.0 ppm
formaldehyde 6 hours/day,
5 days/week for 13 weeks
40 ppm-hours (8 hours at
5 ppm, 4 hours at 10 ppm) or
80 ppm hours (10 ppm
continuous or 20 ppm
intermittently)
0, 8, or 16 ppm, given either
continuously or intermittently
0.0.2, 1, or 3 ppm
22 hours/day, 7 days/week,
26 weeks
Minimal squamous metaplasia in 1 of 10 mice
(4 ppm). Squamous metaplasia observed in all mice
at 10 and 20 ppm.
Increased respiratory epithelial hyperplasia and
keratinization at 20 ppm; squamous metaplasia at
10 ppm in males and females.
Rhinitis, hyperplasia, and squamous metaplasia in
respiratory epithelium at all doses (number of weeks
not specified).
Squamous metaplasia of olfactory epithelium at
20 ppm (number of weeks not specified)
Keratinized and nonkeratinized squamous
metaplasia in level 2 epithelium in 37/50 male and
21/50 female rats at 3 ppm for 13 weeks.
Thinning and disarrangement of respiratory
epithelium, squamous metaplasia, most severe in
20 hours intermittent exposure
Disarrangement and squamous metaplasia at 4 ppm.
Continuous exposure yielded less severe lesions
than intermittent exposure
Rats: rhinitis at 3 ppm, increased incidence of nasal
lesions at 3 ppm.
Monkeys: rhinitis at all doses, increased incidence
of nasal lesions at 3 ppm.
Hamsters: no significant nasal lesions.
NOAEL = 4 ppm
NOAEL = 1 ppm
NOAEL = 1 ppm
NOAEL = 1 ppm
NOAEL = 10 ppm
LOAEL = 8 ppm
NOAEL = 1 ppm
Maronpot et al.
(1986)
Woutersen et al.
(1987)
Feronetal. (1988)
Zwartetal. (1988)
Wilmer et al.
(1987)
Wilmer et al.
(1989)
Rusch et al.
(1983a, b)
Tracheal and lung pathology
Wistar rats (male)
6
0, 5, 15 ppm for 8 hours/day,
5 days/week, 4 or 13 weeks
Significant decreases in zinc content in lung, copper
unchanged, iron increased in lung.
LOAEL = 5 ppm
Ozen et al. (2003)
o
O
II
§
s
rS"
-
I
to
a\
to
-------
o
o
II
§
TO
Table 4-36. Summary of respiratory tract pathology from inhalation exposures to formaldehyde, subchronic studies
(continued)
Species/strain
Sprague-Dawley
rats (male)
Sprague-Dawley
rats (male)
No./
group
6 but n = 5
in some
trials
5
Treatment"
0,0.5,3.0, 15ppm6
hours/day, 5 days/week, for 1
day, 4 days, 12 weeks, 24
weeks
0 or 10 ppm 6 hours/day,
4 days using both formalin and
paraformaldehyde
Respiratory effects
Increased P450 levels after 4 days at 3 ppm.
P450 levels increased at 10 ppm only in groups
treated with paraformaldehyde.
LOAEL/NOAEL
NOAEL = 0.5 ppm
LOAEL = 10 ppm
Reference
Dallas et al.
(1989)
Dinsdale et al.
(1993)
Extrapulmonary effects
Rats and guinea
pigs
SPF Wistar rats
(male and
female)
15
10
3. 7 ppm for 90 days
0, 1, 10, or 20 ppm
6 hours/day, 5 days/week for
13 weeks
Focal chronic inflammation in heart and kidney
tissue.
Relative brain weight increased in female rats at
20 ppm; increased AST, ALT, ALP in plasma at
20 ppm.
LOAEL = 3.7 ppm
NOAEL = 10 ppm
Coon etal. (1970)
Woutersen et al.
(1987)
-
I
J^.
K>
-------
1 absorbed dose in the respiratory tract, given the same exposure concentration (Chang and
2 Barrow, 1983). Additionally, differences in the nasal architecture result in species-dependent
3 variation of formaldehyde absorption (flux) within the respiratory tract (see Section 3.4).
4 Therefore, chronic studies are discussed by species for greater clarity.
5
6 4.2.2.5.1. Mice.
1 Early experiments by Horton et al. (1963) subjected mice (C3H, sex unspecified) to
8 extreme formaldehyde concentrations (0, 0.05, 0.1, and 0.2 mg/L or 41-163 ppm) in an attempt
9 to simulate lung pathology reported in humans exposed to cigarette smoke. The mice were
10 exposed 1 hour/day, 3 days a week for up to 35 weeks. The authors did not note the effects of
11 KB or provide any information on pathology of the URT. There was a clear increase in
12 histologic changes in the tracheobronchial epithelium by exposure, including basal-cell
13 hyperplasia, stratification squamous cell metaplasia and atypical metaplasia. Subsequent
14 exposures to various combinations of formaldehyde and coal tar did result in squamous cell
15 tumors. The findings of Horton et al. (1963) suggest a role for formaldehyde in lung cancer
16 under some conditions. However, the exposure design and early deaths in the treatment groups
17 severely limit the usefulness of these data in human health risk assessment.
18 In a comprehensive study conducted by Swenberg et al. (1980) (also reported in Kerns et
19 al. [1983]) in conjunction with Chemical Industry Institute of Toxicology (CUT) and Battelle
20 Columbus Laboratories, male and female C57BL/6 x C3H FI (B6C3Fi) mice (approximately
21 120/sex/concentration) were exposed to 0, 2.0, 5.6, or 14.3 ppm (0, 2.45, 6.87, or 17.5 mg/m3)
22 formaldehyde 6 hours/day, 5 days/week for 24 months. This exposure period was followed by
23 up to 6 months of nonexposure to evaluate recovery. Interim sacrifices were conducted at 6, 12,
24 18, 24, 27, and 30 months (due to unscheduled deaths, no male mice were sacrificed at 18 or
25 27 months). Exposure generation was accomplished by sublimation of paraformaldehyde, and
26 exposures were conducted in whole-body chambers. Detailed sectioning and examination of the
27 nasal passages were conducted at each interim sacrifice, beginning at 12 months, and for all
28 unscheduled deaths. Gross organ pathology was noted for all animals and complete
29 histopathologic examination was conducted on all animals in the control and high-exposure
30 groups. There were no differences in survival in any exposure group compared with controls.
31 Generally, poor survival in all groups of male mice was attributed to fighting and infections of
32 the urogenital tract associated with group housing; 78, 77, 81, and 82 unscheduled deaths were
33 recorded before 24 months in the 0, 2.0, 5.6, and 14.3 ppm treatment groups, respectively (
34 n= 119, 120, 120, and 119 males, respectively). After the interim sacrifices (6 and 12 months)
35 only 17-22 male mice survived to the 24-month scheduled sacrifice. Female mice had much
This document is a draft for review purposes only and does not constitute Agency policy.
4-264 DRAFT—DO NOT CITE OR QUOTE
-------
1 greater survival with only 30, 34, 19, and 34 unscheduled deaths prior to the 24-month sacrifice.
2 The authors did not note the effects of KB in mice, although the RDso for a 10-minute exposure
3 for male B6C3F1 mice has been reported at 4.9 ppm and 4.4 ppm (Steinhagen and Barrow, 1984;
4 Chang et al., 1981).
5 The first examination of the nasal cavities was conducted at the 12-month interim
6 sacrifice. Inflammation in the nasal turbinates was evident in mice in the 2 and 6 ppm treatment
7 groups (14/20 and 18/20, respectively), including adenitis of the nasal lacrimal duct, lacrimal
8 duct, and vomeronasal gland. Inflammation was not present in mice exposed at 15 ppm,
9 although serous rhinitis was seen in 4 of 20 animals. At 18 months, mice exposed at 2 and
10 6 ppm no longer exhibited adenitis in the nasoturbinates. Epithelial dysplasia was evident in 4 of
11 20 mice at 6 ppm exposure. Mice in the high-exposure group had significantly greater nasal
12 pathology; epithelial dysplasia and squamous metaplasia were reported in 18/19 and 17/19
13 female mice, respectively, exposed to 15 ppm. After 24 months, squamous epithelial hyperplasia
14 of the nasolacrimal duct (29/45) and atrophy of the olfactory epithelium (18/45) were also noted
15 in animals from the high-exposure group (male and female) (Battelle Columbus Laboratories,
16 1981). Similar pathology was reported in only a small fraction of mice exposed at 2 and 6 ppm
17 (5/48 and 11/60, respectively).
18 Three months after cessation of exposure, only nine female mice were available for
19 sacrifice, but within this small sample the data suggested recovery of nasal lesions: epithelial
20 dysplasia (4/9), squamous metaplasia (2/9), atrophy of the olfactory epithelium (1/9), and
21 squamous epithelial hyperplasia of the nasolacrimal duct (1/9) (Battelle Columbus Laboratories,
22 1981).
23 Of the 17 male mice that survived to 24 months in the 14.3 ppm exposure group, 2 had
24 SCC in the nasal cavity (p < 0.05). Of the two tumor-bearing mice, one exhibited significant
25 epithelial pathology, including rhinitis, dysplasia, squamous metaplasia, and hyperplasia.
26 Squamous metaplasia of the nasolacrimal duct was the only related pathology reported for the
27 second mouse. No SCCs were found in female mice, although 48 mice survived to 24 months.
28 The authors reported no other formaldehyde-related tumors. However, comparisons were based
29 on summary tables by organ. Although lymphomas were analyzed by organ and site (e.g.,
30 increase in salivary gland lymphoma considered separately from mandibular lymphoma), later
31 reanalysis of lymphoma in female mice, based on tumor-bearing animals (TBAs), does indicate
32 an association with formaldehyde exposure.
33
This document is a draft for review purposes only and does not constitute Agency policy.
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
4.2.2.5.2. Rats.
Holmstrom et al. (1989a) evaluated coexposure of inhaled formaldehyde with wood dust
in 16 female Sprague-Dawley rats/group. Rats were exposed in whole-body chambers for 6
hours/day, 5 days/week for 104 weeks to formaldehyde alone at 12.4 ±1.1 ppm (15.21 ± 1.35
mg/m3), wood dust alone (25 mg/m3), or both wood dust (25 mg/m3) and formaldehyde (12.7 ±
1.0 ppm) or to room air as the control. The wood dust was generated from grinding of beech.
Microscopic measurements of the wood particles indicated that approximately 70% had a
geometric diameter of about 10 um, while 10-20% were about 5 um or less. Animals were
sacrificed at 104 weeks and histopathology was performed on five transverse sections of the
nasal cavity (see Figure 4-6) and the lungs (not otherwise specified).
There were no differences in mortality among the groups at any time during the study
period. Rats exposed to formaldehyde were reported to have exhibited yellow discoloration of
the fur, and many displayed eye irritation. Formaldehyde exposure, with and without wood dust,
induced squamous metaplasia, keratinization, and dysplasia of the nasal epithelium (see
Table 4-37).
Table 4-37. Histopathologic findings and severity scores in the naso- and
maxilloturbinates of female Sprague-Dawley rats exposed to inhaled
formaldehyde and wood dust for 104 weeks
21
22
23
24
25
26
27
Treatment
Formaldehyde
group (n = 16)
Formaldehyde-
wood dust group
(n = 15)
Wood dust group
(n = 15)
Control group
(n = 15)
Pronounced
squamous
metaplasia
7
7
0
0
Pronounced
squamous
metaplasia
with
keratinization
2
1
0
0
Pronounced
squamous
metaplasia
with
presence of
dysplasia
1
4
0
0
Sum of rats
with
pronounced
metaplasia
and/or
dysplasia
10
12
0
0
ccscc
1
0
0
0
Histologic scores
at the level of
naso- and
maxilloturbinates
(mean ± SD)
2.25±1.73a
2.6±1.88a
1.86±0.83b
1.07 ±0.70
><0.01.
V < 0.05.
Source: Holmstrom et al. (1989a).
This document is a draft for review purposes only and does not constitute Agency policy.
4-266 DRAFT—DO NOT CITE OR QUOTE
-------
1 Among the five levels of the nasal cavity that were examined, Holmstrom et al. (1989a)
2 presented findings for the naso- and maxilloturbinates since formaldehyde-induced tumors had
3 been associated with this level (Morgan et al., 1986a, b). The data also suggested an effect of
4 wood dust on formaldehyde-induced nasal pathology, with a slightly higher histologic score and
5 greater incidence of dysplasia than formaldehyde exposure alone. One SCC (1/16) occurred in
6 the group exposed to formaldehyde only but not in the group exposed to formaldehyde and wood
7 dust. Microscopic examination of the lungs revealed that emphysema (diagnostic criteria not
8 specified) was more prevalent in both groups exposed to wood dust compared with the control
9 group (p < 0.05). There was no significant difference in pulmonary epithelial histopathology
10 among the groups.
11 Tobe et al. (1985) also evaluated F344 rats (32/group) exposed to inhaled formaldehyde
12 for 28 months. Exposures were for 6 hours/day, 5 days/week to formaldehyde concentrations of
13 0, 0.3, 2, and 14 ppm (0, 0.37, 2.45, and 17.2 mg/m3). Fourteen of 32 rats (44%) in the high
14 concentration group developed nasal SCCs, compared with none in the other exposed groups and
15 the control group. Tobe et al. (1985) reported increased rhinitis, hyperplasia, and squamous
16 metaplasia of the nasal respiratory epithelium, including in the low-exposure group (0.3 ppm.)
17 However, some level of rhinitis, hyperplasia, and metaplasia were also present in controls.
18 Without a more complete report, it is unknown whether or not the pathology reported at 0.3 ppm
19 was a formaldehyde-related effect.
20 Kamata et al. (1997) evaluated the effects of inhaled formaldehyde in male F344
21 (F344/DuCrj) rats (32/group) exposed for 28 months. Formaldehyde exposure was generated by
22 metering 37% formalin (containing 10% methanol) into a sprayer in a glass bottle and diluting
23 with room air. Concentration in the chamber was monitored twice daily by the acetyl acetone
24 method. Exposures were for 6 hours/day, 5 days/week at nominal formaldehyde concentrations
25 of 0, 0.3, 2.0, and 15 ppm (0, 0.37, 2.45, and 18.4 mg/m3). Actual levels were 0, 0.3 ± 0.07, 2.17
26 ± 0.32, and 14.85 ± 2.22 ppm (mean ± SD). Rats in the 0 ppm group were given methanol to
27 inhale at the same concentration (4.2 ppm) as the 15 ppm group. A room control no-exposure
28 group was also included in the study. All animals were observed for clinical signs once a day
29 during the study. Body weights and food consumption were recorded weekly. Five animals per
30 group, randomly selected at the end of 12, 18, and 24 months, and all surviving animals at
31 28 months were sacrificed for hematological measurements (Hb, RBCs, PCV, MCV, mean
32 corpuscular hemoglobin [MCH], MCHC, and WBCs), biochemical determinations (TP, ALB,
33 BUN, ALP, AST, ALT, glucose, albumin/globulin ratio, phospholipids, triglycerides, and total
34 cholesterol), and pathological examinations. Wet weights were taken on brain, heart, lungs,
35 liver, kidneys, spleen, testes, and adrenal gland of each rat. Histopathology was performed on all
This document is a draft for review purposes only and does not constitute Agency policy.
4-267 DRAFT—DO NOT CITE OR QUOTE
-------
1 moribund or dead animals and those at specified sacrifices on all gross lesions and the following
2 tissues: pituitary, thyroid, nasal cavity, trachea, esophagus, stomach, small and large intestines,
3 prostate gland, urinary bladder, muscle, femur, sciatic nerve, spinal cord, and mesenteric lymph
4 nodes. Histopathologic sections of the nose were obtained from five anatomical levels, but these
5 did not correspond to the typical levels taken in other bioassays. Most notably, section level B
6 was anterior and not posterior to the incisor teeth. The incidence data for nasal histopathology
7 were not reported with respect to section level location, with the exception that the
8 nonproliferative lesions and tumors reported were described to occur predominantly at levels B
9 and C.
10 Yellow discoloration of the coats occurred in animals exposed at the 2 and 15 ppm levels.
11 Significant decreases in body weight and food consumption were observed in the high
12 concentration (15 ppm) group throughout the exposure period, and elevated mortality was noted
13 at 28 months (88.3 versus 31.8% in controls). The first death occurred after 6 versus 18 months
14 in the control group. Other effects noted in the 15 ppm exposure group include decreased
15 triglycerides, reduced liver weight (both relative and absolute), and increased relative adrenal
16 weights.
17 Treatment-related macroscopic and histopathologic findings were limited to the nasal
18 cavity. Squamous cell metaplasia was reported in all treatment groups: 16% (0.3 ppm), 37.5%
19 (2 ppm), and 91% (15 ppm) of exposed rats. Epithelial hyperplasia was similarly present in 12.5,
20 22, and 91% of the animals, respectively. Since a no-effect level could not be determined, the
21 authors reported benchmark doses (BMDs) of 0.25 and 0.24 ppm for squamous cell metaplasia
22 and epithelial hyperplasia (10% response.). Additional lesions only occurring in the 15 ppm
23 dose group were papillary hyperplasia (2/32), SCC (13/32), squamous cell papilloma (3/32), and
24 sarcoma (1/32). The majority of the tumors were located at levels B and C of the nasal cavity.
25 Albert et al. (1982) and Sellakumar et al. (1985) reported on a set of lifetime studies
26 performed in male Sprague-Dawley rats to evaluate the effects of inhaled formaldehyde alone
27 and in combination with hydrochloric acid (HC1). Rats were exposed 6 hours/day, 5 days/week
28 for life. In the first experiment (Albert et al., 1982), 8-week-old male inbred Sprague-Dawley
29 rats (n = 99) were exposed to a mixture of 10 ppm (12.3 mg/m3) HC1 and 14 ppm (17.2 mg/m3)
30 formaldehyde, and there were two control groups: air-sham and untreated (n = 50).
31 Bis(chloromethyl)ether (BCME), a known animal carcinogen (Albert et al., 1975; Kuschner et
32 al., 1975; Figueroa et al., 1973; Laskin et al., 1971), is formed when formaldehyde and HC1 are
33 mixed. BCME concentrations were estimated at about 1 ppb in the formaldehyde-HCl mixed
34 exposures, based on levels in the mixing chamber. Complete necropsies were conducted when
35 animals died naturally or were killed when moribund. Histologic sections were taken from the
This document is a draft for review purposes only and does not constitute Agency policy.
4-268 DRAFT—DO NOT CITE OR QUOTE
-------
1 nasal cavity, larynx, trachea, pulmonary lobes, liver, bladder, kidney, spleen, and other organs
2 with gross pathologic alterations.
3 Exposure to the mixed gases (formaldehyde-HCl-BCME) induced nasal lesions,
4 including epithelial hyperplasia (71%), squamous metaplasia (64%), squamous papilloma (3%),
5 and SCC (25%) (Albert et al., 1982). Although a few squamous metaplasias were noted in the
6 larynx, trachea, and bronchi, these lesions were also noted in controls. Mortality in exposed rats
7 was significantly increased over controls and was approximately 30% when the first carcinoma
8 was reported (233 days). Mortality in exposed rats rose quickly to approximately 60% after the
9 first year of exposure. Therefore, the authors used a life-table method to calculate a mortality -
10 corrected cumulative incidence, reporting a corrected cumulative incidence of 77% at 720 days
11 after first exposure.
12 In the second experiment performed in this laboratory (Sellakumar et al., 1985; Albert et
13 al., 1982), Sprague-Dawley rats were similarly exposed to HC1 (10 ppm) alone, formaldehyde
14 alone (15 ppm), or a combination of both. The combination exposure was generated in two
15 different ways to better understand the influence of BCME formation on study results: premixed
16 at high concentrations and gases fed separately into the inlet air supply at the target
17 concentrations. BCME concentration measured by a gas chromatography/mass spectrometry
18 method in the premixed chamber varied between 0.1 and 0.4 ppb. Cage-side observations and
19 necropsy procedures were as described in Albert et al. (1982) with the exception of the histologic
20 preparation of the head. The head was cut transversely into four tissue blocks, and sections were
21 taken from the face of each.
22 Animals exposed to formaldehyde alone and formaldehyde-HCl (premixed or
23 nonpremixed) showed a marked decrease in body weight after 16 weeks. After 32 weeks rats
24 exposed to the premixed formaldehyde-HCl (with BCME) had higher mortality compared with
25 the other mixed gas exposures (p < 0.05). Nasal pathology was similar among rats exposed to
26 formaldehyde alone or the mixed gases (see Table 4-38). Desquamation of respiratory epithelial
27 cells was reported in the respiratory epithelium that covers the nasomaxillary turbinates and the
28 nasal septum (approximately section levels 2 and 3). Olfactory epithelium in the ET frequently
29 showed an inflammatory reaction with seropurulent exudate filling the lumen. Squamous
30 metaplasia and hyperplasia were reported in the larynx and trachea in all treatment groups.
31
32
33
34
35
36
This document is a draft for review purposes only and does not constitute Agency policy.
4-269 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
Table 4-38. Histopathologic changes (including tumors) in nasal cavities of
male Sprague-Dawley rats exposed to inhaled formaldehyde or HC1 alone
and in combination for a lifetime
Observation
Number of animals examined
Rhinitis
Epithelial or squamous hyperplasia
Squamous metaplasia
Polyp or papilloma
sec
Adenocarcinoma
Mixed carcinoma
Fibrosarcoma
Esthesioneuroepithelioma
Premixed
HC1-HCHO
100
74
54
64
13
45
1
0
1
1
Nonpremixed
HC1-HCHO
100
75
53
68
11
27
2
0
0
0
HCHO
100
74
57
60
10
38
0
1
1
0
HC1
99
81
62
9
0
0
0
0
0
0
Air
99
72
51
5
0
0
0
0
0
0
Colony
99
70
45
6
0
0
0
0
0
0
Larynx
Hyperplasia
Squamous metaplasia
11
10
22
15
21
4
22
0
2
0
2
0
Trachea
Hyperplasia
Squamous metaplasia
18
9
32
8
21
7
26
0
6
0
2
0
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Source: Sellakumar et al. (1985).
Tumors arose primarily from the nasomaxillary turbinates and nasal septum. The SCCs
were predominantly moderate to well differentiated, with excessive amounts of keratin occluding
the lumen, killing the animals by asphyxiation. Statistical comparisons by the log rank test (Peto
test) showed that tumor incidence was increased in the premixed formaldehyde-HCl combined
exposure group over formaldehyde alone or the combined formaldehyde-HCl (not premixed).
There were no significant differences in the latency among groups, with the average latency
varying from 603 to 645 days. Rats exposed to HC1 exposure alone did not develop tumors.
The esthesioneuroepithelioma is a unique tumor type observed with a high incidence in
an earlier inhalation study of rats exposed to BCME (Kuschner et al., 1975), suggesting that the
higher incidence of nasal tumors observed in the premixed-combined formaldehyde-HCl-
exposure group may have been due to BCME (Krimsky, 1986) since this premixed protocol was
the one most likely to generate BCME. Sellakumar et al. (1985) refuted this assertion, stating
This document is a draft for review purposes only and does not constitute Agency policy.
4-270 DRAFT—DO NOT CITE OR QUOTE
-------
1 that this singular tumor occurred in the absence of other changes in the ethmoid region or in the
2 lungs where BCME was also demonstrated to cause tumors. Furthermore, exposure was
3 approximately one-tenth the cumulative dose in the Kuschner et al. (1975) study that was
4 associated with a single similar tumor. Sellakumar et al. (1985) attributed the higher incidence
5 in the premixed-combination group to traces of other alkylating agents (not BCME) that could
6 have been formed. The results demonstrate that animals exposed to either a combination of
7 formaldehyde-HCl or to formaldehyde alone develop nasal tumors, principally SCCs, at about
8 the same frequency, indicating that HC1 plays little or no role in the carcinogenicity of inhaled
9 formaldehyde.
10 In a companion study to the chronic mouse study described above (Kerns et al., 1983;
11 Swenberg et al., 1980), groups of F344 rats (approximately 120/sex/concentration) were exposed
12 to 0, 2.0, 5.6, or 14.3 ppm (0, 2.45, 6.87, or 17.5 mg/m3) formaldehyde 6 hours/day, 5 days/week
13 for 24 months. This exposure period was followed by up to 6 months of nonexposure to evaluate
14 recovery. Interim sacrifices were conducted at 6, 12, 18, 24, 27, and 30 months. Study
15 parameters and methods were as described above.
16 Formaldehyde exposure increased mortality of both male and female rats in all treatment
17 groups (p < 0.05 for 6 and 15 ppm groups). Severe treatment-related mortality was seen at the
18 highest exposure group beginning at 12 months with only 30% surviving to the 24 months.
19 There were no alterations in clinical chemistry, neurofunctional, or ophthalmological
20 measurements considered to be related to formaldehyde exposure. A concentration-dependent
21 increase in yellow discoloration of the hair coat was observed. This discoloration dissipated over
22 the 3-month postexposure period. Rats in the highest-concentration group were dyspneic
23 (p < 0.01) and emaciated (p < 0.05) and had many facial swellings that on closer examination
24 were revealed to be carcinomas protruding through the nasal cavity. Neoplastic lesions in the
25 URT were first observed clinically at day 358 in females and day 432 in males.
26 Macroscopically, these lesions originated in the anterior portion of the nasal cavity and, in a few
27 instances, extended into the ETs.
28 Figure 4-13 shows the frequency of squamous metaplasia by location in the noses of rats
29 sacrificed at various time points along the 2-year exposure period. Histopathologic lesions were
30 confined to the nasal cavity and proximal trachea in concentration-dependent fashion. The
31 morphologic diagnosis of squamous metaplasia was used to designate zones of altered
32 epithelium that were characterized by the presence of a well-differentiated germinal layer
33 (stratum germinativum) and superficial layers of epithelium (stratum spinosum and stratum
34 corneum). Keratin was produced only in areas of squamous metaplasia. Epithelial dysplasia was
35 detected earlier than squamous metaplasia and was characterized by a mucosa that had
This document is a draft for review purposes only and does not constitute Agency policy.
4-271 DRAFT—DO NOT CITE OR QUOTE
-------
1 undergone a transition from nonciliated simple cuboidal to one that was several cells thick and
2 squamoid with an organization and polarity of the individual cells that had changed from vertical
3 to horizontal with respect to the basement membrane. Similar histomorphologic changes have
4 also been called basal cell hyperplasia and epidermoid metaplasia (e.g., Albert et al. [1982]).
5 Figure 4-13 clearly illustrates that concentration is a dominant determinant of lesion distribution.
6 At low concentrations the lesions occur only in the most anterior region (cross-section level 1).
7 At 5.6 ppm, the squamous metaplasia in levels 1, 2, and 3 was also associated with purulent
8 rhinitis and epithelial dysplasia. At the highest concentration, the lesions progress to the more
9 distal URT, with lesions evident in level 5 and no difference in the incidence at level 1 or 2
10 across the various sacrifice times. Statistically significant (p < 0.05) regression of the lesion was
11 evident at most locations at the 27-month sacrifice (3 months postexposure) (e.g., level 1 in the
12 2 ppm group, all levels of the 5.6 ppm group, and levels 4 and 5 of the 14.3 ppm group).
13 Furthermore, progression of lesions distally to the lower respiratory tract (LRT) occurred
14 only in the high concentration group. Tracheal pathology observed at 18 months included
15 multifocal areas of minimal to mild epithelial hyperplasia, epithelial dysplasia, or squamous
16 metaplasia. There were no significant tracheal lesions present in the 0, 2.0, or 5.6 ppm exposure
17 groups, and tracheal lesions were not observed during the postexposure period in the 14.3 ppm
18 exposure group.
19 Table 4-39 provides the summary data of all neoplastic lesions in the nasal cavity of
20 exposed rats. The adjusted cumulative incidence rates of SCC in male and female rats from the
21 14.3 ppm exposure group at 24 months were 67 and 87%, respectively. In this group, the
22 formation of zones of squamous metaplasia with zones of squamous epithelial hyperplasia and
23 increased keratin production appeared to precede areas of squamous papillary hyperplasia with
24 foci of cellular atypia. More advanced lesions included carcinoma in situ and invasive SCC of
25 the nasal turbinates. The neoplasia were extremely osteolytic and were associated with excessive
26 keratin production and mild to severe purulent rhinitis. In many animals from the high-exposure
27 group (with or without carcinoma), the excessive accumulation of keratin and inflammatory
28 exudates within the lumen of the URT caused severe dyspnea and death. Polypoid adenomas
29 were also observed in eight rats (four/sex) from the low-exposure group, six male rats from the
30 intermediate-exposure group, and six rats (five males, one female) from the high-exposure group
31 in level 1, 2, or 3. One control male rat had a similar lesion. When adjusted and unadjusted data
32 were analyzed, no significant differences were observed in pair-wise analyses; however, a
33 significant adjusted trend (p < 0.05) was reported for male rats. There was no evidence of
34 progression from polypoid adenoma to SCC.
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-272 DRAFT—DO NOT CITE OR QUOTE
-------
A
100
80
60
40
20
r
_
-
"
-
I
" F7!
I
v
X
X
\
ICTI
1
i
£
I
1
i
*$
E3 6 Month
ESS! 12 Month
El 18 Month
E3 24 Month
ESI 27 Month
^
||
;5l
100
C? 80
D §
60
CT 40
S>
u_
20
100
80
60
40
20
Level I
Level I
Level I
Level IV
Level V
Location of squamous metaplasia in nasal cavity
4
5
6
1
8
9
10
11
12
13
14
15
Figure 4-13. Frequency and location by cross-section level of squamous
metaplasia in the nasal cavity of F344 rats exposed to formaldehyde via
inhalation.
Note: Exposure concentrations were 2.0 ppm (A), 5.6 ppm (B), or 14.3 ppm (C).
Nasal cavity levels 2, 3, 4, and 5 were not evaluated at the 6- and 12-month
interim sacrifices in the 14.3 ppm exposure group.
Source: Redrawn from Kerns et al. (1983).
This document is a draft for review purposes only and does not constitute Agency policy.
4-273 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-39. Summary of neoplastic lesions in the nasal cavity of F344 rats
exposed to inhaled formaldehyde for 2 years
Formaldehyde
(ppm)
0
2.0
5.6
14.3
Sex
M
F
M
F
M
F
M
F
No. of nasal
cavities
evaluated
118
114
118
118
119
116
117
115
sec
0
0
0
0
1
1
51
52
Nasal
carcinoma
0
0
0
0
0
0
1"
1
Undifferentiated
carcinoma or
sarcoma
0
0
0
0
0
0
T
0
Carcino-
sarcoma
0
0
0
0
0
0
1
0
Polypoid
adenoma
1
0
4
4
6
0
4
1
Osteo-
chondroma
1
0
0
0
0
0
0
0
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
"One rat in this group also had an SCC.
Source: Kerns etal. (1983).
Morgan et al. (1986b) performed an additional analysis of the slides and tissues from the
Kerns et al. (1983) study to more precisely determine the location of each tumor recorded.
Additional sections were cut from the existing tissue blocks if a full slide set (i.e., five sections)
was unavailable for each animal. For each animal, the location of each tumor was recorded on
diagrams of the cross section of the nose, and an attempt to determine the site of origin was made
based on the center of the tumor mass. The results for each case were assigned an accuracy
rating that was based on the degree of confidence that the pathologist had in the designated site
of origin. Results for SCCs are shown in Table 4-40.
Table 4-40. Apparent sites of origin for the SCCs in the nasal cavity of F344
rats exposed to 14.3 ppm of formaldehyde gas in the Kerns et al. (1983)
bioassay
Sex
Male
Female
Accuracy
rating
High
Low
High
Low
Totals
Number of
animals
36
25
45
15
121
Total SCC (%)a
Area lb
56
56
62
47
57
Area 2b
28
20
27
33
26
Area 3b
14
8
7
13
10
Area 4b
o
J
0
4
0
3
Unable to determine
NA
16
NA
7
4
21
22
23
24
25
aRounded to nearest whole number.
bArea 1 = lateral aspect of the nasoturbinate and adjacent lateral wall; Area 2 = midventral septum; Area 3 = dorsal
septum and roof of dorsal meatus; Area 4 = dorsal and lateral aspect of the maxilloturbinate.
NA = not applicable.
Source: Morgan et al. (1986b).
This document is a draft for review purposes only and does not constitute Agency policy.
4-274 DRAFT—DO NOT CITE OR QUOTE
-------
1 In the 14.3 ppm exposure group, 98/103 rat noses had adequate numbers and quality of
2 slides for mapping the SCC distribution. Single neoplasia were present in 80 (40/sex), while
3 multiple neoplasia were present in 9 males (21 neoplasia) and 9 females (20 neoplasia). The
4 results were similar for cases with high or low accuracy. For example, more than half (57%) of
5 the SCCs occurred on the lateral side of the nasoturbinate and adjacent lateral wall at the front of
6 the nose (levels 1 and 2); approximately 25% were located on the midventral nasal septum
7 (levels 2 and 3); and about 10% were on the dorsal septum and roof of the dorsal meatus
8 (levels 1, 2, and 3). A small number (3%) were found on the maxilloturbinate (levels 2 and 3),
9 which only involved the medial aspect. All other regions of the nose where SCC was found were
10 considered to be involved as a result of invasion from one or more of the above sites. There
11 were two tumors in the 5.6 ppm group: one male had a single neoplasm on the ventral nasal
12 septum (level 3) while a female had an SCC from the lateral aspect of the maxilloturbinate to the
13 adjacent lateral wall (level 2).
14 On the basis of the morphology of 19 small neoplasia in this study and in additional work
15 described below (Morgan, 1997; Monticello et al., 1996), it was further concluded that the SCCs
16 arose from the epithelium lining the airway and not from the underlying glandular epithelium.
17 This mapping procedure and that of Monticello et al. (1996) described below were in good
18 concordance and showed a clear site specificity; most of the SCC arose in the anterior lateral
19 meatus (ALM) (57%), which is lined by transitional epithelium, and the midventral nasal septum
20 (26%), which is lined by respiratory epithelium (Morgan, 1997).
21 The CUT performed a second bioassay on inhaled formaldehyde in 9-week-old male
22 F344 (CDF[F344]/CrlBr) rats (Monticello et al., 1996). The rats were exposed 6 hours/day, 5
23 days/week for 24 months to 0, 0.7, 2, 6, 10, and 15 ppm (0, 0.86, 2.45, 7.36, 12.3, and
24 18.4 mg/m3) formaldehyde. Study objectives were to repeat the Kerns et al. (1983) bioassay,
25 better defining the concentration response relationship and to seek a correlation between
26 localized data on tumor sites and concomitant cell proliferation assays. Histopathology was
27 performed on six cross-section levels of the nasal cavity on every animal of an unscheduled
28 death and all those of the terminal sacrifice after 24 months. The distribution of lesions for each
29 individual animal was recorded onto epithelial maps of the nasal cavity at 30 selected levels
30 designed to permit accurate localization (Mery et al., 1994). Cell proliferation was measured in a
31 subset of animals (five per treatment group) at 3, 6, 18, and 24 months of exposure in each of the
32 nasal regions to which tumors were mapped (see Table 4-41).
33
34
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-275 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-41. Incidence and location of nasal squamous cell carcinoma in male
F344 rats exposed to inhaled formaldehyde for 2 years
Formaldehyde
concentration
(ppm)
0
0.7
2
6
10
15
No. of
nasal
cavities
examined
90
90
96
90
90
147
Nasal location
Anterior
lateral
meatus
0
0
0
1
12
17
Posterior
lateral
meatus
0
0
0
0
2
9
Anterior
mid-
septum
0
0
0
0
0
8
Posterior
mid-
septum
0
0
0
0
0
1
Anterior
dorsal
septum
0
0
0
0
0
3
Anterior
medial
maxillo-
turbinate
0
0
0
0
0
4
Maxillary
sinus
0
0
0
0
0
0
No. of
animals
with
scca
0
0
0
1
20
69
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
aTotal number of animals with SCCs, including those too large to allocate and those located in a site not listed in this
table.
Source: Monticello et al. (1996).
Yellow discoloration of the fur, a consistent response to formaldehyde in rats, was
observed in the rats exposed to 10 and 15 ppm formaldehyde. There were numerous premature
deaths in the 15 ppm exposure group, resulting in significantly decreased survival relative to
controls (18.8 versus 35.7%;p < 0.001). Survival was higher in the three lowest exposure
groups and statistically comparable to controls in the 10 ppm exposure group (35.7 versus
31.3%, respectively).
Control animals showed no histopathologic evidence of disease in the nasal passages.
Buccal cavity SCC, not associated with the nasal cavity, was present in 2 of 90 control animals.
This was considered an incidental finding and within the spontaneous incidence range reported
for this strain of rat. Buccal SCCs were observed in three animals at 15 ppm and in one animal
at 2 ppm. All other neoplastic responses in the respiratory tract were confined to the nose and
considered to have originated from the epithelium lining the nasal airways. The nasal neoplasia
included SCCs and polypoid (transitional) adenomas and were similar in morphologic
characteristics to those described in the Kerns et al. (1983) chronic bioassay. The incidence of
nasal SCCs by location is summarized in Table 4-41, which demonstrates a clear concentration-
response relationship. No SCCs occurred in the two lowest exposure groups or in the controls.
One nasal rhabdomyosarcoma and two nasal adenocarcinomas were reported in animals in the
highest treatment groups.
This document is a draft for review purposes only and does not constitute Agency policy.
4-276 DRAFT—DO NOT CITE OR QUOTE
-------
1 Regional analysis indicated that the SCCs arose in nasal regions lined with transitional or
2 respiratory epithelium and were most common in the lateral meatus and the midseptum (see
3 Table 4-41). Within the lateral meatus and mid-septum, there was clear evidence of a higher
4 tumor incidence rate in the anterior sample site (p = 0.001 and 0.02, respectively). Smaller
5 numbers of SCCs were observed on the medial aspect of the maxilloturbinate and the dorsal
6 septum and on the posterior lateral wall and lining of the nasopharyngeal meatus (data not
7 shown). No SCCs were observed in the maxillary sinus, with the exception of one animal
8 exposed to 15 ppm that had a small tumor in the wall of the ostium of this sinus. Tumor rates
9 across the seven nasal epithelial sites are presented in Table 4-41. There was an increasing
10 tumor response between the 10 and 15 ppm exposure groups in all sites, except in the ALM. The
11 SCC rates at 10 and 15 ppm were virtually identical (13.3 and 11.6%, respectively), which is
12 probably attributable to the occurrence of many large neoplasia in the lateral meatus site that
13 were not suitable and not counted in the analysis.
14 The nonlinear tumor response is mirrored by a highly nonlinear response in cell
15 proliferation measured after 3, 6, 12, and 18 months of exposure. Significant treatment-induced
16 responses in cell proliferation indices at these time points were only observed at the two highest
17 exposure concentrations (10 and 15 ppm). Other treatment-induced lesions, predominantly
18 epithelial hypertrophy, hyperplasia, squamous metaplasia, and mixed inflammatory cell
19 infiltrate, were also most severe at these two exposure concentrations. Significant distortion and
20 destruction of the nasoturbinate architecture occurred in many animals exposed to 15 ppm.
21 Nasal turbinate adhesions and olfactory degeneration (usually confined to the walls of the
22 anterior dorsal medial meatus) also occurred in animals exposed to 10 and 15 ppm. Lesions in
23 the 6 ppm exposure group were limited to focal squamous metaplasia in the anterior regions.
24 As discussed briefly above, small numbers of polypoid adenomas were also induced by
25 formaldehyde exposure and were similar in acinar-like structure and location to those in the
26 Kerns et al. (1983) bioassay. No polypoid adenomas occurred in the control animals or in the
27 0.7, 2, or 6 ppm exposure groups. A clear concentration response was observed in the 10 and
28 15 ppm exposure groups. Five of 90 animals (5.6%) in the 10 ppm exposure group and 14 of
29 147 animals (9.5%) in the 15 ppm exposure group had a polypoid adenoma. Most of these
30 polypoid adenomas (79%) were located in or adjacent to the lateral meatus. The significance of
31 these tumors for risk assessment remains to be determined (Morgan, 1997).
32 Appelman et al. (1988) studied the effects of bilateral intranasal electrocoagulation
33 damage on susceptibility to inhaled formaldehyde in male SPF Wistar (Cpb: WU) rats. Rats
34 were exposed 6 hours/day, 5 days/week for 13 or 52 weeks to 0, 0.1, 1.0, or 10 ppm (0, 0.12,
35 1.23, or 12.3 mg/m3) formaldehyde. These concentrations were chosen because the various
This document is a draft for review purposes only and does not constitute Agency policy.
4-277 DRAFT—DO NOT CITE OR QUOTE
-------
1 short-term studies performed in the same laboratory (described in Section 4.2.1.2) showed that
2 formaldehyde was noncytotoxic to the nasal mucosa at levels of 0.3, 1.0, and 2.0 ppm, slightly
3 cytotoxic at 3 and 4 ppm, and strongly cytotoxic at 10 and 20 ppm (Zwart et al., 1988; Wilmer et
4 al., 1987; Woutersen et al., 1987). Furthermore, because nasal tumors have only been found at
5 exposure concentrations that also induced severe degenerative, hyperplastic, and metaplastic
6 changes in the nasal epithelium (Griesemer et al., 1985; Squire and Cameron, 1984), Feron et al.
7 (1984) and the investigators at the TNO-CIVO Toxicology and Nutrition Institute postulated that
8 formaldehyde at a subcytotoxic concentration was only a very weak initiator without promoting
9 activity. Appelman et al. (1988) used an electrocoagulation method in this study to evaluate if
10 damage to the mucosa followed by compensatory cell proliferation might render the epithelium
11 vulnerable to subcytotoxic levels of formaldehyde. One-half of the rats used in the study
12 (10/group) were damaged bilaterally and then subjected to the first 6-hour exposure to
13 formaldehyde approximately 20-26 hours after the electrocoagulation procedure. Ten
14 undamaged rats/group were also exposed at each concentration for either 13 or 52 weeks.
15 Histopathologic examination included six standard cross-section levels in the nose; livers of all
16 rats killed at 14 weeks and of all control and 10 ppm exposed rats killed in week 53; larynges,
17 tracheas, and lungs of all rats of the control and 10 ppm exposed rats killed in week 53; and
18 organs and tissues of control and 10 ppm exposed rats with an undamaged nasal mucosa killed in
19 week 53.
20 Yellow discoloration of the fur occurred in all animals of the two highest exposure
21 groups. Growth retardation was observed in the animals killed with or without damaged noses
22 after 2 weeks of exposure to 10 ppm formaldehyde. No lexicologically significant findings in
23 the body weights or organ weights of any animals in the other exposure groups were observed.
24 No relevant differences between groups were found in any of the hematological or urinary
25 parameters with the exception of frequent oliguria (p < 0.05) in the top exposure group without
26 nasal coagulation and killed in week 53. Three-way ANOVA revealed a significant increase in
27 TP content of the liver in rats with damaged noses as compared with rats with undamaged noses,
28 and there was a significant negative correlation between the formaldehyde exposure level and TP
29 in these same rats. Hepatic GSH was positively correlated with both nasal damage and age of
30 the animals. No treatment-related gross findings were observed in animals sacrificed at either 14
31 or 53 weeks except for yellow discoloration of the fur in rats exposed at the two highest
32 concentrations. No changes observed in the larynx, trachea, lungs, liver, or other tissues
33 evaluated were regarded as related to formaldehyde.
34 Few nasal lesions were noted in intact rats exposed at 0.1 or 1 ppm for either 13 or 52
35 weeks (n = 10/group). Focal squamous metaplasia was noted in a single animal exposed at
This document is a draft for review purposes only and does not constitute Agency policy.
4-278 DRAFT—DO NOT CITE OR QUOTE
-------
1 1 ppm for 13 weeks. Rats exposed at 10 ppm formaldehyde demonstrated clear pathology in the
2 respiratory epithelium progressing from 13 to 52 weeks, including squamous metaplasia, basal
3 cell hyperplasia, and focal rhinitis. Additionally, focal nest-like infolds of the epithelium were
4 present in 4 of 10 rats at 52 weeks, and minor changes to the olfactory epithelium were noted
5 (thinning/disarrangement and focal basal cell hyperplasia.)
6 All rats with damaged nasal passages exhibited similar minor pathology of the respiratory
7 epithelium at 13 and 52 weeks (squamous metaplasia, focal basal cell hyperplasia, and focal
8 rhinitis). Formaldehyde-related effects were note at 52 weeks, where the squamous metaplasia
9 of the respiratory epithelium was no longer noted in controls (versus 13 weeks) but was clearly
10 present in all formaldehyde-treatment groups, including progression from focal to diffuse lesions
11 (at 1 and 10 ppm) and keratinization (3/10 and 4/10 at 0.1 ppm and 10 ppm, respectively). The
12 formaldehyde effects on the respiratory epithelium were much more severe in rats with damaged
13 nasal passages, with all animals demonstrating thinning and disarrangement of the olfactory
14 epithelium and 8 of 10 rats exhibiting "loosely arranged submucosal tissue." Squamous
15 metaplasia and focal rhinitis of the olfactory epithelium were seen in less than half of the
16 formaldehyde-treated rats with damage. No changes in the olfactory epithelium due only to
17 electrocoagulation were encountered.
18 The most notable effects of nasal damage from electrocoagulation were the ones at the
19 highest formaldehyde exposure (10 ppm) on the olfactory epithelium. Damage to the respiratory
20 epithelium also occurred more posteriorly in rats with damaged noses. Since electrocoagulation
21 often induced damage that included partial or complete loss of turbinates and septal perforation,
22 a likely explanation for the posterior distribution of the damage is an abnormal airflow pattern.
23 This gross damage to the nasal structure may have also disrupted normal mucous production and
24 flow. Therefore, formaldehyde-induced pathology appearing deeper in the nasal passages,
25 including the respiratory epithelium, may be due to formaldehyde penetrating more deeply into
26 the nasal passages and resulting in greater tissue doses in these areas.
27 Woutersen et al. (1989) conducted a lifetime study in parallel to the 1-year study
28 described above for Appelman et al. (1988). Male Wistar rats with nasal damage induced by
29 electrocoagulation (60/group) or without nasal damage (30/group) were exposed 6 hours/day,
30 5 days/week to the same concentrations as in the previous study (0, 0.1, 1.0, and 10 ppm) for
31 28 months or for 3 months followed by a 25-month observation period. The general condition
32 and behavior of the animals were checked daily. Body weight, organ weight, and gross
33 pathology were evaluated as described for Appelman et al. (1988). Histopathologic examination
34 was conducted on all animals at the standard six cross sections (see Figure 4-6).
This document is a draft for review purposes only and does not constitute Agency policy.
4-279 DRAFT—DO NOT CITE OR QUOTE
-------
1 No remarkable findings on behavior were observed except for yellowing of the fur in
2 animals at the two highest concentrations. There were no relevant differences in mortality (data
3 not shown). Growth retardation was observed relative to controls in animals with or without
4 damaged noses exposed to 10 ppm from day 14 onward. Body weights were generally slightly
5 lower in formaldehyde-exposed animals with an intact nasal mucosa and slightly higher in
6 exposed animals with damaged noses than in the corresponding controls.
7 The effects of formaldehyde exposure on the respiratory and olfactory epithelium after
8 28 months of exposure were similar to those reported for 52 weeks exposure (Appelman et al.,
9 1988): rhinitis, squamous metaplasia with some keratinization of the respiratory epithelium, and
10 thinning/disarrangement and slight squamous metaplasia of the olfactory epithelium at the
11 10 ppm exposure. Effects attenuated from the anterior to posterior sections (I-II, III, IV, and
12 V-VI). A low incidence of olfactory epithelium replaced by respiratory epithelium (<10%) and
13 vacuolation and atrophy of olfactory cells (<10%) was reported, this in part may be due to the
14 larger study size (30 rats per group versus 10). Squamous metaplasia in levels I-II of the
15 respiratory epithelium at 10 ppm was the only treatment-related pathology remaining in rats
16 exposed for 3 months followed by a 25-month recovery period.
17 Similarly, as reported by Appelman et al. (1988), rats with noses damaged by
18 electrocoagulation did demonstrate increased pathology of the respiratory epithelium.
19 Formaldehyde exposure at 10 ppm exacerbated these changes, and effects were noted in more
20 posterior sections than in rats without nasal damage (levels III, IV, and V). Olfactory pathology
21 was also greater in formaldehyde-treated rats: basal cell hyperplasia, replacement of olfactory
22 epithelium by respiratory epithelium (10-20% at level III and <10% at level IV). Although the
23 incidences are low, there is some evidence that effects on the olfactory epithelium may be
24 increased at the lower formaldehyde exposures (0.1 and 1 ppm.) Analysis of the number of
25 animals with olfactory pathology would be helpful to better understand the potential of low-level
26 formaldehyde effects on these less frequent lesions. Interestingly, the recovery of the olfactory
27 and respiratory epithelium seen in rats with undamaged nasal cavities after a 25-month recovery
28 period was not evident in rats with damaged noses. Formaldehyde-exposure effects are only
29 present at the 10 ppm exposure for the respiratory epithelium (squamous metaplasia, basal cell
30 hyperplasia), and the formaldehyde-related effects on the olfactory epithelium
31 (thinning/disarrangement, basal cell hyperplasia, and replacement by respiratory epithelium) are
32 seen at 0.1 and 1.0 ppm as well.
33 A single SCC, 1 out of 30 rats, was found in each 28-month formaldehyde-treatment
34 group (1/26, 1/28, and 1/26, respectively) but not in any control animals (n = 52). SCCs were
35 also noted in rats with noses damaged by electrocoagulation (1/54, 1/58, 0/56, and 15/58 for
This document is a draft for review purposes only and does not constitute Agency policy.
4-280 DRAFT—DO NOT CITE OR QUOTE
-------
1 control rats and the formaldehyde-treatment groups, respectively). These data clearly indicate a
2 synergistic effect of high formaldehyde exposure and nasal damage on the formation of SCCs in
3 rats. One adenosquamous carcinoma and one adenocarcinoma were also reported as increasing
4 the frequency to 17/58 for all tumors. Additionally SCC was present in two rats in the 0.1 and
5 1 ppm 3-month exposure groups with damaged noses only, although only one SCC was reported
6 in the 10 ppm 3-month groups with and without damaged noses. Rats not surviving to
7 28 months are included in these results, as well as the histopathology reported above. Since no
8 mortality data are reported, it should be noted that the incidence of both nasal lesions and tumors
9 are not controlled for early deaths.
10 In total, 30 tumors were examined from this study. In general, the tumors (26/30 or 87%)
11 were SCCs, and 69% (18/26) of these clearly originated from the respiratory epithelium lining
12 the septum or nasal turbinates. The eight other SCCs, derived from the epithelium lining the
13 nasolacrimal duct, were seen in connection with severe odontodystrophy and periodontitis or
14 might have originated from the skin or salivary glands. Four remaining rats bearing a nasal
15 tumor developed a small polypoid adenoma located on the nasoturbinate, an adenocarcinoma
16 originating from the olfactory epithelium, an adenosquamous carcinoma of the respiratory
17 epithelium lining the septum or turbinates, or a carcinoma in situ of epithelium in the
18 nasolacrimal duct.
19
20 4.2.2.5.3. Hamsters.
21 Dalbey (1982) examined the effects of inhaled formaldehyde alone for a lifetime or
22 combined with diethylnitrosamine (DEN) in an initiation-promotion study design using male
23 Syrian golden hamsters. For the first experiment, hamsters were exposed at either 0 or 10 ppm
24 (0 or 12.3 mg/m3) formaldehyde in whole body chambers 5 hours/day, 5 days/week for a lifetime
25 (132 controls, 88 exposed). Histopathologic evaluations were carried out on two transverse
26 sections of the nasal turbinates (otherwise not specified), longitudinal sections of larynx and
27 trachea, and all lung lobes cut along the bronchus prior to embedding. In the formaldehyde-only
28 (10 ppm) experiment, mortality was increased relative to unexposed controls (p < 0.05). No
29 tumors and little evidence of toxicity to the nasal epithelium were observed. There was no
30 increase in rhinitis. Epithelial hyperplasia and metaplasia were increased in formaldehyde-
31 treated animals (5% incidence) versus none observed in controls.
32 The second set of experiments by Dalbey (1982) examined interaction of formaldehyde
33 exposure on tumor formation from DEN administered subcutaneously. The five treatment
34 groups included: (1) controls (n = 50); (2) formaldehyde only (n = 50); (3) DEN 0.5 mg, once
35 per week for 10 weeks (n = 100); (4) formaldehyde exposure for life with DEN injection for the
This document is a draft for review purposes only and does not constitute Agency policy.
4-281 DRAFT—DO NOT CITE OR QUOTE
-------
1 first 10 weeks given 48 hours after formaldehyde exposure (n = 27); and (5) DEN injection for
2 10 weeks, followed by formaldehyde exposure for life (n = 23). In all groups hamsters were
3 exposed at 30 ppm formaldehyde 5 hours/day, once a week. Histopathologic examinations were
4 conducted as above.
5 Although weekly exposures to formaldehyde alone (30 ppm once a week) did not
6 influence mortality, treatment with DEN alone significantly (p < 0.05) increased mortality above
7 that of untreated controls, and mortality was further elevated (p < 0.05) in the two groups
8 exposed to both DEN and formaldehyde compared with DEN alone. No respiratory tract tumors
9 were observed in untreated animals or those receiving only formaldehyde. DEN treatment alone
10 resulted in a high incidence (77%) of tumors (nasal, larynx, trachea, and lung). Formaldehyde
11 pre- or post-treatment did not further increase the number of TBAs-. All tumors observed were
12 classified as adenomas. Formaldehyde pretreatment nearly doubled the number of tumors per
13 animal in the trachea (but not lung or larynx) (p < 0.05). This increase in tumors initiated by
14 DEN given 48 hours after formaldehyde exposure suggests a role of formaldehyde- induced
15 changes in the respiratory tract in tumor promotion (e.g., cell proliferation and inflammation).
16
17 4.2.2.5.4. Summary.
18 Chronic rodent studies of inhalation exposure to formaldehyde provide a consistent
19 picture of the agent's toxicity—especially on the URT—on which most studies focus. All three
20 species tested—hamsters, mice, and rats—had some degree of hyperplastic and metaplastic
21 change in the nasal passages. The pathology defined in acute and subchronic exposures is
22 similarly described in chronic studies, where progression, severity, and presence in more
23 posterior sections of the nose increase with both the concentration and duration of exposure.
24 Pathology of the respiratory epithelium includes rhinitis, goblet cell hyperplasia,
25 pseudoepithelial cell hyperplasia, squamous metaplasia, and dysplasia (see Table 4-42). At
26 higher exposures and longer durations of exposure, similar effects are seen on the olfactory
27 epithelium, present further into the nasal passages. In addition to hyperplasia and squamous
28 metaplasia, thinning and disarrangement of the olfactory epithelium noted and, in a few cases,
29 cell damage and replacement of olfactory epithelium with respiratory epithelium appear,
30 including loss of sensory cells (Woutersen et al., 1989; Kerns et al., 1983; Battelle Columbus
31 Laboratories, 1981).
32
This document is a draft for review purposes only and does not constitute Agency policy.
4-282 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-42. Summary of respiratory tract pathology from chronic inhalation exposures to formaldehyde
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
3 ,
Species/strain
No./
group
Treatment"
Respiratory effects
Noncancer
LOAEL/NOAEL
Reference
Chronic bioassays
Mice
C3H mice (sex
unstated)
Male and female
B6C3F1 mice
60
120/sex
0,41, 82, or 163 ppm
1 hour/day, 3 days/week for up
to 35 weeks.
Low- and mid-group mice then
exposed at either 122 or
244 ppm during weeks 35-70.
0, 2, 5.6, or 14.3 ppm
6 hours/day, 5 days/week for 24
months.
The protocol featured a 6-month
recovery period. Interim
sacrifices occurred at 6, 12, 18,
24, and 30 months.
Female Sprague-
Dawley rats
Male and female
F344 rats
16
32/sex
0 or 12.4 ppm
formaldehyde ± wood dust
6 hours/day, 5 days/week for
104 weeks.
0, 0.3, 2, or 14 ppm
6 hours/day, 5 days/week for
28 months.
Pathology: Histologic changes in the tracheobronchial
epithelium by exposure, including basal-cell
hyperplasia, stratification squamous cell metaplasia, and
atypical metaplasia.
Carcinogenicity: No SCC formation was evident in
mice exposed to formaldehyde alone.
Pathology: Rhinitis; hyperplasia, dysplasia, and
squamous metaplasia of the nasal epithelium; atrophy of
the olfactory epithelium; glandular adenitis and
nasolacrimal duct hyperplasia and metaplasia.
Carcinogenicity: Nasal SCC in male mice at 24 months
(2/17). No SCC in female mice.
Rats
Pathology: Squamous metaplasia and dysplasia.
Carcinogenicity: One of 16 rats exposed to
formaldehyde alone developed SCCs.
Pathology: Increased rhinitis, hyperplasia, and
squamous metaplasia of the nasal respiratory epithelium
Carcinogenicity: Nasal SCCs in high concentration rats
(44%).
LOAEL = 41ppm
No evidence of
Carcinogenicity
LOAEL = 2 ppm
Evidence of
Carcinogenicity
Hortonetal. (1963)
Swenberg et al.
(1980); Kerns etal.
(1983); CUT
(1982) ; Battelle
Columbus
Laboratories (1981)
LOAEL = 12.4 ppm
Support for
Carcinogenicity
LOAEL = 0.3 ppm
Support for
Carcinogenicity
Holmstrom et al.
(1989a)
Tobe etal. (1985)
to
oo
fe
H
o
o
-------
ts
^ 8
1!
f^ S
3 ,
Table 4-42. Summary of respiratory tract pathology from chronic inhalation exposures to formaldehyde (continued)
Species/strain
Male F344 rats
Male Sprague-
Dawley rats
Male and female
F344 rats
Male F344 rats
No./
group
32
100
120/sex
90 and 150
controls
Treatment"
0, 0.3, 2, or 15 ppm
6 hours/day, 5 days/week for
28 months.
0 or 15 ppm 6 hours/day,
5 days/week for life.
0, 2, 5.6, or 14.3 ppm
6 hours/day, 5 days/week for
24 months. The protocol
featured a 6-month recovery
period. Interim sacrifices
occurred at 6, 12, 18, 24, and
30 months.
0, 0.7, 2, 6, 10, or 15 ppm
6 hours/day, 5 days/week for
24 months.
Respiratory effects
Pathology: Squamous cell metaplasia and epithelial
hyperplasia.
Carcinogenicity: SCC (13/32), squamous cell
papilloma (3/32), and sarcoma (1/32).
Pathology: Squamous metaplasia, epithelial
hyperplasia, and polyps/papillomas.
Carcinogenicity: SCCs formed in the nasomaxillary
turbinates and nasal septum (25%).
Pathology: Lesions of the nasal cavity were the primary
effects, including squamous metaplasia and epithelial
dysplasia, hyperkeratosis, goblet cell hyperplasia, and
rhinitis.
Salivary gland: atrophy, squamous metaplasia, and
sialadenitis.
Carcinogenicity: SCCs were evident in the nasal cavity
of high concentration rats, plus some polypoid
adenomas.
Pathology: Olfactory degeneration, squamous
metaplasia, epithelial hypertrophy and hyperplasia, and
mixed inflammatory cell infiltrate.
Carcinogenicity: SCCs and polypoid adenomas in the
nasal cavity
Noncancer
LOAEL/NOAEL
LOAEL = 0.3 ppm
BMD10 = 0.24 ppm
Evidence of
Carcinogenicity
LOAEL = 15 ppm
Evidence of
Carcinogenicity
LOAEL = 2 ppm
Evidence of
Carcinogenicity
LOAEL = 2 ppm
Evidence of
Carcinogenicity
Reference
Kamataetal. (1997)
Albert etal. (1982);
Sellakumar et al.
(1985)
Swenberg et al.
(1980); Kerns etal.
(1983); CUT (1982);
Battelle Columbus
Laboratories (1981);
Morgan et al.
(1986b)
Monticello et al.
(1996)
to
oo
fe
H
o
o
-------
a TO
o <
"* TO'
Table 4-42. Summary of respiratory tract pathology from chronic inhalation exposures to formaldehyde (continued)
•
Species/strain
No./
group
Treatment"
Respiratory effects
Noncancer
LOAEL/NOAEL
Reference
Male SPF Wistar
rats
10
0,0.1, 1, or 10 ppm
6 hours/day, 5 days/week for 13
or 52 weeks.
An electrocoagulation method
was applied to damage the noses
of !/2 of each study group.
Pathology: Formaldehyde-induced focal changes to the
respiratory and olfactory epithelium, including rhinitis,
hyperplasia, and metaplasia (10 ppm).
In rats with damaged noses: squamous metaplasia of the
respiratory epithelium increased at all formaldehyde
exposures. Pathology of the olfactory epithelium
increased at the 10 ppm exposure.
Carcinogenicity: No tumors noted; 1-year study
LOAEL = 0.1ppm
in rats with damaged
nasal passages
NOAEL = 1 ppm for
rats with intact noses
Appelman et al.
(1988)
Male Wistar rats
30 (without
nasal
damage),
60 (with
nasal
damage)
0,0.1, 1, and 10 ppm
6 hours/day, 5 days/week for 28
months or for 3 months with a
25-month observation period.
An electrocoagulation method
was applied to damage the nasal
cavity.
Pathology: Intact noses: squamous metaplasia in the
high concentration group exposed for 28 months and
degeneration of the olfactory epithelium.
Changes were more severe in animals with damaged
noses .
Carcinogenicity: SCCs developed in 15/60 rats with
damaged noses exposed at 10 ppm. In other groups, the
incidence of nasal tumors was low irrespective of the
state of nasal damage.
NOAEL = 1 ppm
Woutersen et al.
(1989)
Evidence of
Carcinogenicity
Hamsters
Male Syrian golden
hamsters
oo
88 treated
132
controls.
0 or 10 ppm formaldehyde
5 hours/day, 5 days/week for
life.
Pathology: Increased mortality. Epithelial hyperplasia
and metaplasia increased in formaldehyde-treated
animals (5% incidence)
Carcinogenicity: No tumors reported.
LOAEL = 10 ppm
No evidence of
Carcinogenicity
Dalbey (1982)
fe
H
O
O
-------
Table 4-42. Summary of respiratory tract pathology from chronic inhalation exposures to formaldehyde (continued)
Species/strain
Male Syrian golden
hamsters
No./
group
50
Treatment"
0 or 30 ppm
5 hours/day, 1 day /week for life
± injections with 0.5 mg DEN.
Respiratory effects
Pathology: Increased mortality in conjunction with
DEN — above DEN-only treated animals. Respiratory
pathology not reported.
Carcinogenicity: Only hamsters receiving DEN
developed tumors (77%, adenomas). There was an
increase in the number of tumors per TBAs in the
trachea of animals exposed to formaldehyde 48 hours
prior to DEN (but no increase in TBAs).
Noncancer
LOAEL/NOAEL
LOAEL = 30 ppm.
Evidence for
formaldehyde as a
promoter
Reference
Dalbey (1982)
& s
II
s ,
to
oo
fe
H
o
o
-------
1 Clear species differences in the severity of lesions are present. Although the bioassays in
2 mice, hamsters, and rats do represent similar exposure concentrations and duration of exposure,
3 hamsters exhibit little pathology and rats (three strains tested) exhibit gross toxicity and even
4 increased mortality. Mice similarly exposed exhibit a range of effects on the respiratory
5 epithelium but not near the severity seen in rats. Many factors may contribute to these observed
6 species differences. As Chang and Barrow (1983) reported, the increased KB of mice seems to
7 be protective of POE damage in comparison to that of rats. The reduced ventilation rate and
8 minute volume of rodents in the presence of a reactive gas can reduce the effective delivered
9 dose at the same exposure concentration (Chang and Barrow, 1983). Additionally, as illustrated
10 in the computational fluid dynamic (CFD) modeling (see Section 3.5), there are species
11 differences in nasal architecture that influence areas of formaldehyde absorption or flux into the
12 tissue. Localized differences in mucus flow and production as well as metabolic enzymes have
13 also been posited as having roles in differential toxicity of formaldehyde on the URT (see
14 Chapter 3).
15 Formaldehyde-induced tumors were present in exposed rats and mice and primarily
16 involved SCCs later in life (Kamata et al., 1997; Tobe et al., 1985; Kerns et al., 1983; Swenberg
17 et al., 1980). Although exposure of male Syrian hamsters to either 10 or 30 ppm did not result in
18 formaldehyde-induced nasal tumors, a classic initiation-promotion assay with DEN-induced
19 tumor formation did indicate that formaldehyde increased the tumor burden per animal, where
20 DEN induced tumors in 77% of the animals (Dalbey, 1982). This study suggests a role for
21 promotion in the observed carcinogenicity of formaldehyde. Less clear are the implications of
22 the synergistic effect of formaldehyde exposures and gross damage to the respiratory epithelium
23 by electrocoagulation on tumor formation (Woutersen et al., 1989).
24
25 4.2.2.6. Summary of Respiratory Pathology and Carcinogenic Potential
26 The progressive pathology of the nasal passages from inhalation exposure to
27 formaldehyde is well documented, especially in rodents (rats and mice) (see Tables 4-12, 4-23,
28 4-36, 4-42). Although there are species differences in tissue dose (see Section 3.4) due to
29 variations in nasal architecture and breathing patterns, the nature and progression of the
30 pathology is fairly well conserved across species, including nonhuman primates. The observed
31 formaldehyde-induced pathology includes disruption of the mucociliary apparatus, rhinitis
32 (serous and purulent), hyperplasia (cell proliferation), metaplasia (transition of cell type),
33 dysplasia (disarrangement of cells), nest-like infolds and invaginations of the epithelium,
34 thinning of the epithelial layer and focal to diffuse lesions, atrophy of the olfactory epithelium,
This document is a draft for review purposes only and does not constitute Agency policy.
4-287 DRAFT—DO NOT CITE OR QUOTE
-------
1 thickening and keratinization (usually of squamous metaplasia), tumors (adenoma, sarcoma,
2 carcinoma) (see Section 4.2.2).
3 Progression of lesions can be viewed as progression from the anterior to posterior
4 sections of the nasal cavity or as a progression in severity of lesions at a particular location (e.g.,
5 level or region) of the nasal passages. In both cases, progression is evident with increasing
6 exposure concentration and with increasing duration of exposure (Kamata et al., 1997;
7 Monticello et al., 1996; Morgan et al., 1986b; Takahashi et al., 1986; Sellakumar et al., 1985;
8 Kerns et al., 1983; Albert et al., 1982). The data suggest that concentration and duration of
9 exposure do not act in a simply cumulative manner (e.g., C x t). Additionally the influence of
10 concentration, duration, and repeated exposure may be different for various effects. For
11 example, some lesions may be transient (e.g., low-exposure cell proliferation), others may have a
12 threshold and vary little after that (e.g., rhinitis). Additionally, as the nasal epithelium responds
13 with both adaptive and adverse epithelial changes, the absorption of formaldehyde into the tissue
14 at that location may be reduced. As respiratory epithelium transitions to squamous metaplasia,
15 the effective tissue dose of formaldehyde increases posterior to these lesions. As barriers to
16 formaldehyde flux into the tissue develop (e.g., squamous metaplasia, keratinization),
17 formaldehyde penetrates more deeply into the nasal passages (Kimbell et al., 2006). Therefore,
18 although both concentration and duration of exposure do affect the adverse effect, the
19 relationship is difficult to define and in fact may be different for various adverse effects.
20 Respiratory histopathology has been commonly reported in response to exposure to
21 formaldehyde in rats and mice (Lino dos Santos Franco et al., 2006; Javden and Taher, 2000;
22 Kamata et al., 1996a, b; Cassee and Feron, 1994; Bhalla et al., 1991; Monteiro-Riviere and Popp,
23 1986; Buckley et al., 1984; Chang et al., 1983), rabbits (lonescu et al., 1978), hamsters
24 (Schreibner et al., 1979), and rhesus monkeys (Monticello et al., 1989). The histopathologic
25 lesions ranged from inflammation to ulceration, necrosis, and metaplasia that occurred in nasal
26 turbinates, maxilloturbinates, and goblet and microvillus cells (Bhalla et al., 1991). These effects
27 were observed at a variety of doses (e.g., 10 ppm for 4 hours, 3.13 ppm for 6 hours for 1, 2, or
28 4 days, 6 or 15 ppm). Wilmer et al. (1989, 1987) assessed whether a dose and time-dependent
29 interaction (C x t) is associated with histopathologic lesions. Results indicated that
30 concentration, rather than duration or cumulative exposure, correlates best with severity of
31 lesions (Wilmer et al., 1989, 1987).
32 Histopathologic lesions and changes to biochemistry have been reported in the lung as
33 well, though these effects were observed following a high dose of formaldehyde. In addition,
34 changes in clinical chemistry, P450 expression and activity in lung tissue, and gene expression
35 that is phenotypically anchored to the observed respiratory pathology have been reported.
This document is a draft for review purposes only and does not constitute Agency policy.
4-288 DRAFT—DO NOT CITE OR QUOTE
-------
1 Extrapulmonary effects have also been noted, including changes in liver chemistry, relative brain
2 weight, and focal, chronic inflammation in the heart and kidney. Most of these changes occurred
3 at exposures of 20 ppm, and those that occurred at lower formaldehyde exposures (3.7 ppm)
4 could not be strictly correlated with formaldehyde exposure.
5 Some researchers have reported formaldehyde-induced effects in the pulmonary region in
6 rats, mice, and rabbits. Kamata et al. (1996a) observed reduced lipid content of pulmonary
7 surfactant in rats exposed to 128.4 or 294.5 ppm formaldehyde. Kamata et al. (1996b) reported
8 biochemical changes in lung homogenates and altered lipid content of BAL at 145.6 ppm
9 formaldehyde. Lino dos Santos Franco et al. (2006) observed increased leukocytes (and
10 neutrophils) and degranulated mast cells recovered in BAL fluid (concentration of 1% formalin
11 not provided). In rabbits, lonescu et al. (1978) observed frank necrosis of lung parenchyma after
12 aerosol inhalation of 3% formalin for 3 hours/day for 50 days (concentration of formaldehyde
13 not provided). These pulmonary effects may be due to frank toxicity resulting from the high
14 dose of formaldehyde used in these studies.
15 Several recent toxicogenomics studies have assessed gene expression changes in nasal
16 and lung tissue in animals and in humans by using in vivo and in vitro approaches. Hester et al.
17 (2005, 2003) documented changes in gene expression associated with DNA repair and apoptosis
18 in nasal tissue from male rats after a single instillation of formaldehyde. Other gene expression
19 changes were observed in those genes related to xenobiotic metabolism and in cell cycle and
20 repair. These preliminary results provide an initial basis for forming a phenotypically anchored
21 set of gene expression changes associated with exposure to formaldehyde and may assist in
22 determining the underlying MOA, as will be discussed in Section 4.5. Sul et al. (2007)
23 investigated gene expression genes in lung tissue from formaldehyde-exposed rats. Yang et al.
24 (2005) performed a proteomics analysis by using lung tissue extracted from formaldehyde-
25 exposed rats. Two studies used human tracheal cell lines to investigate formaldehyde-induced
26 gene expression changes in vitro (Lee et al., 2008, 2007). However, the relevance of these
27 findings to actual exposures remains unknown. In total, toxicogenomics studies hold promise,
28 but they must be interpreted with caution until results can be replicated and phenotypically
29 linked to observable changes.
30 Thus, formaldehyde-induced respiratory pathology has been commonly described in the
31 nasal passages and includes cellular proliferation, mucociliary function, and histopathologic
32 lesions. Pulmonary effects have been documented as well but at high doses. The nasal
33 pathology may occur as a result of both concentration and duration components of exposure.
34
This document is a draft for review purposes only and does not constitute Agency policy.
4-289 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.2.2.6.1. Carcinogenic potential.
2 In the respiratory tract, only nasal tumors are considered formaldehyde induced in rodent
3 studies. The majority of studies were conducted using rats (F344, Wistar, or Sprague-Dawley),
4 and all studies of 18 months or greater in mice and rats show evidence of formaldehyde-induced
5 nasal carcinogenicity. The nasal tumors are primarily SCCs, although papillomas, polypoid
6 adenoma, adenocarcinoma, fibrosarcoma, and esthesioneuroepithelioma have been reported
7 (Kamata et al., 1997; Monticello et al., 1996; Morgan et al., 1986a, b; Takahashi et al., 1986;
8 Sellakumar et al., 1985; Kerns et al., 1983; Albert et al., 1982). Although hyperplasia, dysplasia,
9 and squamous metaplasia of the respiratory epithelium have been observed beyond the nasal
10 cavity, other respiratory tract tumors have not been significantly increased by formaldehyde
11 exposure alone.
12 Increased tumor incidence and decreased latency are correlated with increasing
13 formaldehyde exposure concentration. Reviewing data from the only lifelong inhalation study
14 with multiple exposure groups, SCC is first noted at 8 and 9 months for high exposed (15 ppm)
15 female and male F344 rats autopsied as "early deaths" prior to the 12 month sacrifice, with an
16 incidence of 43% over the course of the study (unadjusted for mortality) (Kerns et al., 1983). In
17 contrast only two SCCs were found in male and female rats sacrificed after 24 months of
18 exposure (incidence of SCC 2.5% at 24 months) (Kerns et al., 1983). In a follow-up study by
19 Monticello et al. (1996), the incidence of SCC in rats exposed at 15 ppm was 47% with the first
20 tumor noted at 12 months. The incidence of SCC in male rats exposed at 10 ppm was 22% with
21 the first SCC noted at 18 months. Moreover, of 90 rats exposed at 6 ppm for 20 months only one
22 SCC was noted. No SCCs were detected in rats exposed at 0.7 or 2 ppm formaldehyde. These
23 incidence rates are not mortality adjusted and include animals from each scheduled sacrifice (3,
24 6, 12, and 18 months). In a lifelong study of male Sprague-Dawley rats exposed at 15 ppm
25 formaldehyde, the cumulative nasal tumor incidence was calculated as a function of time of
26 exposure (see Figure 4-14) (Sellakumar et al., 1985). After 2 years of exposure, the probability
27 of nasal carcinoma was greater than 60%.
28 There is some evidence that less-than-lifetime exposure to formaldehyde can induce nasal
29 tumors over an extended observation period. Two studies, both in male Wistar rats, report nasal
30 tumors in response to less-than-lifetime exposures (Woutersen et al., 1989; Feron et al., 1988).
31 A 13-week exposure at 20 ppm resulted in four nasal tumors (three SCCs), a cystic SCC of the
32 nasolacrimal duct, and an epithelial tumor on the mandible, for a total of six tumors observed
33 over 30 months of observation (Feron et al., 1988). No tumors were noted in 13-week controls.
34 A limited number of formaldehyde-related tumors were noted due to 4 or 8 weeks of exposure
35 followed by 30 months of observation. Although the tumor incidence of these less-than-lifetime
This document is a draft for review purposes only and does not constitute Agency policy.
4-290 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
exposures is low, this is consistent with the 2-year bioassays in Wistar rats. Wistar rats are more
resilient to formaldehyde-induced nasal toxicity than F344 or SD rats (see Section 4.2.1), and
only 1 of 26 (4%) Wistar rats exposed at 10 ppm for 28 months developed SCC (Woutersen et
al., 1989) versus 22% in F344 rats (Monticello et al., 1996).
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
80
'I
£ 60
3 20
10 ppm HCI + 14ppm HCHO (PremU)
14 ppm HCHO
10 ppm HCI + 14ppm HCHO (Not Premix)
240 360 480 600 720
Days After First Exposure
840
Figure 4-14. Mortality corrected cumulative incidences of nasal carcinomas
in the indicated exposure groups.
Source: Sellakumar et al. (1985).
Woutersen et al. (1989) also examined the effect of severe nasal damage from
electrocoagulation on formaldehyde-induced SCC in Wistar rats. Nasal tumors were noted in
formaldehyde-exposed rats without damaged noses (exposed for only 3 months and observed for
25 months). However, the low incidence of tumors in each treatment group (1/26, 2/60, 2/60,
1/58) indicates these data should be considered suggestive even though no SCCs were noted in
control rats with or without damaged noses (n = 83). The studies by Woutersen et al. (1989) did
demonstrate a synergistic effect of nasal damage from electrocoagulation and 10 ppm
formaldehyde exposure (3 months), where 15/58 rats had SCC versus 1/26 with undamaged
noses. The study was originally designed to examine the effect of formaldehyde on the damaged
tissue on cancer promotion. However, it is unclear if the synergistic effect of formaldehyde
exposure on damaged nasal tissue is an effect of formaldehyde on the damaged cells and joint
effects of a mutagen with regenerative proliferation from the nasal damage. It is also possible
This document is a draft for review purposes only and does not constitute Agency policy.
4-291 DRAFT—DO NOT CITE OR QUOTE
-------
1 the damaged nasal passages may alter airflow in the nasal passages, resulting in significantly
2 different flux of formaldehyde into the tissue.
3 There is a single inhalation study (Dalbey, 1982) that investigates the role of promotion
4 in formaldehyde-induced cancer. Although hamsters exhibit little to no effects of formaldehyde
5 on the nasal mucosa or other respiratory tract tissues (Rusch et al., 1983a, b; Dalbey, 1982),
6 DEN-induced respiratory adenomas were increased with formaldehyde exposure (10 ppm)
7 48 hours prior to DEN injection (but not by formaldehyde alone or formaldehyde exposure after
8 DEN injection). The number of tracheal tumors per TEA was doubled by formaldehyde
9 exposure. The study authors note that adenomas should be considered independent tumors and
10 that the increase in tracheal tumors is of biological significance even given the incidence of
11 TBAs (77%, DEN alone), was not further increased by formaldehyde exposure. It is of
12 particular interest that a promotion study in hamsters is positive, since so little nasal pathology
13 occurs with formaldehyde exposure. The absence of significant hyperplasia and tissue damage
14 in these animals suggests that formaldehyde may induce subtle changes in the respiratory tract
15 mucosa that permit formaldehyde to act as a tumor promoter.
16
17 4.2.3. Gastrointestinal Tract Pathology
18 As with inhalation, the POE is thought to be the principal target tissue in response to oral
19 exposure. A concentration-dependent pattern of toxicity longitudinally down the GI tract has
20 been observed upon oral exposure. Some evidence (Til et al., 1989, 1988) suggests that, with
21 regard to oral exposure, duration in addition to concentration is important in the development of
22 toxicity.
23 Formalin and paraformaldehyde were used to dose animals in oral toxicity studies.
24 Formalin contains 12-15% methanol as a preservative to inhibit the polymerization of
25 formaldehyde and subsequent precipitation as paraformaldehyde (Kiernan, 2000). The presence
26 of methanol in formalin may confound the results of a formaldehyde study. Methanol has been
27 shown to be a developmental and neurologic toxin (e.g., Degitz et al. [2004a, b]; Rogers et al.
28 [2004, 2002]; Weiss et al. [1996]; Sharpe et al. [1982]). Oral dosing with paraformaldehyde is
29 preferred because it allows for the preparation of methanol-free formaldehyde in the laboratory
30 by dissolving paraformaldehyde in slightly basic water.
31
32 4.2.3.1. Short- Term and Subchronic Ingestion Studies
33 Til et al. (1988) evaluated the oral toxicity of formaldehyde and acetaldehyde in a
34 subacute study in Wistar (Cpb:WU; Wistar random) rats. Groups of rats (10/sex/dose) were
35 exposed to paraformaldehyde dissolved in drinking water at 0, 5, 25, and 125 mg/kg-day for 4
This document is a draft for review purposes only and does not constitute Agency policy.
4-292 DRAFT—DO NOT CITE OR QUOTE
-------
1 weeks. The control group was comprised of 20 rats of each sex. To account for potential effects
2 of decreased water consumption in treated animals, an additional control group of 10 male and
3 10 female rats was given drinking water in an amount equal to the amount of liquid consumed by
4 the group given the highest dose. Examination of the GI tract was performed in all dose groups
5 and included the tongue, esophagus, and stomach. Histopathology for the other tissues was
6 performed on high-dose and control animals.
7 The rats appeared to be healthy throughout the study, and no effects on growth occurred
8 despite significant decreases in food and water intake that occurred at the high dose (125
9 mg/kg-day). Yellow discoloration of the fur occurred in the rats on the high dosage from week 3
10 onward. There were no significant changes in hematology among the exposed groups except for
11 slight (not statistically different) increases in PCVs in the water-restricted group and in high-dose
12 males. The high-dose groups of the formaldehyde exposed and in the water-restricted controls
13 had slightly increased urine density, but again this was not statistically significant. Plasma TP
14 and ALB levels were decreased in the males of the highest dose group. No changes in organ
15 weights occurred except for relative kidney weights that were slightly increased in the females of
16 the high-dose group. Gross pathological findings were restricted to the GI tract and revealed a
17 thickening of the limiting ridge of the forestomach in all animals exposed at the highest dose that
18 was accompanied by a yellowish discoloration of the mucosa. These latter changes were not
19 observed in the acetaldehyde-exposed animals. Treatment-related histopathologic changes were
20 seen in the GI tract only. Slight (8/20) or moderate (12/20) focal hyperkeratosis of the
21 forestomach and slight focal atrophic gastritis occurred in animals of the high-dose groups only
22 (see Table 4-43). One female had moderate focal papillomatous hyperplasia. No
23 histopathologic changes were observed in any animals of the lower-dose groups. The study
24 established a LOAEL and NOAEL for epithelial changes in the GI tract of male and female
25 Wistar rats exposed to formaldehyde in drinking water at 125 mg/kg-day and 25 mg/kg-day,
26 respectively.
27 Johannsen et al. (1986) performed a subchronic study by using rats exposed to
28 paraformaldehyde dissolved in drinking water and dogs exposed to paraformaldehyde in the diet.
29 Groups of albino Sprague-Dawley rats (15/sex) were administered the equivalent of 0, 50, 100,
30 or 150 mg/kg-day in their drinking water for 91 consecutive days. Pure-bred beagle dogs
31 (four/sex/group) were fed a diet with added aqueous formaldehyde to approximate 0, 50, 75, or
32 100 mg/kg-day. Dogs were observed daily and rats at frequent intervals for behavioral reactions.
33 Body weights and food and water intake were recorded on a weekly basis in both species.
34 Hematology (HCT, Hb, total and differential leukocyte counts), clinical chemistry (blood sugar,
35 BUN, ALP, AST and ALT in dogs only), and urine analyses (color, appearance, pH, specific
This document is a draft for review purposes only and does not constitute Agency policy.
4-293 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-43 Summary of lesions observed in the gastrointestinal tracts of
Wistar rats after drinking-water exposure to formaldehyde for 4 weeks
Type of lesion
Formaldehyde (mg/kg-day)
0
5
25
125
Number of male rats examined
20
10
10
10
Focal hyperkeratosis of forestomach
Very slight
Slight
Moderate
Focal gastritis
Slight
Moderate
Dilated fundic glands (single or a few)
Submucosal mononuclear cell infiltrate
Type of lesion
Focal hyperkeratosis of forestomach
Very slight
Slight
Moderate
Focal gastritis
Very slight
Slight
Moderate
Focal papillomatous hyperplasia
Polymorphonuclear leukocytic infiltration
o
J
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
6
2
1
0
1
Number of female rats examined
20
6
0
0
0
0
0
0
0
10
0
0
0
0
0
0
0
0
10
0
0
0
0
0
0
0
0
10
2
2
6
1
1
1
1
1
4
5
6
7
9
10
11
12
Source: Til etal. (1988).
gravity, sugar, protein, and microscopic elements) were evaluated in 10 male and 10 female rats
selected from each test group and in all dogs. Organ weights were recorded for the adrenals,
gonads, hearts, kidneys, livers, lungs, and thyroids in each species. Histopathology was
performed on a set of over 20 or 30 tissues and organs from rats or dogs, respectively, in the
high-dose and control groups only.
This document is a draft for review purposes only and does not constitute Agency policy.
4-294 DRAFT—DO NOT CITE OR QUOTE
-------
1 No deaths or abnormal reactions were observed in either species. Significant reductions
2 in weight gain were observed in dogs of both sexes at 100 mg/kg-day, in rats of both sexes at
3 150 mg/kg-day, and in male rats at 100 mg/kg-day of formaldehyde. There was a dose-related
4 decrease in liquid consumption of both sexes in rats given formaldehyde, but there was no
5 overall difference in mean food intake or feed efficiency, so the reductions in body weight gain
6 were considered to be systemic effects. Dogs administered formaldehyde had reduced food
7 consumption and feed efficiency at all doses tested. No significant effects on hematology,
8 clinical chemistry, or urine analyses were observed in either species. No effects in either species
9 were reported on organ weights. The GI mucosa in both species was reported to appear normal
10 with no indication of irritation. This study suggests a NOAEL of 150 mg/kg-day in Sprague-
11 Dawley rats and of 100 mg/kg-day in beagle dogs for formaldehyde in drinking water.
12 Differences in the results for the rats with those reported in other studies (Til et al., 1989, 1988;
13 Tobe et al., 1989) may be due to strain differences or duration of the exposure. The dog may be
14 a more sensitive species than the rat based on these results and on those of 2-week pilot studies.
15
16 4.2.3.2. Chronic Ingestion Studies
17 The same laboratory that tested formaldehyde and acetaldehyde in a 4-week study (Til et
18 al., 1988) performed a chronic bioassay with formaldehyde in drinking water. Til et al. (1989)
19 administered paraformaldehyde dissolved in drinking water to Wistar rats (Cpb: WU; Wistar
20 random) (70/sex/dose). Interim sacrifices (10/sex/dose) were performed at 12 and 18 months.
21 Formaldehyde was administered in drinking water to provide target doses of 0, 5, 25, and 125
22 mg/kg-day. The mean formaldehyde doses administered were 0, 1.2, 15, or 82 mg/kg-day for
23 males and 0, 1.8, 21, or 109 mg/kg-day for females. Concentrations were adjusted weekly for
24 the first 12 weeks based on dose estimates derived from body weight and liquid consumption
25 data. Such adjustments were made every 4 weeks from weeks 12 to 52 and kept constant. Fresh
26 solutions of the test concentrations were prepared weekly and stored at 15°C.
27 Endpoints examined included daily observations for condition and behavior, body weight
28 at weekly intervals for the first 12 weeks and then every 4 weeks thereafter, liquid intake weekly,
29 and food intake weekly for the first 12 weeks and then every 2 weeks for the remainder of the
30 study. Samples of blood were taken for hematological and clinical chemistry analyses on weeks
31 26 and 103. Analysis of blood glucose and urine pH, density, and volume was performed on
32 samples at weeks 27, 52, 78, and 104. Pooled urine samples were also evaluated for glucose,
33 occult blood, ketones, urobilinogen, and bilirubin in samples at weeks 27 and 104. Weights of
34 all major organs were recorded at interim sacrifices and at term. Gross and histopathologic
35 examinations were carried out on all major tissues of the rats in the high-dose and control
This document is a draft for review purposes only and does not constitute Agency policy.
4-295 DRAFT—DO NOT CITE OR QUOTE
-------
1 groups. The livers, lungs, stomach, and noses were examined in all rats. Additionally, the
2 adrenals, kidneys, spleens, testes, thyroids, ovaries, pituitaries, and mammary glands (for
3 females) were examined in all sacrificed animals at weeks 53 and 79 and at term.
4 The general health and behavior of the rats were not affected in any of the formaldehyde-
5 exposed groups. Slight yellowing of the fur did occur in the animals exposed at the mid and high
6 doses from week 3 onward. The mean body weights were decreased in the males from week 1
7 and in the females from week 24 onward. At the high dose, liquid consumption was significantly
8 decreased in both sexes, and food intake was significantly decreased in the males. There were no
9 lexicologically significant effects on hematological, urinary, or clinical chemistry parameters.
10 Decreases in absolute heart, liver, and testis (males) weights were attributed to lower body
11 weights. Relative kidney weights were increased in females of the high-dose group, and relative
12 brain weights were increased in both sexes of the high-dose group. Relative testis weight was
13 increased in males. Treatment-related changes in gross pathology were restricted to the
14 forestomach. Histopathologic examinations at the two interim sacrifices and final sacrifice
15 revealed GI tract changes. Renal changes were observed in the high-dose group at final
16 sacrifice. There was no indication of treatment-related effects in other tissues.
17 As shown in Table 4-44, significant histopathology in the GI tract was limited to the
18 forestomach and stomach of rats in the high-dose groups. Some progression with duration of
19 exposure may have occurred by week 105 because GI lesions were observed in the lower dose
20 groups at this time point, whereas none were observed in these groups at interim sacrifices. The
21 histopathologic changes included papillary epithelial hyperplasia in the forestomach that was
22 frequently accompanied by hyperkeratosis on the limiting ridge or its vicinity. The mucosa
23 showed an irregular layer of hyperplastic basal cells, but no atypical nuclei or other subcellular
24 structures were observed. Chronic atrophic gastritis occurred to varying degrees in the stomachs
25 of all high-dose rats. In some cases the inflammatory process involved the entire mucosa and
26 was seen to extend to the whole muscularis mucosae and met the criteria for ulceration.
27 Histologic examination also showed that the incidence and degree of renal papillary
28 necrosis was increased in animals of the high-dose groups at the terminal sacrifice. This change
29 was located at the tip of the papilla and was characterized by patchy necrosis of interstitial cells,
30 capillaries, and loops of Henle. There was no evidence of a dose-related response in chronic
31 nephropathy. The incidence of chronic nephropathy was lower in the males of the high-dose
32 group than in controls. In females, the incidence was slightly higher in the test groups than in
33 controls but only achieved statistical significance at the lowest dose. It is likely that the decrease
34 in liquid intake incurred in the high-dose groups contributed to the increased incidence and
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-296 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-44. Incidence of lesions observed in the gastrointestinal tracts of
Wistar rats after drinking-water exposure to formaldehyde for 2 years
Incidence of lesions with formaldehyde dose (mg/kg-day)a
Males
0
1.2
15
82
Females
0
1.8
21
109
Week 53
Number of rats examined1"
9
10
10
10
10
10
10
9
Forestomach
Focal papillary epithelial hyperplasia
0
0
0
7
0
0
0
5
Glandular stomach
Chronic atrophic gastritis
Focal ulceration
Focal mononuclear cell infiltrate
Atypical glandular hyperplasia
0
0
1
0
0
0
0
0
0
0
3
0
10C
3
0
0
0
0
2
0
0
0
0
0
0
0
0
0
9C
1
0
1
Week 79
Number of rats examined
10
10
10
10
10
9
10
10
Forestomach
Focal papillary epithelial hyperplasia
2
1
1
8
1
0
1
9
Glandular stomach
Chronic atrophic gastritis
Focal ulceration
Focal squamous metaplasia
Submucosal inflammatory cell infiltrate
Focal mononuclear cell infiltrate
Glandular dilation
0
0
0
1
0
2
0
0
0
2
1
4
0
0
0
0
0
4
10C
2
1
0
0
1
0
0
0
0
1
2
0
0
0
0
1
2
0
0
0
0
0
4
10C
0
0
0
0
0
Week 105
Number of rats examined1"
47
45
44
47
48
49
47
48
Forestomach
Focal papillary epithelial hyperplasia
Focal hyperkeratosis
Focal ulceration
Focal acanthosis
Focal basic cell hyperplasia
Diverticurum
Exophytic papilloma
1
2
1
1
0
0
0
2
6
1
0
1
0
1
1
4
1
2
1
1
0
45C
24C
8
1
0
0
0
1
3
0
0
1
0
1
0
5
0
0
0
0
0
2
o
J
2
0
0
0
0
45C
33C
5
1
0
0
0
Glandular stomach
Chronic atrophic gastritis
Focal ulceration
Glandular hyperplasia
Mineralization
Focal inflammatory cell infiltrate
0
0
0
o
J
5
0
0
1
2
o
J
0
0
0
1
2
46C
11°
20C
0
0
0
0
0
0
2
0
0
0
0
3
0
0
0
0
1
48C
10C
13C
0
0
4
5
6
7
8
9
10
""Incidence in rats that died or were killed when moribund during the experiment or were killed at week 53,
79, or 105.
bA few rats were lost because of advanced autolysis.
°The values differ significantly (Fisher's exact test) from the control value (p < 0.001).
Source: Til etal. (1989).
This document is a draft for review purposes only and does not constitute Agency policy.
4-297 DRAFT—DO NOT CITE OR QUOTE
-------
1 degree of renal papillary necrosis observed in the high-dose animals because dehydration has
2 been shown to enhance its production by various analgesics.
3 The results of this chronic bioassay indicated that formaldehyde is cytotoxic to the
4 epithelial mucosa of the nonglandular (forestomach) and glandular stomach with a LOAEL of 82
5 and 109 mg/kg-day and a NOAEL of 15 and 21 mg/kg-day in males and females, respectively.
6 The findings provided no evidence of carcinogenicity in either the GI tract or systemic sites for
7 formaldehyde administered in drinking water to Wistar rats at doses as high as 82 mg/kg-day.
8 Tobe et al. (1989) performed a chronic toxicity study of Wistar rats (Slc:Wistar) exposed
9 to paraformaldehyde dissolved in drinking water. Groups of 20 male and 20 female rats were
10 given formaldehyde solution in their drinking water at concentrations of 0, 0.02, 0.10, and 0.50%
11 for 24 months. Interim sacrifices of six randomly chosen rats from each group were performed
12 after 12 and 18 months. Based on the estimated average amount of water intake and body
13 weight, the actual doses of formaldehyde in either sex were reported to be 0, 10, 50, and
14 300 mg/kg-day. Fresh test solutions were prepared twice each week. The rats were observed
15 daily for the entire study. Body weights and water and diet intake were measured once weekly
16 or biweekly. Hematology (RBC, WBC, and Hb) and serum clinical chemistry (TP, ALB, BUN,
17 uric acid, total cholesterol, inorganic phosphorous, ALP, AST, and ALT) were made at each
18 necropsy. Organ weights were measured for the brain, heart, lung, liver, kidney, spleen, adrenal,
19 testis or ovary, pituitary, and thyroid. These organs and the stomach, small and large intestine,
20 pancreas, uterus, lymph nodes, and all tumors were examined histopathologically.
21 The general condition of animals in the high-dose group was poor with significantly
22 reduced body weight gain as well as intake of water and diet. An increase in mortality was also
23 observed in this group. Some clinical chemistry parameters were altered in this group. No
24 significant changes in absolute or relative organ weights were observed. Mortality was 100% in
25 the high-dose group by 24 months. At the 12-month sacrifice, hyperplasia of the squamous
26 epithelium with or without hyperkeratosis was observed in the forestomach of all high-dose
27 animals (12/12). Basal cell hyperplasia with growth into the submucosa was also observed in
28 most cases (10/12). Erosions and/or ulcers with submucosal inflammatory cell infiltrates were
29 observed in the glandular stomach of most rats (10/12). Regenerative changes of the glandular
30 epithelium (glandular hyperplasia) were noticed in most cases (10/12) along the limiting ridge of
31 the fundic mucosa. No lesions were observed in the glandular stomach at the 50 mg/kg-day
32 dose, and forestomach hyperplasia was observed in only one of six males and in one of eight
33 females at 18 and 24 months. No lesions in either the forestomach or glandular stomach were
34 observed in rats treated at 10 mg/kg-day.
This document is a draft for review purposes only and does not constitute Agency policy.
4-298 DRAFT—DO NOT CITE OR QUOTE
-------
1 This study corroborates the Til et al. (1989) study and shows that the main targets for
2 formaldehyde toxicity administered by drinking water to rats are the forestomach and glandular
3 stomach. Although the lesions observed at the 50 mg/kg-day were minimal in this study, Tobe et
4 al. (1989) designated the NOAEL at 10 mg/kg-day, further supporting the NOAEL of 15
5 mg/kg-day from the Til et al. (1989) study.
6 Takahashi et al. (1986) studied the effects of formaldehyde in an initiation-promotion
7 model of stomach carcinogenesis in male outbred Wistar rats (Shizuoka Laboratory Center,
8 Shizuoka). Rats (n = 17) were given 100 mg/L of TV-methyl-TV1-nitro-jV-nitrosoguanidine
9 (MINING) in drinking water and a diet supplemented with 10% sodium chloride (NaCl) for the
10 first 8 weeks as an initiation phase. This was followed by 0.5% formalin (which contains
11 12-15% methanol) in drinking water for 32 weeks as the promotion phase of the protocol. A
12 comparison group (n = 10) was given stock water and diet without any supplementation for the
13 first 8 weeks followed by 0.5% formalin in drinking water for 32 weeks. Animals were observed
14 daily and weighed once every 4 weeks. Small pieces of the stomach and other tissues in the
15 peritoneal cavity were fixed for histopathologic examination.
16 Body weight gain was reduced by exposure to MINING with sodium chloride, and
17 formaldehyde exposure during the promotion phase exacerbated this effect. Histopathologic
18 investigations were restricted to the GI tract. Formaldehyde was shown to statistically increase
19 the incidence of lesions in the forestomach and stomach in the animals initiated with MNNG
20 with NaCl as compared with controls receiving no initiation (see Table 4-45). Increases in
21 papilloma in the forestomach, adenomatous hyperplasia in the fundus, and adenocarcinoma in
22 the pylorus were observed. Histopathology in the animals receiving formaldehyde alone during
23 weeks 9 through 32 showed an increase in forestomach papillomas but with no lesions in the
24 glandular stomach (see Table 4-45). The adenomatous hyperplasia were defined as proliferative,
25 noninvasive mucosal lesions, and the adenocarcinomas were defined as well differentiated and
26 composed of typical glandular structures, demonstrating a tubular pattern and cellular or
27 structural atypism without metastasis. No definition of criteria for papilloma diagnosis was
28 provided. The findings in this study are inconsistent with those of Til et al. (1989), who found
29 no evidence of carcinogenicity in a 2-year bioassay at comparable concentrations (assuming 37%
30 formaldehyde in formalin results in 0.19% formaldehyde in this study). As discussed above, the
31 differences may be due to differences in the strains of rat or in the diagnostic criteria. The lack
32 of more than one test concentration precludes dose-response analysis of this study and provides
33 only a stand-alone LOAEL of 0.2% formaldehyde in drinking water. The lack of consumption
34 data precludes an estimation of dose in mg/kg-day.
This document is a draft for review purposes only and does not constitute Agency policy.
4-299 DRAFT—DO NOT CITE OR QUOTE
-------
it
& s
II
Table 4-45. Effect of formaldehyde on gastroduodenal carcinogenesis initiated by MNNG and NaCl in male
Wistar rats exposed to formaldehyde (0.5% formalin) in drinking water for 8 weeks
No MNNG initiation prior to 8-week oral exposure to formaldehyde (0.5% formalin in drinking water)
Control
Formaldehyde
Gastroduodenal
carcinoma
0%
0%
Forestomach
papillomas
0%
80%a
Glandular stomach tumors
Fundus
Adenocarcinoma
0%
0%
Adenomatous
hyperplasia
0%
0%
Pylorus
Adenocarcinoma
0%
0%
Preneoplastic
hyperplasia
0%
0%
Duodenum
Adenocarcinoma
0%
0%
MNNG initiation (100 mg/L in drinking water for 8 weeks) prior to 8-week oral exposure to formaldehyde (0.5% formalin in drinking water)
Control
Formaldehyde
Gastroduodenal
carcinoma
13.3%
29.4%
Forestomach
papillomas
0%
88.2%a
Glandular stomach tumors
Fundus
Adenocarcinoma
0%
0%
Adenomatous
hyperplasia
0%
88.2%a
Pylorus
Adenocarcinoma
3.3%
23.5%b
Preneoplastic
hyperplasia
23.3%
41.2%
Duodenum
Adenocarcinoma
10.0%
5.9%
5'
3 ,
oo
o
o
fe
H
Significantly different from control animals with MNNG initiation, p < 0.01.
bSignificantly different from control animals with MNNG initiation, p < 0.05.
Source: Takahashi et al. (1986).
O
O
-------
1 4.2.3.3. Summary of Gastrointestinal Effects and Evaluation of Carcinogenic Potential
2 Short-term and subchronic exposures to formaldehyde via drinking water for 4 weeks
3 yielded slight to moderate histopathologic lesions (focal hyperkeratosis) at 125 mg/kg-day in
4 male and female Wistar rats, as well as slight focal gastritis and submucosal infiltrate in one to
5 two animals of both sexes (Til et al., 1988). No histopathologic lesions were noted in albino
6 Sprague-Dawley rats or beagle dogs that received oral doses of formaldehyde in drinking water
7 for 91 days (Johannsen et al., 1986). In both studies, decreases in weight gain were noted in
8 exposed animals compared with controls.
9 As with the respiratory tract, the proximal portion of the GI tract exhibits formaldehyde-
10 induced lesions in the forestomach and glandular stomach (Til et al., 1989; Tobe et al., 1989;
11 Takahashi et al., 1986). In a chronic drinking water study, Til et al. (1989) reported that
12 formaldehyde is cytotoxic to the epithelial mucosa of the nonglandular (forestomach) and
13 glandular stomach with a LOAEL of 82 and 109 mg/kg-day and a NOAEL of 15 and 21
14 mg/kg-day in males and female Wistar rats, respectively. The findings provided no evidence of
15 carcinogenicity in either the GI tract or systemic sites for formaldehyde administered in drinking
16 water to Wistar rats at doses as high as 82 mg/kg-day. The incidence and degree of renal
17 papillary necrosis was increased in animals of the high-dose groups at the terminal sacrifice (Til
18 et al., 1989). Findings by Tobe et al. (1989) corroborate the Til et al. (1989) study and show that
19 the main targets for formaldehyde toxicity administered by drinking water to rats are the
20 forestomach and glandular stomach.
21 There is evidence that formaldehyde may act as a tumor promoter by the oral route as
22 well as the inhalation route (discussed above). Takahashi et al. (1986) studied the effects of
23 formaldehyde in an initiation-promotion model of stomach carcinogenesis in male outbred
24 Wistar rats (Shizuoka Laboratory Center, Shizuoka, Japan Takahashi et al. (1986) reported an
25 increase in MNNG-initiated GI cancers with formaldehyde exposure (29.4 versus 13.3% TEA in
26 controls); the greatest difference in tumor-containing versus nontumorigenic mice was associated
27 with adenocarcinoma in the glandular stomach (23.5 versus 3.3% in controls). Additionally,
28 forestomach papillomas and preneoplastic hyperplasia in the glandular stomach were increased
29 with formaldehyde exposure alone.
30
31 4.2.4. Immune Function
32 Leach et al. (1983) documented potential immunomodulatory effects of formaldehyde
33 inhalation exposure. F344 rats were exposed nose only to formaldehyde 6 hours/day,
34 5 days/week for up to 30 days. The target concentrations for exposure were 0, 3, 16, 61, and
35 99 ppm formaldehyde (0, 3.7, 19.7, 75.0, and 122 mg/m3). Body weight and food consumption
36 were recorded, and blood samples for standard hematology and immune assays were collected
This document is a draft for review purposes only and does not constitute Agency policy.
4-301 DRAFT—DO NOT CITE OR QUOTE
-------
1 (details not given). Immune measures referenced include in vitro lymphocyte transformation,
2 hemagglutination assays, and enumeration of B cells, WBCs, and RBCs. No effects were seen at
3 3 ppm formaldehyde. However, dose-dependent responses were reported for weight loss,
4 decreased food consumption, increased WBCs, increased segmented neutrophils and nucleated
5 RBCs, and decreased ability to produce antibodies to sheep RBCs. The results of the
6 lymphocyte transformation assay were inconsistent, with a 25-30% reduction in stimulation after
7 exposure to 99 ppm but an initial stimulation seen after 16 and 61 ppm exposures. Further
8 details were not available, making it difficult to determine if these reported immunomodulatory
9 effects may have been, in part or in full, secondary to effects on the URT. Subchronic exposures
10 at 61 and 99 ppm formaldehyde would be expected to result in frank toxic effects in mice (see
11 Section 4.2.1). However, these findings suggest possible immunomodulatory effects due to
12 formaldehyde exposure and require further exploration.
13 Dean et al. (1984) investigated the effects of formaldehyde exposure on a range of
14 indicators of immune function. Female B6C3F1 mice were exposed to 15 ppm formaldehyde
15 (18.4 mg/m3) 6 hours/day, 5 days/week for 3 weeks. Three trials were run with a total of 255
16 formaldehyde-treated mice. Body and organ weights were recorded at sacrifice for control and
17 formaldehyde-exposed mice (10 per group). Measures of host susceptibility, cell-mediated
18 immunity MP function, and antibody reactions were conducted 2 to 6 days after the end of
19 exposure (see Table 4-46). Lymphocyte subsets, spleen cellularity, bone marrow cellularity, and
20 progenitor cell subsets were enumerated. Host susceptibility and delayed type hypersensitivity
21 were measured in vivo. Lymphocyte proliferation, natural killer cell activity, phagocytosis,
22 hydrogen peroxide production, and IgM plaque-forming cells (PFCs) were measured ex vivo
23 after in vivo stimulation in some cases (see Table 4-46).
24 Body weight, organ weights and cellularity, progenitor cell populations, blood cell
25 counts, and differentials were unchanged in formaldehyde-treated mice (Dean et al., 1984).
26 Circulating blood monocytes were decreased in treated mice, which may be a reflection of the
27 local inflammatory response expected in the nasal epithelium (Dean et al., 1984). However,
28 there was no corresponding decrease in peritoneal MPs. There was a trend, but no statistical
29 significance, for decreased spleen weight, cellularity, and B cell precursors (87, 83, and 78% of
30 controls, respectively). The mean body weight of formaldehyde-treated mice was 21.1 versus
31 20.9 g in control mice, and thymus and spleen weights were not normalized by body weight.
32 All indicators of natural killer cell function, cell-mediated immunity, and humoral
33 immunity in formaldehyde-treated mice were unchanged from controls (Dean et al., 1984).
34 Phagocytic capacity of both resident and elicited peritoneal MPs was unchanged by formaldehyde
35 treatment. However, hydrogen peroxide production in elicited peritoneal MPs was significantly
This document is a draft for review purposes only and does not constitute Agency policy.
4-302 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
increased in formaldehyde-treated mice, 78 versus 42 nmol/mg protein (p < 0.05) (Dean et al.,
1984).
Table 4-46 Battery of immune parameters and functional tests assessed in
female B6C3F1 mice after a 3 week, 15-ppm formaldehyde exposure
(6 hours/day, 5 days/week)
Immune function
Host susceptibility
Cell-mediated
immunity
MP function
(both resident and
MVE-1 elicited
MP)
Humoral cell
immunity
Progenitor cells
Model
Tumor resistance
Tumor resistance
Bacterial resistance
Delayed type
hypersensitivity
Lymphocyte
proliferation
Lymphocyte subsets
Natural killer cell
activity
Phagocytosis
Hydrogen peroxide
production
Antibody PFC
responses, IgM PFCs
Bone marrow
cellularity (femur)
Granulocyte-MP
progenitors
B-cell precursors
Challenge
PYB6 sarcoma cells
16F10 melanoma cells
Listeria monocytogenes
Keyhole limpet
hemocyanin
T-cell mitogen, PHAa
B-cell mitogen, LPSb
(ex vivo)
None
Yac-1 target cells
(51Cr labeled)
(ex vivo)
Sheep RBCs
(51Cr labeled)
(ex vivo)
Pharmacologic
stimulation (ex vivo)
Sheep RBCs,
TVF-LPS, or TNF-Ficoll
None
None
None
Metric
Subcutaneous injection, folio wed by skin
palpation to track tumor development
Lung tumor burden determined by [125I]UdR
incorporation
Survival after challenge
Radiometric index of delayed
hypersensitivity responses
Ex vivo proliferation, 3 days, measured by
[3H]-thymidine incorporation
Percentage of cells positive for cell surface
markers (Thy-1, Mac-1, Lyt-1)
% cytotoxicity by 51Cr release
51 Cr incorporation as a measure of RBCs
phagocytized
H2O2 release in culture
Plaques formed
Cell enumeration by a Coulter counter
Cell enumeration by a Coulter counter
Clonogenic assay
9
10
11
12
13
14
15
aT-cell mitogen, phytohemagglutinin (PHA-P).
bB-cell mitogen, lipopolysaccharide (Escherichia coli).
Source: Dean etal. (1984).
As shown in Table 4-47, several indicators of host resistance in the female B6C3F1 mice
were increased after formaldehyde exposure (Dean et al., 1984). Tumor mass and pulmonary
foci after B16F10 melanoma cell challenge were significantly reduced in formaldehyde-treated
This document is a draft for review purposes only and does not constitute Agency policy.
4-303 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
mice, indicating improved tumor immunity (p < 0.05). However, following PYB6 sarcoma cell
challenge, formaldehyde-treated mice had a 7.1% tumor incidence versus 11.1% in controls,
which was not statistically different. Mortality due to Listeria monocytogenes (LM) was
decreased from 70 to 30% (p < 0.05). Because resistance to LM is primarily MP dependent, the
authors speculated that this enhanced resistance might be due in part to increased bactericidal
activity as was also suggested by increased hydrogen peroxide production ex vivo in elicited
peritoneal MPs from female mice (Dean et al., 1984).
Table 4-47. Summary of the effects of formaldehyde inhalation on the
mononuclear phagocyte system (MPS) in female B6C3F1 mice after a
3-week, 15 ppm formaldehyde exposure (6 hours/day, 5 days/week)
In vivo indicators of MPS
Cellularity
In vivo test of host resistance
Ex vivo indicators of MPS
H2O2 production
Phagocytosis
Assessment of MP
maturation
Leucine aminopeptidase content
Acid phosphatase content
Binding of tumor cells
Lysing of tumor cells
Metric
Circulating monocytes
CMF progenitor cells
Resident peritoneal MP
Elicited peritoneal MP
LM
B16F10 tumor challenge
PYB6 tumor challenge
Cell type/activation
Resident, no PMA°
Resident, with PMA
Elicited, no PMAd
Elicited, with PMA
Resident
Elicited
Resident
Elicited
Resident
Elicited
Resident
Elicited
Resident
Elicited
Formaldehyde effect
Decreased3
No change3
No change3'15
No change3'15
Increased resistance3
Increased resistance3
No significant increase3
Formaldehyde effect
None detected 3'b
None detected3-13
None detected3-13
Increased 3>b
No change3
No change3
Decreased15
No changeb
No changeb
No changeb
No changeb
No changeb
No changeb
Increased at mid-range target-to-effector cell ratiob
13
14
15
16
17
3Deanetal. (1984).
bAdamsetal. (1987).
"Phorbol 12-myristate 13-acetate (PMA).
dPeritoneal MPs were elicited with the pyran copolymer Murray Valley encephalitis virus (MVE-2).
Sources: Adams et al. (1987); Dean et al. (1984).
This document is a draft for review purposes only and does not constitute Agency policy.
4-304 DRAFT—DO NOT CITE OR QUOTE
-------
1 Overall, the observations of increased hydrogen peroxide production and increased host
2 resistance in peritoneal MPs distant from the POE suggest that formaldehyde has an effect on the
3 mononuclear phagocyte system (MPS). The authors postulated that this effect may be indirect,
4 due in part to the tissue inflammatory response in the URT or a direct systemic effect on the
5 MPS by formaldehyde exposure (Dean et al., 1984). Subsequent studies by the same researchers
6 explored the possibility of systemic effects of formaldehyde exposure on MPS function and
7 maturation stage (Adams et al., 1987). Female B6C3F1 mice were exposed to 15 ppm
8 (18.4 mg/m3) formaldehyde 6 hours/day, 5 days/week for 3 weeks, as before (Adams et al.,
9 1987). Both resident and Murray Valley encephalitis virus (MVE-2)-elicited peritoneal MPs
10 were examined for hydrogen peroxide production, enzymatic activity, phagocytic ability,
11 binding, and lysis of tumor cells (Adams et al., 1987).
12 Similar to the findings of Dean et al. (1984), formaldehyde treatment increased hydrogen
13 peroxide production almost twofold in MVE-2 elicited peritoneal MPs (Adams et al., 1987). As
14 summarized in Table 4-47, no treatment differences were seen in phagocytic ability in either
15 resident or elicited MPs (Adams et al., 1987). Resident peritoneal MPs from formaldehyde-
16 treated mice were not different in their ability to bind or lyse tumor cells. Although
17 formaldehyde treatment did not increase the ability of elicited MPs to bind tumor cells, lysis of
18 the target cells (P815 tumor cells) was increased from 28 to 37% by formaldehyde treatment but
19 only at the midrange target-to-effector-cell ratio tested in the assay (p < 0.05) (Adams et al.,
20 1987). Although this is statistically significant, the authors questioned the biological
21 significance of this result since it was not observed at all three target cell ratios tested. However,
22 an increase in tumor cell lysis in vitro would be consistent with the in vivo increased tumor
23 resistance previously reported (Dean et al., 1984). The in vitro lysis response curve suggests that
24 assay conditions may result in a maximum cytolysis near 40%. If so, any treatment effects on
25 lysis would be difficult to discern at higher effector cell ratios.
26 Jakab (1992) investigated the effect of formaldehyde exposure on the alveolar MPs and
27 resistance to respiratory infections. The first set of experiments assessed bactericidal activity by
28 directly quantifying the pulmonary bacterial loading after exposure to Staphylococcus aureus.
29 White female Swiss mice were exposed to formaldehyde after bacterial infection (regimens A
30 and C), before bacterial infection (regimen B), or before and after infection (regimen D) (see
31 Table 4-48).
32
This document is a draft for review purposes only and does not constitute Agency policy.
4-305 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-48. Formaldehyde exposure regimens for determining the effects of
formaldehyde exposure on pulmonary S. aureus infection
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Regimen A
Regimen B
Regimen C
Regimen D
Preinfection treatment
None
18 hours
0, 0.5, or 1 ppmb
None
18 hours
0, 0.5, or 1 ppm
Postinfection treatment
4 hours
0, 1, 5, 10, or 15 ppnf
None
4 hours
0, 0.5, or 1 ppm
4 hours
0, 0.5, or 1 ppm
Results
15 ppm, increased
bacterial loading
No effect
No effect
1 ppm, increased
bacterial loading
aO, 1.2, 6.2, 12.3, or 18.5 mg/m3 formaldehyde.
bO, 0.62, or 1.2 mg/m3 formaldehyde.
Source: Jakab (1992).
For regimen A, mice were exposed to 0, 1, 5, 10, or 15 ppm (0, 1.2, 6.2, 12.3, or
18.5 mg/m3) formaldehyde. For regimens B-D, mice were exposed to 0, 0.5, or 1 ppm (0, 6.2,
or 1.2 mg/m3) formaldehyde. A 30-minute exposure to an infectious aerosol of S. aureus
deposited 2 x 105 staphylococci in the lungs. Bacterial loading was determined in homogenized
lung tissue by culturing diluted aliquots for an estimate of bacteria present immediately after
loading and 4 hours later. Bacterial loading was expressed as a percentage change between
control and formaldehyde-exposed animals. Mice exposed to 15 ppm formaldehyde for the
4 hours following bacterial infection (regimen A) had approximately an 8% increase in bacteria,
indicating decreased host resistance (p = 0.006) (Jakab, 1992) (see Table 4-48). Mice receiving
lower concentrations of formaldehyde following bacterial infection did not have increased
pulmonary bacterial loading. Preinfection exposure to 0.5 or 1.0 ppm did not change bacterial
loading 4 hours after infection (regimen B). However, combining an 18-hour preinfection
formaldehyde exposure with a 4-hour postinfection 1 ppm formaldehyde exposure increased
pulmonary bacterial loading by approximately 6.5% (p < 0.05). This effect was not seen with
only a 0.5 ppm pre- and post-treatment regimen. Increased bacterial loading indicates that
formaldehyde exposure (regimens A and D) reduced pulmonary bacterial resistance. This is in
apparent contradiction to the findings of increased host resistance by Dean et al. (1984).
However, there are important differences between the studies. The studies by Jakab (1992) are
acute studies examining effects at the respiratory tract where direct effects are possible.
Additionally, in some cases, the exposures were concurrent with bacterial infection, and it is
This document is a draft for review purposes only and does not constitute Agency policy.
4-306 DRAFT—DO NOT CITE OR QUOTE
-------
1 difficult to distinguish the potential for formaldehyde effects directly on the mucociliary
2 apparatus as a barrier to infection.
3 A second set of experiments in the same report (Jakab, 1992) examined the effects of
4 coexposure to formaldehyde and carbon black on pulmonary infection with S. aureus. The
5 particle size distribution of the carbon black aerosol was less than a 5 jim aerodynamic diameter
6 and, therefore, 98% respirable. Female Swiss mice were exposed nose only to formaldehyde and
7 carbon black. Experiments were run at two target concentrations: (1) 2.5 ppm (3.1 mg/m3)
8 formaldehyde and 3.5 mg/m3 carbon black or (2) 5 ppm (6.2 mg/m3) formaldehyde and
9 10 mg/m3 carbon black. Coexposure was given either for 4 hours after a 30-minute S. aureus
10 infection or 4 hours/day for 4 days as a pretreatment prior to S. aureus infection. Bacterial
11 loading was determined 0 and 4 hours after the S. aureus infection to assess bacterial survival.
12 Formaldehyde-carbon black coexposure did not alter bacterial survival either as a pretreatment or
13 post-treatment to bacterial exposure. However, this exposure regimen was not run for
14 formaldehyde or carbon black separately, and the 4 hours/day for 4 days pretreatment was not
15 included in the formaldehyde alone experiments (see Table 4-48).
16 Jakab (1992) also assessed the phagocytic activity of alveolar MPs collected by lavage at
17 various time points after formaldehyde, carbon black, or coexposure. Female Swiss mice were
18 coexposed to 5 ppm (6.2 mg/m3) formaldehyde and 10 mg/m3 carbon black 4 hours/day for
19 4 days. Mice were sacrificed and alveolar MPs harvested 1, 3, 5, 25, and 40 days after exposure.
20 Mice exposed only to formaldehyde or carbon black were sacrificed 3, 10, 25, and 40 days after
21 exposure. Fc-receptor-mediated phagocytosis was assessed ex vivo by using sensitized sheep
22 RBCs. The phagocytic index (PI) was reported as the total number of RBCs in 100 MPs.
23 Neither formaldehyde nor carbon black exposure alone significantly changed the PI (Jakab,
24 1992). These findings are consistent with the first coexposure experiment, since no changes in
25 PI were seen immediately after exposure. However, coexposure did decrease the PI of alveolar
26 MPs in a time-dependent manner, with maximal decrease to less than 70% of controls by 25 days
27 after exposure (see Figure 4-15). Decreases in the PI reflect changes in both the percentage of
28 phagocytic MPs and the number of RBCs phagocytized (Jakab, 1992). The PI recovered to
29 control levels by 40 days postexposure.
30
This document is a draft for review purposes only and does not constitute Agency policy.
4-307 DRAFT—DO NOT CITE OR QUOTE
-------
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
— 110
c
O
o
c
o
o
o
U)
03
.C
Q.
.C
Q.
O HCHO
D Carbon
• Carbon + HCHO
10 20 30
Days after last exposure
40
Figure 4-15. Alveolar MP Fc-mediated phagocytosis from mice exposed to
5 ppm formaldehyde, 10 mg/m3 carbon black, or both.
Note: Exposure was 4 hours/day for 4 days. Each value represents the mean ±
SEM of five determinations.
Source: Redrawn from Jakab (1992).
Holmstrom et al. (1989b) evaluated the effects of long-term formaldehyde exposure on
antibody production. Female Sprague-Dawley rats were exposed to 12.6 ppm formaldehyde
(15.5 mg/m3) 6 hours/day, 5 days/week for 22 months. Body weight, tumor incidence, and
pathology were reported elsewhere (Holmstrom et al., 1989b). Rats were given a subcutaneous
injection of pneumococcal polysaccharide antigens or tetanus toxoid 21 to 25 days prior to
sacrifice. The two vaccines chosen represent T-cell-dependent and T-cell-independent antigens,
respectively. Antibody liters (IgG and IgM) were determined prior to vaccination and at
sacrifice. Formaldehyde treatment had no effect on antibody liters either before or after
vaccination (Holmstrom et al., 1989b).
Summary of Formaldehyde Effects on Immune Function
Although there were initial reports of systemic immunomodulation attributed to
formaldehyde exposure (Leach et al., 1983), formaldehyde effects on measures of humoral and
cell-mediated immunity were not confirmed by Dean et al. (1984). The authors did report
increased host resistance to both tumor and bacterial tumor challenges after a 3-week exposure to
This document is a draft for review purposes only and does not constitute Agency policy.
4-308 DRAFT—DO NOT CITE OR QUOTE
-------
1 15 ppm formaldehyde. An increased resistance to these challenges, presented distal to the site of
2 formaldehyde exposure (administered subcutaneously or intravenously), suggests a systemic
3 effect of formaldehyde exposure. In addition, increased host resistance and hydrogen peroxide
4 release from peritoneal MPs were reported and confirmed (Adams et al., 1987; Dean et al.,
5 1984). Chronic inflammation and tissue damage to the respiratory mucosa expected with
6 formaldehyde exposure may result in an up regulation of the MPS and therefore increase host
7 immunity. It is unclear if this response would be specific to formaldehyde or similar to
8 enhancement of immune function seen with chronic inflammation.
9 Jakab (1992) demonstrated decreased pulmonary resistance to bacterial infection where
10 animals were exposed to 15 ppm formaldehyde immediately after bacterial loading or when they
11 were given an 18-hour pre-exposure to formaldehyde followed by 1 ppm formaldehyde exposure
12 after bacterial loading. The authors speculated that formaldehyde may directly act on pulmonary
13 MPs, reducing their effectiveness. However, Jakab (1992) showed that there was no change in
14 Fc-mediated phagocytosis of alveolar MPs immediately after formaldehyde exposures.
15 Degradation of the protective mucus layer and possible epithelial cell damage may contribute to
16 more effective bacterial infection in the presence of formaldehyde without a direct action on MP
17 function. As mentioned above, degradation of the mucus layer may result in a more potent
18 inoculation and therefore higher bacterial loading.
19 Although neither formaldehyde nor carbon black alone impacted Fc-mediated
20 phagocytosis of alveolar MPs, Jakab (1992) demonstrated that there was decreased Fc-mediated
21 phagocytosis after formaldehyde and carbon black coexposure. Carbon black may have acted as
22 a carrier for formaldehyde, allowing higher levels of formaldehyde to be delivered more deeply
23 into the lungs than would be seen with formaldehyde alone.
24 Formaldehyde is known to break down the mucus layer protecting the respiratory tract,
25 allowing exposure of the underlying epithelium (Morgan et al., 1986a, c, d). Additionally,
26 formaldehyde can directly induce tissue inflammation through sensory irritation via substance P
27 from the trigeminal nerve (Fujimaki et al., 2004a). These actions together could contribute to
28 some of the observed effects on immune response attributed to formaldehyde exposures.
29 Degradation of the protective mucus layer would make antigens more available to the immune
30 system. It has been shown that direct application of an antigen to the nasal associated lymph
31 tissue, bypassing the mucus layer, is a more effective delivery of antigen (Hou et al., 2002).
32 Therefore, increased availability of these antigens to the immune system may in part explain
33 observed increased antibody production seen against ovalbumin (OVA) or common dust mite
34 allergen (Der f) during formaldehyde exposure (Sadakane et al., 2002; Riedel et al., 1996;
35 Tarkowski and Gorski, 1995). Neurogenic inflammation may also contribute to more efficient
This document is a draft for review purposes only and does not constitute Agency policy.
4-309 DRAFT—DO NOT CITE OR QUOTE
-------
1 antigen processing and presentation by activation of resident MPs. These factors are consistent
2 with the observation that formaldehyde exposures do not affect antibody production to antigens
3 administered outside of the respiratory tract, even after chronic exposures (Holmstrom et al.,
4 1989b).
5 This effect was initially observed several days after exposure was ended with maximal
6 suppression seen 25 days after a 4-day formaldehyde exposure. The delayed onset of this
7 response, however, suggests an effect beyond the POE effects observed at the time of exposure.
8 Table 4-49 presents a summary overview of the effects of formaldehyde on immune function in
9 laboratory animals.
10
11 4.2.5. Hyper sensitivity and Atopic Reactions
12 Adverse reactions in humans exposed to formaldehyde in the workplace and homes have
13 been reported, which are consistent with an allergic response or a chemical sensitivity (see
14 Section 4.1.1 for details). Rashes and skin reactions are reported in some individuals after
15 dermal exposures, and in some cases exacerbation of asthma is reported after inhalation of
16 formaldehyde. However, the reports of human reactions do not allow a clear determination of
17 whether this sensitization is immunogenic or neurogenic in origin. Formaldehyde-induced
18 sensitization may have both neurogenic and immunologic components. Numerous animal
19 studies have been conducted in order to understand the potential for sensitization to
20 formaldehyde. Although hypersensitivity and allergic sensitization are often considered solely
21 immunologic in origin, neurogenic mechanisms may result in bronchial hypersensitivity and
22 increased immunologic sensitization. Therefore, the animal studies regarding formaldehyde-
23 induced sensitization are evaluated discretely in order to examine these etiologic possibilities.
24 Classically, hypersensitivity is characterized as an immune response to an antigen,
25 resulting in an inflammatory reaction that itself damages the tissues or is otherwise harmful
26 (Kuby, 1991). These reactions may be localized, as in topical dermatitis, or systemic, as in
27 anaphylactic shock from an allergen. Hypersensitivity can be mediated by a humoral immune
28 response or by a cell-mediated immune response. Four classes of hypersensitivity are generally
29 recognized that differ in their immune system components and functions. Although a single
30 agent (e.g., penicillin) may induce all four types of hypersensitivity, it is more usual for an agent
31 to primarily induce one form of hypersensitivity.
This document is a draft for review purposes only and does not constitute Agency policy.
4-310 DRAFT—DO NOT CITE OR QUOTE
-------
ts
Table 4-49. Summary of immune function changes due to inhaled formaldehyde exposure in experimental
animals
^ 8
1!
f^ S
s ,
Species
F344 rats
B6C3F1
mice
(female)
B6C3F1
mice
(female)
White Swiss
mice
(female)
White Swiss
mice
(female)
Sprague-
Dawley rats
(female)
No./
group
8
10
Pooled
MPs from a
number of
mice
18
18
5
Treatment"
0,3, 16,61, 99ppm6
hours/day,
5 days/week for
4 weeks
15 ppm 6 hours/day, 5
days/week for
3 weeks
15 ppm 6 hours/day, 5
days/week for
3 weeks
0, 1,5, 10, or 50 ppm
for 18 hours before
and/or 4 hours after a
30-minute exposure to
bacterial infection (S.
aureus)
5 ppm (2.6 mg/m3)
formaldehyde and
10 mg/m3 carbon
black 4 hours/day for
4 days
12.6 ppm 6 hours/day,
5 days/week,
22 months
Observations
No effects at 3 ppm. Mixed results at higher doses that were not
consistent.
Increased H2O2 production, and increased host resistance to tumor
formation, but other immune parameters unchanged.
Increased H2O2 production in MVE-2 -elicited peritoneal MPs.
Combining an 18-hour pre-exposure to formaldehyde with 4-hour
postexposure to formaldehyde increased bacterial loading at 1 ppm by
6.5%.
Phagocytic index was decreased by coexposure to formaldehyde and
carbon black but not by either insult alone.
Formaldehyde treatment had no effect on antibody liters either before or
after vaccination with pneumococcal polysaccharide antigen or tetanus
toxoid.
LOAEL/
NOAEL
NAb
LOAEL 15 ppm
LOAEL 15 ppm
LOAEL 1 ppm
NA
NA
Reference
Leach et al.
(1983)
Dean et al.
(1984)
Adams et al.
(1987)
Jakab (1992)_
Jakab (1992)
Holmstrom et
al. (1989b)
fe
H
NA = not applicable.
O
o
-------
1 Chemical sensitivity generally implies a neurogenically induced sensitization (Meggs,
2 1995). A chemical may directly interact with sensory nerves, releasing mediators or
3 neuropeptides such as substance P (a tachykinin) that trigger inflammation, hence called
4 neurogenic inflammation,. Repeated exposure to the same chemical is hypothesized to potentiate
5 neurogenic inflammation (Meggs, 1995). The resulting signs of tissue inflammation may be
6 similar to immunogenic inflammation (occurs by binding of antigen to antibody or leukocyte
7 receptor), but there would be no requirement that the immune system recognize the chemical as
8 an antigen for this type of response. Therefore, a chemical may induce one or more clinical
9 signs of atopic asthma without IgE-mediated type 1 hypersensitivity response. One form of
10 sensitivity directly affects sensory nerve endings, resulting in neurogenic inflammation and is a
11 well-known health effect attributed to formaldehyde. Neurogenic responses may result from the
12 direct and acute interaction of the chemical with sensory nerve ending receptors of the trigeminal
13 nerve that may lead to persistent rhinitis and an asthma-like reactive airway dysfunction
14 syndrome that may develop after short-term human exposures (Brooks et al., 1985). Thus, there
15 is evidence to suggest that neurogenic inflammation may contribute to observed increases in
16 formaldehyde-induced airway hyperresponsiveness and atopic responses. The available animal
17 studies that have investigated formaldehyde-induced airway hyperresponsiveness and atopic
18 responses are summarized below.
19
20 4.2.5.1. Inhalation Studies in Experimental Animals
21 This section summarizes animal studies informing the role of formaldehyde-induced
22 chemical sensitization. The symptoms of sensitization (atopy, airway hyperresponsiveness) are
23 frequently associated with immunologic markers (cytokine production, leukocyte infiltration
24 histamine release, and antibody production) but may be mediated by neurogenic sensory
25 irritation, principally by activation of the trigeminal nerve (see Section 4.1.1.1 for a discussion of
26 sensory irritation). The animal studies that illuminate these neurogenic and immunologic
27 responses are discussed outside of the classic neurotoxicology and immunotoxicology study
28 summary sections to allow synthesis of these data.
29 Sensitization to chemical exposure by inhalation often manifests as an allergic or
30 asthmatic response as characterized by BC or BHR. This sensitization may be a result of
31 immune involvement, as in the case of hypersensitivity, or a neurogenic sensitization, where a
32 chemical may directly stimulate inflammation. Asthma is a specific manifestation of IgE-
33 mediated hypersensitivity, characterized by BHR and airway inflammation, resulting in lower
34 airway obstruction (Fireman, 2003; Kuby, 1991). In asthma, an allergen capable of cross-linking
35 membrane-bound IgE on mast cells initiates immunogenic inflammation resulting in an influx of
This document is a draft for review purposes only and does not constitute Agency policy.
4-312 DRAFT—DO NOT CITE OR QUOTE
-------
1 eosinophils, neutrophils, and lymphocytes. Mediators of BC, including histamine, eicosanoids,
2 and bradykinin (Kuby, 1991), are released during this process. Prior exposure to the allergen can
3 increase allergen-specific IgE, potentiating the allergic reaction; this is immunogenic
4 sensitization.
5 Biagini et al. (1989) evaluated the effect of a single pulmonary exposure of formaldehyde
6 on pulmonary mechanics, including BC. The researchers chose cynomolgus monkeys known to
7 be hyperreactive to methacholine (acetyl-p-methacholine chloride), which is a direct-acting
8 stimulant of BC (Cain, 2001). Measures of pulmonary mechanics included pulmonary flow
9 resistance; dynamic compliance; PEFR; FVC; FEV; FEF25-75o/0, and 50% of VC; and FEFs
10 normalized for VC. Nine cynomolgus monkeys were exposed to increasing levels of
11 methacholine for 1 minute at 10-minute intervals (0, 0.125, 0.5, 2, and 8 mg/mL) as an aerosol
12 (0.065 mL/minute with a mean aerodynamic diameter of 1.0-1.5 um). Pulmonary mechanics
13 were measured to establish each monkey's response to methacholine. Methacholine challenge,
14 as the positive control, increased pulmonary flow resistance at increasing levels of methacholine
15 (0.125, 0.5, 2, and 8 mg/mL) to 196 ± 16, 285 ± 57, 317 ± 64, and 461 ± 120% of baseline levels,
16 respectively. After a 2-week recovery period, each methacholine-sensitized monkey was
17 exposed to 2.5 ppm formaldehyde (generated from formalin, 15% methanol) for 10 minutes.
18 Measures of pulmonary function were performed at 2, 5, and 10 minutes after exposure.
19 Formaldehyde exposure increased pulmonary flow resistance from 11.3 ± 1.4 cm H2O
20 prior to formaldehyde exposure to 16.1 ±2.1, 16.9 ±2.8, and 20.0 ± 3.4 cm H2O at 2, 5, and
21 10 minutes after 2.5 ppm formaldehyde exposure (with 142, 150, and 177% change,
22 respectively). All other measures of formaldehyde-induced pulmonary mechanics were not
23 significantly different from controls. Increased pulmonary flow resistance, a measure of
24 increased BC, was induced by formaldehyde challenge in previously sensitized mice. However,
25 the differences between methacholine challenge and formaldehyde challenge were not
26 statistically significant. Although both formaldehyde challenge and methacholine challenge
27 increased pulmonary flow resistance, there was no correlation between individual methacholine
28 responsiveness and the magnitude of effect after formaldehyde exposure (p > 0.1). Therefore,
29 although formaldehyde exposure stimulated BC similarly to a known direct stimulating agent,
30 formaldehyde may not work through the same site of action as methacholine.
31 Swiecichowski et al. (1993) assessed pulmonary resistance and airway reactivity due to
32 formaldehyde exposure alone and in response to increasing doses of acetylcholine chloride (a
33 direct-acting BC agent) after formaldehyde exposure in vivo. Male Hartley guinea pigs (eight
34 per group) were exposed at 0.86, 3.4, 9.4, or 31.1 ppm (1.1, 4.2, 11.6, or 38.3 mg/m3)
35 formaldehyde for 2 hours or at 0.11, 0.31, 0.59, or 1.05 ppm (0.14, 0.38, 0.73, or 1.29 mg/m3)
This document is a draft for review purposes only and does not constitute Agency policy.
4-313 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde for 8 hours. Total pulmonary resistance increased after 2 hours formaldehyde
2 exposure at 9.4 and 31.1 ppm and reached similar peak resistance at the end of the exposure
3 period. This effect was rapidly reversible, with values returning to baseline within 30 minutes
4 after exposure. Although 2-hour exposures at 3 and 1 ppm did not increase pulmonary
5 resistance, 8-hour exposures at 0.3 and 1 ppm did increase pulmonary resistance to similar levels
6 as the 2-hour exposure at 30 ppm. The results indicate that both concentration and exposure time
7 impacted the measured increase in pulmonary resistance. However, a simple multiplicative
8 model (e.g., C x t) does not adequately represent the effects observed. It is noted that an 8-hour
9 exposure at 1 ppm (8 ppm-hours), reached approximately the same pulmonary resistance as 2
10 hours at 9.4 ppm (19 ppm-hours). This may in part be due to a maximum practical increase in
11 pulmonary resistance in the animals. Conversely, there was no effect at 3 ppm for 2 hours
12 (6 ppm-hours), although significant increase in pulmonary resistance was recorded after an
13 8-hour exposure at 0.3 ppm (2.4 ppm-hours). Formaldehyde does not appear to exert its effects
14 via a classic C x t paradigm. Exposure concentration, however, did seem to impact recovery
15 time.
16 In addition, specific pulmonary resistance and airway reactivity to increasing doses of
17 intravenous acetylcholine chloride, a direct respiratory stimulant, were measured immediately
18 after formaldehyde exposure for up to 60 minutes. Formaldehyde-induced airway
19 hyperreactivity was defined as a decrease in the level of acetylcholine chloride needed to
20 produce twice the basal specific resistance (effective dose [ED]20o). The dose of acetylcholine
21 chloride required to double the specific pulmonary resistance (ED2oo) and airway reactivity was
22 decreased in animals exposed for 2 hours to formaldehyde. When the duration was extended to 8
23 hours of formaldehyde exposure, the effective dose of formaldehyde required to elicit a doubled
24 pulmonary resistance (ED2oo) in the presence of acetylcholine chloride was decreased to
25 1.07 ppm. Lower ED20oS were recorded in formaldehyde-treated animals. This indicates that
26 less acetylcholine was needed to produce BC when formaldehyde was present. Thus,
27 formaldehyde can exacerbate BHR. Additionally the formaldehyde-induced effect increased
28 with duration of exposure, indicating that time as well as exposure concentration are factors in
29 the magnitude of the response. Directly induced increases in airway hyperreactivity peaked 1
30 hour after exposure and persisted 6 hours after exposure.
31 In a second set of experiments, male Hartley guinea pigs were treated for 8 hours at
32 3.4 ppm (4.2 mg/m3) in order to measure airway hyperreactivity ex vivo (Swiecichowski et al.,
33 1993). After formaldehyde exposure, tracheae were excised and mounted in tissue baths, where
34 tracheal contraction was measured in response to direct application of acetylcholine and then
35 carbachol. Tracheae from similarly exposed guinea pigs were fixed and sectioned for histologic
This document is a draft for review purposes only and does not constitute Agency policy.
4-314 DRAFT—DO NOT CITE OR QUOTE
-------
1 examination and were assessed for signs of inflammation. Formaldehyde exposure did not
2 increase ex vivo tracheal constriction and suggests that changes in airway reactivity were
3 produced due to both local humoral changes and neural reflexes. However, no changes in
4 epithelial cell morphology or influx of inflammatory cells were recorded even up to 4 days after
5 formaldehyde exposure ended. The authors speculated on possible MO As for BHR, such as the
6 role of an irritant receptor or altered epithelial cell biochemistry. It may be that the window of
7 acute inflammation occurred early in the exposure protocol and was resolved by the time of first
8 measurement, after 8 hours of exposure. The absence of inflammatory markers may argue
9 against a classic type 1 sensitivity.
10 The binding of an allergen to receptor bound IgE triggers degranulation of mast cells and
11 basophils, releasing mediators of type 1 hypersensitivity, including the histamine responsible for
12 BC. Brown Norway (BN) rats are known for their high capacity for IgE production and airway
13 hyperresponsiveness in response to allergens or other chemicals; they have often been used as a
14 model of allergic respiratory disease. Ohtsuka et al. (1997) compared the effects of
15 formaldehyde exposure on the nasal epithelium of F344 and BN rats. If the formaldehyde-
16 induced inflammatory response in the nasal epithelium is IgE mediated, BN rats would be
17 expected to display more severe effects of formaldehyde exposure than F344 rats. Both strains
18 of age- and sex-matched rats were exposed to formaldehyde aerosol for 3 hours/day, 5
19 days/week for 2 weeks. The aerosol was generated from a 1% formaldehyde solution by a two-
20 fluid atomizer, and formaldehyde level was maintained at 2 mg (1% sol.)/L (approximately
21 16 ppm or 20 mg/m3), by adjusting the flow rate for formaldehyde solution to the atomizer.
22 During the course of exposure, the following clinical signs were monitored: abnormal
23 respiration, stridor wheezing, nasal discharge, and sneezing. Rats were weighed weekly. Two
24 days postexposure, rats were sacrificed and tissues from the head, trachea, and lungs were fixed
25 and sectioned. Transverse sections were taken at the following palatal landmarks from three
26 animals: level 1 (lateral edge of incisor teeth), level 2 (between incisive papilla and the first
27 palatal ridge), and level 3 (on the second upper molar). The nasal septa of the remaining two
28 animals were revealed for examination by electron microscopy.
29 Formaldehyde-treated F344 rats showed less body weight gain over the 2-week
30 treatment, resulting in lower body weight at week 1 and week 2 than F344 controls (p < 0.05 and
31 0.01). Body weights of formaldehyde-treated BN rats were unchanged from BN controls. The
32 authors observed fewer clinical signs of respiratory irritation in the formaldehyde-exposed BN
33 rats compared with formaldehyde-exposed F344 rats, such as abnormal respiration (three versus
34 five) and nasal discharge (three versus five). Histologic analysis of lung and trachea tissues
35 revealed no distinct signs of inflammation in either strain. Formaldehyde exposure induced cell
This document is a draft for review purposes only and does not constitute Agency policy.
4-315 DRAFT—DO NOT CITE OR QUOTE
-------
1 damage in URT tissues. Epithelial cell damage was milder and impacted a smaller portion of the
2 URT in BN rats compared with F344 rats. Squamous metaplasia were present in the respiratory
3 epithelium (levels 1 and 2) in both strains in formaldehyde-treated rats. However, a distinct
4 keratinized layer was noted in level 1 epithelium of F344 rats, and the extent of lesions in level 2
5 respiratory epithelium was much greater than that seen in BN rats. Additionally, the olfactory
6 epithelium (level 2) in formaldehyde-exposed F344 rats exhibited degeneration, necrosis, and
7 desquamation not seen in BN rats. Mild squamous metaplasia was noted in level 3 of the
8 respiratory epithelium in the treated F344 rats but not the treated BN rats. No pulmonary
9 function measurements were taken, and, thus, no direct comparison in BHR or BC between BN
10 and F344 rats in response to formaldehyde can be made. It appears that BN rats are more
11 resistant to formaldehyde-induced cell damage than are F344 rats, despite the fact that BN rats
12 are known to be IgE responders. These results suggest that IgE responsiveness may be
13 protective of formaldehyde-induced cell damage, or IgE may not play a role at all. The authors
14 note that their earlier research indicated the BN rats have well-developed submucosal glands and
15 speculate that greater mucus flow may be partly responsible for the greater resistance of BN rats
16 to the histologic signs of formaldehyde toxicity.
17 In a subsequent study in the same laboratory, Ohtsuka et al. (2003) compared histology
18 and cytokine profiles in the nasal mucosa of formaldehyde-treated F344 and BN rats.
19 Formaldehyde aerosol was generated as above and rats (nine per group) were exposed
20 3 hours/day for 5 days to approximately 16 ppm of formaldehyde (20 mg/m3). Clinical signs
21 were recorded daily, and monitored respiratory parameters included abnormal respiration, stridor
22 wheezing, nasal discharge, and sneezing. Tissue sections of the nose (five rats per group) were
23 prepared for light microscopy as above: transverse sections at levels 1, 2, and 3. Th-1 cytokines
24 (LFN-y, IL-2) and Th2 cytokines (IL-4 and IL-5) were determined from the whole nasal mucosa
25 in four rats of each treatment group.
26 As expected, lesions and neutrophilic infiltration were more severe in F344
27 formaldehyde-exposed rats compared with treated BN rats. In addition, lesions were observed in
28 all three levels of epithelium examined in F344 rats and impacted both respiratory and olfactory
29 epithelium. Mucosal lesions in formaldehyde-treated BN rats impacted the respiratory
30 epithelium of levels 1 and 2 only. Changes in formaldehyde-induced cytokine mRNA
31 expression were modest in both strains. Th-1-related cytokines (IFN-y, 11-2) in formaldehyde-
32 treated BN rats were significantly decreased compared with control BN rats. A similar, although
33 not statistically significant, decrease in Th-2 cytokines (IL-4, IL-5) was observed in
34 formaldehyde-treated BN rats compared with unexposed BN rats. There were no treatment
35 differences in either Th-1 or Th-2 cytokine expression in formaldehyde-treated F344 rats
This document is a draft for review purposes only and does not constitute Agency policy.
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
compared with unexposed F344 rats. The modest changes in cytokine profile reported in
formaldehyde-treated BN rats were not consistent with type 1 hypersensitivity since type 1
hypersensitivity reactions generally result in increased Th-2 cytokines. The mRNA expression
results were not corroborated with protein levels and may not have been captured at their peak
expression levels.
Lee et al. (1984) evaluated the potential for formaldehyde to act as a sensitizing agent
through different routes of exposure in guinea pigs. The inhalation studies will be highlighted
here. Dermal exposure and associated contact sensitivity results will be discussed in the dermal
exposure section (see Section 4.2.5.2). Three groups of male English smooth-haired guinea pigs
(four/group) were exposed via inhalation to either 6 or 10 ppm (7.4 or 12.3 mg/m3)
formaldehyde 6 hours/day for 5 consecutive days. Depending on the group, animals were then
subjected to bronchial provocation challenge with 2 or 4 ppm formaldehyde on day 7 or days 7,
22, and 29 after exposure (see Table 4-50 for clarification).
Table 4-50 Study design for guinea pigs exposed to formaldehyde through
different routes of exposure: inhalation, dermal, and injection
Group I — Inhalation
Group II — Inhalation
Group III — Inhalation
Group IV — Dermal
Group V — Injection
Formaldehyde
exposure
6 ppm formaldehyde3,
days 1-5
10 ppm formaldehyde,
days 1-5
10 ppm formaldehyde,
days 1-5
100 uL formalin, days 1
and 3
37 mg formaldehyde
with Freund's adjuvant
Bronchial provocation
challenge
Day 7
2 ppm formaldehyde3 for 1 hour
Day 7
2 ppm formaldehyde for 1 hour
Days 7, 22, and 29
4 ppm formaldehyde for 4 hours
Day 22
2 ppm formaldehyde for 1 hour
4 ppm formaldehyde for 4 hours
Day 19
2 ppm formaldehyde
Skin test
Day 9
Day 9
Day 31
Day 7
Day 7
Blood drawn
for antibody
titer
Day 14
Day 14
Day 14
Day 14
Day 14
18
19
20
21
22
23
24
25
26
Source: Lee etal. (1984).
Dermal and injection groups are shown for comparison. Pulmonary hypersensitivity was
assessed by measuring respiratory rate and tidal volume in response to exposure to 2 ppm
formaldehyde challenge for 1 hour, 2 days postexposure for all three groups, and additional
measurements were taken 22 and 29 days postexposure for group III. Blood was drawn to
characterize IgE antibodies to formaldehyde in a passive cutaneous anaphylaxis (PCA) assay.
This document is a draft for review purposes only and does not constitute Agency policy.
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
Respiratory rate was measured following initial formaldehyde exposure and again after bronchial
challenge with formaldehyde.
Respiratory rate (exhibited as a pause during expiration) was depressed by 45% following
exposure to 10 ppm formaldehyde during the first hour of exposure. During the first hour of
exposure, decreased respiratory rate was accompanied by a pause during expiration that has been
categorized as RB and indicated sensory irritation. The decreased respiratory rate is consistent
with URT sensory irritation and induction of the trigeminal (neurogenic) reflex (Lee et al.,
1984). After the first hour of exposure, decreased respiratory rate was characterized by a pause
between breaths, which is similar to the breathing pattern seen in mice exposed to formaldehyde
via tracheal cannula (Alarie, 1981). This suggests a separate effect of formaldehyde on the LRT
after deep penetration of formaldehyde and suggests pulmonary irritation (Lee et al., 1984).
However, subsequent bronchial provocation challenge with either 2 or 4 ppm
formaldehyde for either 1 or 4 hours failed to elicit immediate or delayed-onset respiratory
sensitization (see Table 4-51). Respiratory rates were reported as being within ±20% of
prechallenge levels (data not shown) and did not reflect statistical significance (Lee et al., 1984).
Moreover, increased respiratory sensitivity was not observed in animals that had received an
emulsification of formaldehyde and Freund's complete adjuvant by injection. Only two to four
animals given formaldehyde injections in the presence of Freund's complete adjuvant developed
a low titer of antibodies to formaldehyde (Lee et al., 1984).
Table 4-51 Sensitization response of guinea pigs exposed to formaldehyde
through inhalation, topical application, or footpad injection
Exposure route
Inhalation
6 ppm (Group I)
10 ppm (Group II)
10 ppm (Group III)
Topical
Injection
Pulmonary sensitization
0/4
0/4
0/4
0/8
0/4
Dermal sensitization
0/4
0/4
2/4
8/8
4/4
Antibody production
0/4
0/4
0/4
0/8
2/4
Source: Lee etal. (1984).
Thus, inhalation exposure to 6 or 10 ppm formaldehyde (8 hours/day for 5 days) followed
by bronchial challenge with 2 or 4 ppm formaldehyde failed to result in respiratory sensitivity
defined as greater than 20% change in respiratory rate. Second, for animals that received an
injection of formaldehyde with Freud's adjuvant, it was not effective in inducing pulmonary
This document is a draft for review purposes only and does not constitute Agency policy.
4-318 DRAFT—DO NOT CITE OR QUOTE
-------
1 sensitivity. While neither inhaled formaldehyde challenge nor injected formaldehyde and
2 Freud's adjuvant emulsion were effective in producing pulmonary sensitivity, this study relied
3 on increased respiratory rate as an indication of hyperresponsiveness and may not be an accurate
4 measure of hyperresponsiveness. Thus, overall, conclusions are uninformative due to study
5 design flaws.
6 Riedel et al. (1996) tested the effects of formaldehyde inhalation on the development of
7 sensitization to a known allergen. Female Perlbright-white Dunkin-Hartley guinea pigs (12 per
8 group) were exposed to 0, 0.13, or 0.25 ppm formaldehyde 8 hours/day for 5 consecutive days.
9 On day 5, the animals were sensitized to the common model allergen, OVA, in a 3-minute, head-
10 only exposure to an aerosol of a 5% OVA solution. A booster sensitization with OVA occurred
11 on day 19. A compressor nebulizer with an output rate of 0.75 mL/minute generated the aerosol.
12 Particle size ranged from 0.5 to 5.0 um. On day 26, bronchial provocation testing was conducted
13 with 1% OVA challenge (aerosol). Blood samples were taken and anti-OVA IgG antibodies
14 were quantified by ELISA. Significant airway obstruction was defined as an increase in
15 compressed air in the lung that cannot be expired. Three guinea pigs were exposed to
16 formaldehyde (0.20 ppm) or clean air for 5 days. Immediately after exposure, lung and tracheal
17 tissues were fixed for histologic and morphometric evaluation. Wall thickness of bronchial and
18 alveolar septa was measured systematically with a microscope-digitizing-table set.
19 Significant airway obstruction as measured by compressed air was seen in 3 of 12
20 controls, 8 of 12 0.13 ppm-exposed, and 10 of 12 0.25 ppm-treated animals after OVA challenge.
21 The average airway obstruction was increased after 0.25 ppm (mean = 0.35 mL,p < 0.01) but not
22 after 0.13 ppm formaldehyde exposure. However, individual response to OVA sensitization was
23 highly varied, and animals exhibiting a 10-fold increase in obstruction (measured as compressed
24 air) were seen in both treatment groups (0.13 and 0.25 ppm). Even at the lower exposure (0.13
25 ppm), biologically significant responses were seen in individuals (see Figure 4-16).
26 Specific anti-OVA antibodies (IgGl class) were not detected in animals prior to
27 sensitization or in control-treated animals after sensitization . Measurable anti-OVA antibodies
28 were elevated in 3 of 12 (at 0.13 ppm) and 6 of 12 (at 0.25 ppm) formaldehyde-treated guinea
29 pigs after sensitization (see Figure 4-17). The average anti-OVA titer for the high-dose group
30 was significantly higher than for controls (p < 0.05). The individual responses at the 0.13 ppm
31 exposure level indicate that, although the average group OVA titer may not have reached
32 statistical significance, there was a measureable biological response in three individuals. These
33 results indicate that formaldehyde exposure can sensitize previously naive (nonsensitized)
34 animals to OVA.
This document is a draft for review purposes only and does not constitute Agency policy.
4-319 DRAFT—DO NOT CITE OR QUOTE
-------
CA (ml)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
1
0,8
0,6
0,4
0,2
0
-v-
•*•
Controls
FA 0.1 3 ppm
FA 0.25 ppm
Figure 4-16. Compressed air in milliliters as parameter for airway
obstruction following formaldehyde exposure in guinea pigs after OVA
sensitization and OVA challenge
Note: CA = compressed air; FA = formaldehyde; — = median; ** =p < 0.01.
Source: Redrawn from Riedel et al. (1996).
1.000
100
10
Anti-OVA-lgG1 [EU]
Controls
FA 0.13 ppm
FA 0.25 ppm
Figure 4-17. OVA-specific IgGl (IB) in formaldehyde-treated sensitized
guinea pigs prior to OVA challenge.
Note: EU = experimental units; FA = formaldehyde; — = median; ** =p < 0.01.
Source: Redrawn from Riedel et al. (1996).
This document is a draft for review purposes only and does not constitute Agency policy.
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
The only significant, treatment-related histologic change was bronchial edema, with
thickening of the bronchial wall in formaldehyde-exposed animals compared with nontreated
animals subjected to OVA sensitization and subsequent OVA challenge. Bronchial walls were
measured as 40.9 ± 2.5 versus 28.2 ±1.2 um. No signs of inflammation in the bronchial mucosa
were seen with this edema.
Tarkowski and Gorski (1995) exposed female Balb/C mice to 0 or 6.63 ppm (0 or
2 mg/m3) formaldehyde for either 6 hours/day for 10 days or 6 hours/day once a week for
7 weeks. All mice were sensitized intranasally to OVA for 10 days or once a week for 7 weeks.
IgE anti-OVA liters were determined from sera collected from four mice every 8 days (1 day
after OVA booster) by PC A. A parallel experiment to compare the role of the route of
administration was conducted with IP. rather than intranasal sensitization (1 ug OVA once every
7 days).
OVA liters increased similarly in control mice and mice exposed to formaldehyde once a
week (see Figure 4-18). In contrast, mice exposed to formaldehyde 6 hours/day for 10
consecutive days at the beginning of the experiment had increased anti-OVA beginning after the
fourth OVA sensitization, which continued to increase through seven doses of OVA to a peak of
70 PCA units (p < 0.01) (see Figure 4-18). Anti-OVA IgE liters were significantly different
between formaldehyde-treated and nonexposed mice.
c-n
U
+* Afl -
P A
o
°- 20 -
J* *
W*
/
•
_ jt) ^^m •
^'-^=~~*r*'^
34567
The number of OVA doses
Figure 4-18. Anti-OVA titers in female Balb/C mice exposed to 6.63 ppm
formaldehyde for 10 consecutive days or once a week for 7 weeks.
Note: • = control mice; + = formaldehyde once a week x 7; • = formaldehyde
10 days.
Source: Redrawn from Tarkowski and Gorski (1995).
This document is a draft for review purposes only and does not constitute Agency policy.
4-321 DRAFT—DO NOT CITE OR QUOTE
-------
1 Intraperitoneal sensitization to OVA was much more effective than intranasal
2 sensitization, resulting in liters as high as 1,000 after 4 weeks. However, there were no
3 differences between controls and animals treated with formaldehyde via the IP. route of
4 exposure. Thus, formaldehyde administered intranasally 6 hours/day for 10 days may facilitate
5 the sensitization to allergens. These changes were not observed when formaldehyde was
6 administered intranasally once a week for 10 weeks or via IP. injection (Tarkowski and Gorski,
7 1995). The authors speculate that formaldehyde may increase permeability of respiratory
8 epithelium and destruction of immunologic barriers. Thus the respiratory tract may become
9 vulnerable to inhaled allergens after formaldehyde exposure (Tarkowski and Gorski, 1995).
10 Ito et al. (1996) conducted three experiments to examine the effects of acute
11 formaldehyde exposure on bronchoconstriction and the mediators of vascular permeability.
12 Male Wistar rats (five to eight per group) were exposed to 0, 2, 5, 15, or 45 ppm (0, 2.5, 6.2,
13 18.5, or 55.4 mg/m3) formaldehyde for 10 minutes. Baseline pulmonary insufflation and blood
14 pressure were determined prior to formaldehyde exposure and monitored throughout the
15 experiment. Vascular leakage was measured by injection of Evans blue dye prior to the
16 experiment and determining extravasation 5 minutes postexposure. Briefly, lungs were perfused
17 with 0.9% saline through an aortic cannula. The lower portion of the trachea and main bronchi
18 were removed, and the Evans blue dye remaining was determined and expressed as ng dye/g
19 tissue. A second experiment was conducted to determine if dye leakage continued to increase
20 after exposure. Seven rats were exposed to 15 ppm formaldehyde for 10 minutes, as above.
21 Evans blue dye was injected 5 minutes postexposure, and tissues were perfused and excised
22 15 minutes later. The final experiment was conducted to determine the effect of certain receptor
23 agonists on the formaldehyde-induced microvascular leakage. Ten groups of Wistar rats (four to
24 seven per group) were exposed to 15 ppm formaldehyde and injected with Evans blue dye, as
25 before. However, each receptor agonist under test or saline sham was injected 4-5 minutes prior
26 to the 10-minute formaldehyde exposure. Agonists tested included tachykinin NKi receptor
27 antagonist (CP-99,994) at 1, 3, or 6 mg/kg; a bradykinin B2 receptor antagonist (HOE 140) at
28 0.65 mg/kg; and a histamine HI receptor antagonist (ketotifen) at 1 mg/kg.
29 Formaldehyde exposure did not change pulmonary insufflation pressure or blood
30 pressure. Formaldehyde increased vascular permeability in a concentration-dependent manner in
31 both the trachea and main bronchi for the first 5 minutes after exposure, as measured by Evans
32 blue dye extravasation (Ito et al., 1996) (see Figure 4-19). Vascular permeability was not
33 increased by formaldehyde exposure from 5 to 15 minutes postexposure (experiment 2).
34 Administration of a selective NKi receptor antagonist (CP-99,994) inhibited the formaldehyde-
This document is a draft for review purposes only and does not constitute Agency policy.
4-322 DRAFT—DO NOT CITE OR QUOTE
-------
1 induced vascular permeability, reducing Evans dye extravasation to control levels at the 3 and 6
2 mg/kg doses (see Figure 4-20).
Trachea
Main Bronchi
4
5
6
7
8
9
10
11
12
13
14
15
16
•3T 200-r
« 100--
w
UJ
Room 2 5 15 45
Formaldehyde
(ppm)
D
2 5 15 45
Formaldehyde
(ppm)
Figure 4-19. Vascular permeability in the trachea and bronchi of male
Wistar rats after 10 minutes of formaldehyde inhalation.
Note: Vascular permeability was tested by an increase in Evans blue dye
extravasation in the tissue. Solid bars: formaldehyde; open bars: room air, n = 1.
Values are the means ± SEM of five to seven animals. *p < 0.05 and ^p < 0.01
versus room-air-exposed group (Williams' test).
Source: Redrawn from Ito et al. (1996).
This document is a draft for review purposes only and does not constitute Agency policy.
4-323 DRAFT—DO NOT CITE OR QUOTE
-------
200r
Trachea
Main Bronchi
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
3
to
0)
•
100..
UJ
0-L
Sham Formaldehyde
Sham Formaldehyde
Figure 4-20. Effect of select receptor antagonists on formaldehyde-induced
vascular permeability in the trachea and bronchi male of Wistar rats.
Note: Vascular permeability was tested by an increase in Evans blue dye
extravasation. Rats were treated i.v. with 1, 3, or 6 mg/kg CP-99,996 (open bars),
0.65 mg/kg HOE 140 (hatched bars), 1 mg/kg ketotifen (solid bars), or vehicle
(shaded bars) before formaldehyde challenge. Sham: animals were exposed to the
sham gas for 15 ppm formaldehyde (10 minutes) after pretreatment with 0.9%
saline (0.5 mL/kg i.v.). Data are the means ± SEM of six to seven rats/group.
*p < 0.05 versus sham-stimulated group (unpaired Student's t test or Welch's
test), f/? < 0.05. ttP < 0-01 versus 0.9% saline-pretreated, formaldehyde-
exposed control group (Williams' test).
Source: Redrawn from Ito et al. (1996).
Neither the bradykinin 82 nor histamine HI receptor agonists affected formaldehyde-
induced vascular permeability (Ito et al., 1996). Therefore, the immediate effect of
formaldehyde exposure on vascular permeability is mediated, at least in part, through the NKi
receptor but does not seem to require the B2 or HI receptors. This implies a role for tachykinins
in formaldehyde-induced vascular permeability. These findings suggest a neurogenic
inflammatory response because the tachykinins are released from sensory nerve endings in the
trachea and bronchi, whereas bradykinin is released from mast cells.
Sadakane et al. (2002) investigated the effects of formaldehyde exposure on airway
inflammation caused by Der f Two groups of male outbred ICR mice (18/group) were exposed
to an aerosol of 0.5% formaldehyde solution produced by an ultrasonic nebulizer for 15 minutes,
once a week for 4 weeks. Two groups were similarly treated but exposed to saline aerosol only.
Details of the aerosol generation and resulting magnitude of exposure were not given. One
group each of control and formaldehyde-exposed mice was sensitized to Der f by an injection 1
This document is a draft for review purposes only and does not constitute Agency policy.
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
day prior to formaldehyde exposure (1.5 mg/animal). The same groups were challenged with
intratracheal instillation of Der f (10 ug/animal) after 4 weeks. Three days after allergen
challenge, mice were sacrificed and blood plasma and lung tissue were collected. Blood plasma
was analyzed for Der f-specific immunoglobulins (IgGl and IgE). Lungs from nine mice in each
treatment group were homogenized, and Thl cytokine IL-2, Th2 cytokines IL-4 and IL-5,
granulocyte macrophage-colony-stimulating factor (GM-CSF), and the "chemokine regulated
upon activation, normal T-cell expressed and secreted" (RANTES) protein levels were quantified
in the supernatant via ELISA. Lungs from nine mice in each group were fixed, sectioned, and
stained to evaluate eosinophil infiltration, lymphocyte infiltration, goblet cell proliferation, and
localization of RANTES in the airway epithelium.
Der f-specific IgGl was present in blood plasma of sensitized mice but was unchanged
by formaldehyde exposure (Sadakane et al., 2002). IgE was too low to titer. IL-2 and GM-CSF
were undetected in lung homogenate supernatant, and 11-4 was unchanged by sensitization or
formaldehyde exposure. However, RANTES was increased by both formaldehyde exposure and
allergen sensitization and challenge (see Table 4-52). These increases were more pronounced
but less than additive for formaldehyde-exposed, allergen-sensitized mice. IL-5 was increased
by allergen but unaffected by formaldehyde exposure only. However, formaldehyde exposure
potentiated the IL-5 increase seen with allergen challenge.
Table 4-52. Cytokine and chemokine levels in lung tissue homogenate
supernatants in formaldehyde-exposed male ICR mice with and without
Der f sensitization
Group
1
2
o
J
4
Formaldehyde
-
+
-
+
Derf
-
-
+
+
GM-CSF
NDa
ND
ND
ND
IL-2
ND
ND
ND
ND
IL-4
68.1 ±.9
59.5 ±4.3
70.7 ±4.9
62.3 ±5.8
IL-5
4.4 ±0.3
4.1 ±0.2
13.6 ± 1.6 c'e
21.6±2.7c'e'f
RANTES
200.1 ±19.7
390.6 ±37.4b
479.6 ± 80.0C
593.3 ±58.2c'd
24
25
26
27
28
29
30
31
32
33
34
aNone detected.
bp < 0.05 from control.
°p< 0.001 from control.
dp < 0.05 from Group 2.
ep < 0.001 from Group 2.
fp < 0.001 from Group 3.
Source: Sadakane et al. (2002).
This document is a draft for review purposes only and does not constitute Agency policy.
4-325 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
10
11
12
13
14
15
16
17
18
19
20
21
Der f sensitization and challenge increased eosinophil infiltration into the interstitium
around the bronchi and bronchioles as well as goblet cell proliferation in the bronchial
epithelium (see Figure 4-21). Formaldehyde exposure exacerbated the eosinophilic and goblet
cell responses to a challenge dose of Der f (p < 0.05) (Sadakane et al., 2002). Formaldehyde-
induced eosinophilic infiltration in the absence of sensitization and challenge was not different
from nontreated, nonsensitized mice.
Panel A
1
o
LU
+++
«•
I
control
FA
Derf Derf + FA
Panel B
++++
S>
o
o
in
"o
o
IS
.a
§
1 •••
V
••
control
FA
Derf
Derf + FA
Figure 4-21. The effects of formaldehyde inhalation exposures on eosinophil
infiltration (Panel A) and goblet cell proliferation (Panel B) after Der f
challenge in the nasal mucosa of male ICR mice after sensitization and
challenge.
Note: ap < 0.001 compared with control group; bp < 0.001 compared with
formaldehyde group; cp < 0.05 compared with Der f group.
Source: Redrawn from Sadakane et al. (2002).
This document is a draft for review purposes only and does not constitute Agency policy.
4-326 DRAFT—DO NOT CITE OR QUOTE
-------
1
2 These results suggest that formaldehyde exposure may aggravate eosinophilic infiltration and
3 goblet cell proliferation that accompanies allergic responses. This response is associated with an
4 increase in IL-5, an eosinophilic attractant, and an increase in RANTES, which recruits
5 eosinophils by chemotaxis in formaldehyde-exposed and Der f challenged animals, although the
6 effect was not statistically significantly elevated compared with Der f challenge-induced levels
7 of IL-5 and RANTES alone.
8 Fujimaki et al. (2004a) investigated the long-term effects of low-dose formaldehyde
9 exposure on immunologic and neurological inflammation. Female C3FI/He mice were exposed
10 to 0, 0.082, 0.393, or 1.87 ppm (0, 0.1, 0.48, or 2.3 mg/m3) formaldehyde 16 hours/day,
11 5 days/week for 12 weeks. Six mice at each exposure level were given injections of OVA plus
12 adjuvant before the initial exposure and in weeks 3, 6, 9, and 11 of the experiment. Five mice at
13 each formaldehyde-exposure level did not receive OVA injections. One day after the last
14 exposure, mice were weighed and blood, BAL, spleen, and thymus were collected from each
15 animal. After weighing, spleens were disaggregated and spleen cells harvested for cell culture.
16 Immunophenotype of the spleen cells was determined by flow cytometry (CD4, CDS, CD3, and
17 CD 19 positive cells). Lymphocyte proliferation in response to lipopolysaccharide (LPS),
18 phytohemagglutinin A (PHA), or OVA was determined after 72 hours in culture. Splenocytes
19 were cultured for 48 hours in the presence of LPS, PHA, and OVA (immunized mice only), and
20 supernatants were collected for cytokine analysis (IL-4, IL-5, and IFN-y). Splenocytes were
21 cultured for 24 hours in the presence or absence of OVA to assess chemokine production (MCP-
22 1 and MlPl-a). Anti-OVA IgE, IgGi, IgG2, and IgGs were quantified in blood plasma.
23 Body and thymus weights were unchanged by formaldehyde exposure or OVA injection
24 (Fujimaki et al., 2004a), while, in nonimmunized mice, spleen weights were reduced by
25 formaldehyde exposure from 152 mg in controls to 128, 118, and 121 mg in mice exposed to
26 0.08, 0.4, and 1.8 ppm formaldehyde, respectively. Spleen weights tended to increase in groups
27 exposed to 400 and 2,000 ppb formaldehyde compared with controls in OVA-immunized mice
28 (control: 117.8 mg compared with 400 ppb: 168.6 mg and control: 121.0 mg compared with
29 2,000 ppb: 153.2 mg, respectively) but were not statistically significant.
30 To gain insight on the overall pulmonary inflammatory response of mice exposed to
31 formaldehyde in both immunized and nonimmunized mice, the total number and differential
32 count of MPs, neutrophils, lymphocytes, and eosinophils in BAL were counted and were found
33 to be unchanged by formaldehyde in nonimmunized mice. By contrast, in immunized mice
34 exposed to 1.8 ppm formaldehyde, the total number of BAL cells, MPs, and eosinophils were
35 significantly increased compared with nonimmunized controls (9.65 versus 2.84, 7.22 versus
36 2.74, and 2.0 versus 0.02 x 104 cells, respectively).
This document is a draft for review purposes only and does not constitute Agency policy.
4-327 DRAFT—DO NOT CITE OR QUOTE
-------
1 To further assess the pulmonary inflammatory response, protein levels of inflammatory
2 cytokines were determined by ELISA in BAL fluid. Levels of IL-lp in BAL of immunized mice
3 were decreased by formaldehyde exposure (p < 0.05 at 1.8 ppm formaldehyde), but IL-lp levels
4 after formaldehyde exposure were not different from controls in nonimmunized mice (Fujimaki
5 et al., 2004a). All other cytokines or chemokines were either unchanged (TNF-a, IL-6, and GM-
6 CSF) or not detected (eotaxin, MlP-la, and MCP-1).
7 Various neuropeptides, such as brain-derived neurotrophic factor (BDNF), nerve growth
8 factor (NGF), and substance P are released from vagal nerve endings and mediate a neurogenic
9 inflammatory response. Levels of BDNF, NGF, and substance P were assessed in BAL fluid
10 and/or in plasma. BDNF was not detected in BAL or in plasma. NGF levels in immunized mice
11 were significantly higher than in nonimmunized mice in both BAL fluid and in plasma. NGF
12 levels in immunized mice were significantly attenuated by 0.08 and 0.4 ppm formaldehyde
13 exposure (see Figure 4-22) in both BAL fluid and in plasma. Plasma level of substance P (a
14 mediator of neurogenic inflammation) was increased by formaldehyde exposures in
15 nonimmunized mice (see Figure 4-23) in both BAL fluid and plasma. This increase appears to
16 be dose-dependent and reaches statistical significance at 2,000 ppb formaldehyde exposure in
17 nonimmunized mice compared with nonimmunized controls. Similar to NGF, levels of
18 substance P increased in OVA-immunized mice compared with nonimmunized mice in both
19 BAL fluid and plasma. Similar to NGF, levels of substance P in OVA-immunized mice were
20 attenuated by formaldehyde exposure at 80 ppb.
21 Fujimaki et al. (2004a) further investigated the effect of low-level formaldehyde exposure
22 from both immunized and nonimmunized mice on the systemic immune response. Spleens were
23 removed from formaldehyde-exposed mice and were cultured in the presence of LPS or PHA
24 (for nonimmunized samples) or OVA (for immunized samples). The secretory ability of
25 immunized and nonimmunized spleen cells was assessed by measuring IFN-y release by ELISA.
26 Formaldehyde exposure (1.8 ppm) increased IFN-y fourfold in LPS-stimulated cultured spleen
27 cells from nonimmunized mice. No other cytokine or chemokine was changed by formaldehyde
28 exposure in cultured spleen cells from nonimmunized mice. In OVA-immunized mice,
29 formaldehyde had no significant effect on cytokines from stimulated spleen cells. OVA in vitro
30 stimulation significantly increased the chemokines MIP-1 and MCP-1 for control and
31 formaldehyde-treated OVA-immunized mice. The OVA-stimulated release of MCP-1 in vitro
32 was enhanced by formaldehyde exposure in a concentration-dependent manner, increasing
33 threefold and fourfold at 0.40 and 1.8 ppm, respectively. Increases in MCP-1 correlate with
34 reported increases in the associated cytokine, RANTES, which recruits eosinophils by
35 chemotaxis (Sadakane et al., 2002). These formaldehyde-induced increases in cytokine levels
This document is a draft for review purposes only and does not constitute Agency policy.
4-328 DRAFT—DO NOT CITE OR QUOTE
-------
1 contribute to pulmonary inflammation. The inflammatory response is not mediated by
2 lymphocytes, since lymphocyte subsets and in vitro cell proliferation were unchanged by OVA
3 immunization or formaldehyde treatment (Fujimaki et al., 2004a).
6
7
8
9
10
11
12
13
14
15
16
400 2000 (ppb)
Figure 4-22. NGF in BAL fluid from formaldehyde-exposed female C3H/He
mice with and without OVA sensitization.
Note: The day after the final formaldehyde inhalation, BAL fluid was collected
from formaldehyde-exposed, nonimmunized and formaldehyde-exposed, OVA-
immunized mice, and the production of NGF was determined by ELISA. Data
are mean ± SEM from five to six animals. **p < 0.01.
Source: Redrawn from Fujimaki et al. (2004a).
This document is a draft for review purposes only and does not constitute Agency policy.
4-329 DRAFT—DO NOT CITE OR QUOTE
-------
200
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
1
u
cj
d
2000
0 80 400 2000 (ppb)
Figure 4-23. Plasma substance P levels in formaldehyde-exposed female
C3H/He mice with and without OVA sensitization.
Note: The day after the final formaldehyde inhalation, plasma samples were
collected from formaldehyde-exposed, nonimmunized, and formaldehyde-
exposed OVA-immunized mice, and the levels of substance P were determined by
ELISA. Data are mean ± SEM from five to six animals. */><0.05.FA =
formaldehyde.
Source: Fujimaki et al. (2004a) as amended in Fujimaki et al. (2005; errata).
Anti-OVA (IgE and IgG2a) levels in plasma were unchanged by formaldehyde exposure.
Anti-OVA IgGi was reduced in immunized mice exposed to 400 ppb formaldehyde compared
with nonexposed animals. However, this effect did not persist as dose increased. Anti-OVA
IgGs was depressed in immunized mice exposed to 0.08 and 0.4 ppm formaldehyde (Fujimaki et
al., 2004a). Formaldehyde exposure did not induce an inflammatory response in lung or tracheal
epithelium in sections viewed by light microscopy (Fujimaki et al., 2004a). Although there was
a mild infiltration of mast cells into the epithelium of OVA-immunized mice, there were no
effects of formaldehyde treatment on mast cell infiltration.
A recent study by Lino dos Santos Franco et al. (2009) exposed male Wistar rats for
3 days, 90-minutes/day, to 1% formaldehyde (by weight; exact doses not reported) by inhalation.
Of these, one group was sensitized IP. to OVA (10 |j,g), a common allergen, immediately
following formaldehyde exposure, and subsequently challenged with OVA 2 weeks later. Other
rats were sensitized and challenged but were not exposed to formaldehyde. PCA reaction as well
as BAL analysis and whole blood analysis were conducted. Immunohistochemical analysis of
This document is a draft for review purposes only and does not constitute Agency policy.
4-330 DRAFT—DO NOT CITE OR QUOTE
-------
1 platelet endothelial cell adhesion molecule-1 (PECAM-1) expression, an inflammatory mediator,
2 in lung tissue was also measured. When formaldehyde exposure was followed by OVA
3 sensitization and challenge, decreased lung inflammation was reported compared with the group
4 that was OVA-sensitized but had not been exposed to formaldehyde. Reduced lung mast cell
5 degranulation was also reported in the formaldehyde/OVA group compared with the nonexposed
6 OVA group. Total circulating leukocytes, total bone marrow cells, and lung protein expression
7 levels of PEC AM-1 were also significantly decreased in formaldehyde/OVA rats compared with
8 non-formaldehyde exposed OVA rats. The reduction in inflammatory parameters in response to
9 formaldehyde may be attributed to different study designs, since in this study animals were
10 sensitized after exposure rather than prior to exposure. The results suggest that formaldehyde
11 may functionally alter the activity of certain cells, like mast cells, that may downgrade an
12 appropriate immune response to antigen and might serve to threaten lung homeostasis. Due to
13 the unique experimental design of this study, it cannot be directly compared with Sadakane et al.
14 (2002) or Fujimaki et al. (2004a). In addition, this study did not intend to measure whether
15 formaldehyde can exacerbate an asthmatic response but rather set out to identify whether
16 formaldehyde could affect immune homeostasis.
17 In summary, studies suggest that formaldehyde exposure may induce a predominantly
18 neurogenic inflammatory response via release of neuropeptide, such as NGF and substance P
19 from vagal nerve endings. Formaldehyde does not appear to potentiate a systemic immune
20 response. However, localized pulmonary inflammation can be potentiated by formaldehyde
21 exposure, as indicated by the increased presence of eosinophils and certain proinflammatory
22 cytokines (IFN-y). This response does not appear to be mediated by classic immunogenic
23 mechanisms since studies have failed to report elevated levels of anti-formaldehyde-specific IgE.
24 Several studies have shown that exposure to formaldehyde can facilitate allergic sensitization in
25 previously naive animals, and it is thought that this effect may occur due to formaldehyde's
26 ability to increase microvascular leakage in the nasal epithelium and by causing damage to the
27 nasal barrier (Ito et al., 1996). Sadakane et al. (2002) demonstrated that formaldehyde exposure
28 can also exacerbate allergic responses by enhancing the response to challenge allergen. Thus,
29 formaldehyde may exacerbate allergic responsiveness by aggravating the sensitization response
30 in previously naive animals by altering the permeability of the mucosal barrier in nasal
31 compartments. Neurogenically derived inflammation, including stimulation of the trigeminal
32 nerve and release of braykinin, suggests that the MOA for sensitization may ultimately have its
33 roots in neurogenic inflammation rather than an immunogenic response. In addition, using a
34 different protocol, Lino dos Santos Franco et al. (2009) suggest that formaldehyde exposure can
35 adversely affect lung homeostasis by reducing the activity of important inflammatory mediators
This document is a draft for review purposes only and does not constitute Agency policy.
4-331 DRAFT—DO NOT CITE OR QUOTE
-------
1 (mast cells, circulating leukocytes, PECAM-1 expression) when it occurs prior to sensitization,
2 thus downgrading an appropriate immune response.
O
4 4.2.5.2. Dermal Sensitization
5 Wahlberg (1993) used Hartley strain guinea pigs as test animals to determine the skin
6 irritancy of a suite of industrial chemicals, including formaldehyde. Aqueous solutions of the
7 compound in a 0.1 mL volume were applied to the shaved flanks of guinea pigs and gently
8 rubbed into the skin with a cotton-tipped applicator. Sites were left open and the treatments
9 repeated once daily for 10 days. A number of indices of acute skin irritation were monitored,
10 including erythema via visual scoring and edema and skin-fold thickness using Harpenden
11 calipers. Varying concentrations of formaldehyde (up to a 10% solution) induced a dose-
12 dependent increase in skin-fold thickness. Responses also showed shorter latencies at the higher
13 concentrations. For example, erythema was first observed on day 2 when 10% formaldehyde
14 was applied, day 5 (for 3%), and day 6 (for 1%).
15 Lee et al (1984) investigated the role of different routes of exposure in formaldehyde-
16 induced allergic sensitization. Two sets of four male English smooth-haired guinea pigs received
17 topical applications of 100 |jL 37% w/v formalin distributed over two shaved, depilated dorsal
18 sites two times over the course of 2 days at different sites. The total dose was calculated as
19 74 |j,g/animal. In addition, eight animals received a single topical application onto a 15 mm area
20 of the dorsal surface. The applied dose of 25 |jL formaldehyde was dissolved in saline. Two
21 other groups of guinea pigs were exposed to either 6 ppm (6 hours/day for 5 days) or 10 ppm
22 (6 hours/day for 5 days) formaldehyde by inhalation. A third group of guinea pigs was exposed
23 to 10 ppm formaldehyde for 8 hours/day for 5 consecutive days by inhalation. All animals were
24 evaluated for contact sensitivity by topical application of 20 mL formaldehyde diluted with
25 saline and distributed in a 15 mm area on the backs of the shaved guinea pigs (Lee et al., 1984).
26 Sites were visually inspected for erythema at 1, 6, 24, and 48 hours following the topical
27 application, and reactions were scored. No erythema was observed in control animals. None of
28 the guinea pigs in the 6 hours/day inhalation groups (6 and 10 ppm formaldehyde) developed
29 skin sensitivity tested on day 9 (4 days after the initial exposure regimen ended). Two of four
30 guinea pigs exposed to 10 ppm formaldehyde for 8 hours/day for 5 consecutive days developed
31 mild skin sensitization tested on day 31. Contact sensitivity increased in a dose-dependent
32 fashion in groups of animals that had been sensitized via the dermal route. Thus, dermal
33 exposure resulted in contact sensitivity. Inhalation exposure did not consistently produce contact
34 sensitivity.
This document is a draft for review purposes only and does not constitute Agency policy.
4-332 DRAFT—DO NOT CITE OR QUOTE
-------
1 Arts et al. (1997) used a local lymph node assay (LLNA) and the induction of IgE to
2 monitor the sensitization of female Wistar rats (low IgE-responders) and BN rats (high IgE
3 responders). For the LLNA assay, animals were sensitized by the application of varying
4 concentrations of formaldehyde in raffmated olive oil on the dorsum of both ears on days 0, 1,
5 and 2. Control animals were treated with raffmated olive oil alone. Animals received an IP.
6 injection of BrdU on day 5 and were subsequently sacrificed. Ear-draining lymph nodes were
7 collected, fixed, and sectioned, and the mitotic activity was monitored following successive
8 incubation of the sections in anti-BrdU, biotin-labeled rabbit anti-mouse antibody, peroxidase-
9 conjugated streptavidin, and 3,3-diaminobenzidine tetrahydrochloride. For serum IgE responses,
10 150 jiL of different concentrations of formaldehyde were applied to the shaved flanks of rats on
11 day 1, then 75 jiL of the same chemical at 50% of the initial concentration were applied to the
12 dorsum of each ear on day 7. The amount of IgE in the blood was measured using ELISA but
13 appeared to be little affected by formaldehyde treatment in either species of rat. However, the
14 ear-draining lymph nodes of both strains of rat showed a comparative increase in weight in
15 response to formaldehyde, and proliferation (BrdU positive) of paracortical cells was observed in
16 response to increasing doses of the compound. This response was most notable in BN rats
17 treated with 10% formaldehyde. Arts et al. (1997) concluded that the irritant and sensitizing
18 properties of formaldehyde may act through non-IgE-immune mechanisms.
19 Hilton et al. (1998) used the LLNA assay in female CBA/Ca (H-2k haplotype) mice to
20 compare the skin sensitizing potencies of formaldehyde and glutaraldehyde. The comparison
21 was set on a quantitative basis by determining the concentration of each compound necessary to
22 induce a threefold increase in lymph node cell proliferative activity (effective concentration
23 [ECs]). While both aldehydes induced a dose-dependent proliferative response, the
24 incorporation of [3H]-methylthymidine was far greater in animals exposed to glutaraldehyde
25 versus formaldehyde (with EC 3 values of 0.002-0.006 mol/L for glutaraldehyde versus
26 0.11-0.18 mol/L for formaldehyde). These data indicate the potential of both chemicals to
27 induce skin sensitization, although the potency of glutaraldehyde was far greater than that of
28 formaldehyde.
29 Xu et al. (2002) evaluated the extent to which the expression of some cytokines may
30 change as a result of cutaneous exposure to formaldehyde in mice. Female Balb/C mice were
31 skin painted with three topical applications of 100 jiL of 17.5% formaldehyde or distilled water
32 with a 1-day interval between each application. Spleen and draining lymph nodes were
33 harvested on days 3, 5, 7, 9, or 12 after the last skin painting. In some animals, contact
34 hypersensitivity was induced by applying 2% formaldehyde to both sides of mouse ears on day 3
35 following the last skin painting. For this endpoint, the percent increase in thickness of the ears
This document is a draft for review purposes only and does not constitute Agency policy.
4-333 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
was monitored. For the cytokines, mRNA expression levels of IL-2, IL-4, IL-5, IL-10, IL-12,
IL-13, IL-15, IL-18, and INF-y were determined semi quantitatively by measuring the amount of
individual mRNAs following amplification with the reverse transcriptase (RT)-PCR. The
relative amounts of cytokine mRNAs were calculated as the ratio of cytokine mRNA to that of
glyceraldehyde-3-phosphate dehydrogenase, as revealed in specific bands on an agarose gel.
Cutaneous formaldehyde treatment was associated with the long-lasting expression of
IL-4 and IFN-y mRNAs in mouse spleen and draining lymph nodes and with IL-15 mRNA only
in mouse spleen. Only IL-13 mRNAs displayed a transient increase in expression in both spleen
and draining lymph nodes. Levels of IL-2, IL-12, and IL-15 were increased in the mouse spleen
but not the lymph nodes. The mouse ear swelling test gave positive correlations with enhanced
expression of mRNA for IL-4 and IFN-y (see Table 4-53).
Table 4-53. Correlation coefficients among ear swelling responses and skin
mRNA levels in contact hypersensitivity to formaldehyde in mice
Variables
Ear swelling
IL-2
IL-4
Correlation coefficients
IL-2
0.50
-
-
IL-4
0.74a
0.39
-
IFN-Y
0.67a
0.60
0.79a
"Statistically significant (p < 0.05).
Source: Xuetal. (2002).
4.2.5.3. Summary of Sensitization Studies
Several animal studies report increased airway resistance and BC due to inhalation
exposures to formaldehyde (Nielsen et al., 1999; Swiecichowski et al., 1993; Biagini et al., 1989;
Amdur, 1960). Changes in pulmonary resistance were observed as early as 10 minutes after
exposure (Biagini et al., 1989), and reported effect levels ranged from 0.3-13 ppm. Other
pulmonary effects were reported in conjunction with BHR, such as increased tracheal reactivity
and decreased pulmonary elasticity (Swiecichowski et al., 1993; Amdur, 1960). Although BHR
is a common result of Type I hypersensitivity reaction to an allergen, the observation of BHR
alone is not sufficient to demonstrate that an agent induces Type 1 hypersensitivity.
BHR may be directly induced both pharmacologically and neurogenically (Joos, 2003;
Cain, 2001; Meggs, 1995). There is little evidence that formaldehyde itself is an allergen
recognized by the immune system, especially via inhalation (Lee et al., 1984). Although
This document is a draft for review purposes only and does not constitute Agency policy.
4-334 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde exposure has been reported to alter cytokine levels and immunoglobulins in some
2 experimental systems, these immunomodulatory effects do not support a type 1 hypersensitivity.
3 IgE was unchanged (Fujimaki et al., 2004a; Lee et al., 1984), and cytokine profiles were not
4 consistent with the Th-2 cytokines expected in IgE mediated hypersensitivity (Fujimaki et al.,
5 2004a; Ohtsuka et al., 2003).
6 Formaldehyde-induced dermal sensitization show parallel results. The physical signs of
7 irritation and sensitization are consistently shown (e.g., rashes, edema). Some involvement of
8 the immune response has been demonstrated with positive LLNA assays, indicating proliferation
9 of lymphocytes in lymph nodes draining the affected area (Hilton et al., 1998; Arts et al., 1997).
10 Increased expression of Th-2 cytokines in the lymph nodes of mice given dermal applications of
11 formaldehyde does indicate an immune component to the observed sensitization. However, the
12 response does not seem to be mediated by IgE (Arts et al., 1997; Lee et al., 1984).
13 Ito et al. (1996) reported that a tachykinin NKi receptor, but not the histamine HI or
14 bradykinin B2 receptors, is involved in formaldehyde-induced vascular permeability.
15 Neuropeptides NGF and substance P were affected in BAL and stimulated splenocytes from
16 formaldehyde-exposed mice, with greater effects seen in OVA-immunized mice. Tachykinins
17 (e.g., substance P and neurokinin A) are produced by nerve cells and can directly stimulate
18 bronchoconstriction (Van Schoor et al., 2000). Substance P is also a mediator of neurogenic
19 inflammation. Therefore, although formaldehyde may induce some of the symptoms of type 1
20 hypersensitivity, these symptoms are more likely neurogenic than immunogenic in origin.
21 In contrast, formaldehyde enhances immunogenic hypersensitivity of known allergens
22 (Sadakane et al., 2002; Riedel et al., 1996; Tarkowski and Gorski, 1995). This potentiation
23 varied based on sensitization protocols (respiratory tract versus systemic, frequency and timing
24 of immunization, allergen, etc.) and formaldehyde exposure regimens (concentration, continuous
25 versus intermittent exposures). Taken as a whole, the results support the finding that
26 formaldehyde exposure can aggravate a type 1 hypersensitivity response (see Table 4-54).
27
This document is a draft for review purposes only and does not constitute Agency policy.
4-335 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-54: Summary of sensitization and atopy studies by inhalation or dermal sensitization due to
formaldehyde in experimental animals
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
3 ,
Species
No./
group
Treatment"
Observations
LOAEL/
NOAEL
Reference
Inhalation studies
Cynomolgus
monkeys
Hartley
guinea pigs
(male)
Hartley
guinea pigs
(male)
F344 rats and
BN rats
English
smooth-haired
guinea pigs
Perlbright-
white,
Duncan-
Hartley
guinea pigs
(female)
Balb/C mice
(female)
9
8
5-7
5
4
12
4
Methacholine-sensitive monkeys
exposed to 2.5 ppm formaldehyde
for 10 minutes
0.86, 3.4, 9.4, 3 1.1 ppm
formaldehyde for 2 hours or 0. 1 1,
0.31, 0.59, 1.05 ppm for 8 hours
3. 4 ppm for 8 hours
16 ppm 3 hours/day, 5 days
6, 10 ppm, 6 hours/day, 5 days,
combined with provocation
challenge (2 or 4 ppm on day 7, or
days 7, 22, and 29)
0, 0.13, 0.25 ppm 8 hours/day,
5 days. The animals were
sensitized to OVA (3 minutes
exposure to 5% OVA aerosol)
0, 6.63 ppm 6 hours/day for
10 days or 6 hours/day once/week
for 7 weeks. All mice were
sensitized to OVA
Formaldehyde increased pulmonary resistance after 2, 5, and
10 minutes.
Total pulmonary resistance increased after 2 hours exposure at
9.4 and 31.1 ppm. Effect was reversible and returned to baseline
within 30 minutes. Total pulmonary resistance was increased
after 8 hours exposure at 0.3 and 1 ppm. Amount of
acetylcholine needed to achieve doubled pulmonary resistance
was decreased in animals after 2 hours exposure.
No changes in ex vivo trachea! constriction or inflammation.
Modest changes in inflammatory cytokine expression, but
respiratory and olfactory epithelial lesions were more severe in
F344 rats than in BN rats.
Inhalation challenge with 6 or 10 ppm followed by bronchial
challenge failed to increase respiratory sensitivity
Anti-OVA liter was significantly elevated over controls in
animals exposed to 0.25 ppm formaldehyde and showed that
formaldehyde may sensitize previously naive animals to OVA.
Formaldehyde administered intranasally for 6 hours/day for
10 days may facilitate sensitization to allergens since anti-OVA
liters were elevated over control animals. However, Ihe length
and duration of exposure appears to affecl developmenl of
sensitization.
LOAEL
2.5 ppm
NA
NA
NA
NA
NA
NA
Biagini el al.
(1989)
Swiecichowski
el al. (1993)
Swiecichowski
el al. (1993)
Ohlsuka el al.
(2003)
Lee el al. (1984)
Riedel el al.
(1996)
Tarkowski and
Gorski (1995)
J^.
oo
fe
H
O
o
-------
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
3 ,
Table 4-54: Summary of sensitization and atopy studies by inhalation or dermal sensitization due to
formaldehyde in experimental animals (continued)
Species
Wistar rats
(male)
Outbred ICR
mice (male)
C3H/HeJ
mice (female)
Wistar rats
(male)
No./
group
5-8
18
6
NA
Treatment"
0, 2, 5, 15, 45 ppm for 10 minutes
0.5% formaldehyde for 15
minutes, once/week for 4 weeks.
Both control and exposed groups
were exposed to Der f by IP.
injection 1 day before
formaldehyde and then challenged
with Der f after 4 weeks.
0,0.082,0.393, 1.87 ppm
16 hours/day, 5 day /week,
12 weeks. Mice also given OVA
plus adjuvant before exposure,
and again 3, 6, 9, 11 weeks after
exposure. Some formaldehyde
mice did not receive any OVA
1% Formaldehyde by weight for
90 minutes for 3 days. One group
was sensitized to OVA after to
formaldehyde exposure and then
challenged with OVA afterwards.
Others were sensitized and
challenged but not exposed to
formaldehyde.
Observations
Pulmonary insufflation or blood pressure were not altered.
Vascular permeability increased in concentration-dependent
manner and could be reduced by adding a NK1 selective
antagonist.
More pronounced PxANTES production in formaldehyde-treated
and sensitized rats than in sensitized rats that had not been
exposed to formaldehyde. Formaldehyde also potentiated IL-5
production associated with sensitization.
Substance P and NGF were increased dose dependency in
formaldehyde-treated, nonimmunized mice but were attenuated in
formaldehyde-treated immunized mice compared with
nonexposed, immunized controls.
Total circulating leukocytes, bone marrow cells, and lung protein
PECAM expression were significantly decreased in
formaldehyde/OVA rats compared with OVA rats.
LOAEL/
NOAEL
NA
NA
LOAEL
0.082 ppm
NA
Reference
Itoetal. (1996)
Sadakane et al.
(2002)
Fujimaki et al.
(2004a)
Lino dos Santos
Franco et al.
(2009)
J^.
oo
fe
H
o
o
-------
Table 4-54: Summary of sensitization and atopy studies by inhalation or dermal sensitization due to
formaldehyde in experimental animals (continued)
Species
No./
group
Treatment"
Observations
LOAEL/
NOAEL
Reference
Dermal sensitization
Hartley
guinea pigs
English
smooth-haired
guinea pigs
Wistar and
BN rats
(female)
CBA/Ca mice
Balb/c mice
(female)
5
4
4
NA
3-5
Skin painted once/day for 10 days
withO.lmLof 1, 3, or 10%
formaldehyde
Group 1: skin painted, 100 uL
37% formalin twice over 1 days,
Group 2: single topical
application of 25 uL
formaldehyde
Group 3 : 10 ppm 6 hours/day for
5 days by inhalation
Application of formaldehyde to
ears on days 0, 1,2, followed by
an IP. injection of BrdU.
Compared glutaraldehyde to
formaldehyde to induce a local
lymph node assay
Skin painted with 100 uL of
17.5% formaldehyde every other
day for days 3, 5, 7, 9, 12
Varying concentrations (up to 10%) induced dose -dependent
increase in skin-fold thickness. Erythema seen earlier at higher
doses (2 days at 10% formaldehyde vs 5 days at 3% or 6 days at
1%).
Two of four guinea pigs from group 3 had mild skin sensitization
after day 31. Contact sensitivity developed in a dose-dependent
manner in the dermal groups (group 1 and 2).
Ear-draining lymph nodes increased in weight in response to
formaldehyde, reflected in increased number of BrdU-stained
cells, most notably in BN rats (high IgE responders) treated with
10% formaldehyde
Glutaraldehyde and formaldehyde induced a dose-dependent
proliferative response that was greater in glutaraldehyde-treated
animals
Cutaneous treatment associated with long-lasting expression of
various cytokines from draining lymph nodes and spleen.
NA
Wahlberg et al.
(1993)
Lee etal. (1984)
Arts et al.
(1997)
Hilton et al.
(1998)
Xu et al. (2002)
I-
3 ,
oo
OJ
oo
NA = not applicable.
fe
H
o
o
-------
1 4.2.6. Neurological and Neurobehavioral Function
2 4.2.6.1. Inhalation Exposure
3 There are a number of published reports examining the effects of formaldehyde exposure
4 on nervous system structure and function. The reports evaluating behavioral effects fall into
5 three main categories: (1) behavioral responses evaluated during or immediately following
6 formaldehyde exposures, which may include effects due to the potential irritant properties of the
7 chemical, (2) acute or short-term exposures followed by behavioral assessments conducted 2-24
8 hours after termination of formaldehyde exposure, which reflect sustained effects of chemical
9 exposure independent of its irritant properties, and (3) repeated exposures to formaldehyde
10 followed by neurological assessments performed throughout the treatment period or several days
11 to weeks after termination of treatment. In addition to reports evaluating changes in behavior,
12 there are several reports evaluating neuropathological effects or changes in brain chemistry.
13
14 4.2.6.1.1. Behavioral response.
15 4.2.6.1.1.1. Clinical signs.
16 Several studies that were focused on general toxicity or carcinogenicity of formaldehyde
17 also assessed clinical signs in exposed animals, which may be related to adverse effects on the
18 nervous system. Procedural details for the assessments, or specific data regarding findings, were
19 not provided. Signs recorded included uncoordinated locomotion and climbing of cage walls at
20 20 ppm formaldehyde in rats (Woutersen et al., 1987); restlessness at 15 ppm formaldehyde in
21 rats (Morgan et al., 1986a); dyspnea, listlessness, and hunched posture at 20 ppm and ataxia at
22 40 ppm in mice (Maronpot et al., 1986); and dyspnea in rats at 14.3 ppm formaldehyde (Kerns et
23 al., 1983). Given the lack of information regarding procedures used for these evaluations and the
24 limited reporting of results, the utility of these data is limited.
25
26 4.2.6.1.1.2. Irritant threshold detection.
27 Wood and Coleman (1995) evaluated the irritant properties of acute formaldehyde
28 exposure in mice. Adult male Swiss mice (eight/group) were initially trained to terminate a
29 60-second exposure to an irritant gas (ammonia, 1,000 ppm) by poking their noses into a conical
30 sensor five times to produce a 60-second facial shower of clean air. Each test session consisted
31 of 25 exposure trials. Following training, response to formaldehyde was evaluated, using the
32 same testing scenario. Each day mice had a morning exposure session to ammonia and an
33 afternoon session to formaldehyde. Formaldehyde concentrations tested were different each day,
34 in sequence from 0, 1, 1.8, 3, 5.6, and 10 ppm (0, 1.23, 2.21, 3.68, 6.87, and 12.3 mg/m3) and
35 then stepping back again from 10 to 0 ppm. Half of the animals were tested in an ascending
This document is a draft for review purposes only and does not constitute Agency policy.
4-339 DRAFT—DO NOT CITE OR QUOTE
-------
1 order of formaldehyde concentrations, the other half in a descending order. The frequency of
2 terminating exposure, error rate, and the time lapse to termination were recorded. The
3 concentration at which 50% of the formaldehyde deliveries would be expected to be terminated
4 was estimated (ACso) by simple linear regression or by analysis of covariance on the logit
5 transform of percentage terminated as a function of log concentration.
6 All mice were trained successfully to terminate 100% of ammonia exposures, but varied
7 responses were observed with formaldehyde exposure. In general, time taken to terminate
8 formaldehyde exposure decreased significantly with increasing formaldehyde concentration.
9 Mice terminated more exposures to 1 ppm formaldehyde than to air alone (p < 0.0005), and the
10 error rate, generally below 40%, did not significantly differ with formaldehyde concentration
11 tested. Each animal had two test sessions with each formaldehyde concentration (once during
12 the ascending sequence and again during the descending sequence); both the time to termination
13 (p < 0.0012) and ACso were decreased in the second series of tests. One method of estimation by
14 the authors yielded an ACso of 3.63 ppm for the first series of tests versus 1.88 ppm for the
15 second series. A two-way repeated measures ANOVA with replication and concentration as
16 within variables was highly significant (p < 0.00005). These studies indicate mice are sensitive
17 to the irritant properties of formaldehyde at exposure concentrations as low as 1 ppm, and
18 animals reacted more swiftly and with greater accuracy to terminate formaldehyde exposure as
19 the concentration increased. However, a wide variety of responses was noted on an individual
20 animal basis. Two of the eight mice terminated 90% of the trials during 1 ppm exposures and
21 80-100% of trials at all other tested formaldehyde concentrations. One mouse terminated fewer
22 than 10% of the formaldehyde exposure trials (1-10 ppm) during the testing regimen but had a
23 92% response rate to 20 ppm formaldehyde. The remaining five mice responded with increasing
24 termination frequency as formaldehyde concentration increased from 1 to 10 ppm, with an ACso
25 of 2.72 ppm (3.34 mg/m3).
26
27 4.2.6.1.2. Motor activity and habituation.
28 Malek et al. (2003a) examined open field behavior of rats after acute formaldehyde
29 exposures. Male and female LEW. IK rats (15/sex/group) were exposed to 0, 1, 2.5, or 5 ppm (0,
30 1.23, 3.08, or 6.15 mg/m3) formaldehyde for 2 hours. Formaldehyde was vaporized from
31 aqueous solutions directly below the exposure chamber. Formaldehyde levels were checked 16
32 times throughout the 2-hour exposure periods. Mean formaldehyde levels of 1.01 ± 0.29 ppm,
33 2.51 ppm (SD is missing), and 5.0 ± 0.27 ppm were achieved. Locomotor activity was assessed
34 for 3 minutes in an open field 2 hours after termination of formaldehyde exposure and again 24
35 hours later, using an automated device to count the number of squares crossed. Other behaviors
This document is a draft for review purposes only and does not constitute Agency policy.
4-340 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
were noted, including grooming (face cleaning, fur licking, and scratching), rearing, sniffing (air
and floor), wall climbing, and defecation.
The authors reported no signs of irritation or changes in activity or food or water intake
during exposure. In general, sniffing was increased after formaldehyde exposure and movement
was decreased (crossed quadrants and climbing) in both male and female rats (p < 0.05).
Significant reductions in horizontal movements (crossed quadrants) were observed at all dose
levels and were characterized by a U-shaped dose response (see Figure 4-24). The lowest dose
tested (1 ppm) demonstrated a higher level of activity suppression than the two higher doses, but
all groups were still suppressed relative to controls. Although female rats displayed a greater
level of activity overall, a similar U-shaped dose-response pattern was also observed.
Activity as crossed quadrants
Number
100
80
60
40
20
0
D Males
• Females
Control 1 ppm 2.5 ppm 5.0 ppm
Formaldehyde
Figure 4-24. Motor activity in male and female rats 2 hours after exposure to
formaldehyde expressed as mean number of crossed quadrants ± SEM.
Greater reductions were observed in the lowest dose group, a pattern that was
evident in both genders. ** = different from control,p < 0.005.
Source: Drawn from data reported by Malek et al. (2003b).
This document is a draft for review purposes only and does not constitute Agency policy.
4-341 DRAFT—DO NOT CITE OR QUOTE
-------
1 Activity in the same apparatus was reassessed 24 hours later. As expected, controls
2 demonstrated habituation to the test apparatus, exhibiting only 20% of the motor activity
3 observed on day 1 (see Figure 4-25). In contrast, formaldehyde-treated animals failed to
4 demonstrate the same degree of habituation. Activity levels for males observed on day 2 were
5 60-80% of the activity levels seen on day 1 (p < 0.005). Formaldehyde-treated females also
6 failed to habituate and actually demonstrated increases in activity on day 2 relative to day 1 at all
7 formaldehyde exposure levels (p < 0.005).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
-------
1 exposure periods, and mean values were found to be 0.13 ± 0.04, 0.48 ± 0.05, and
2 5.18 ± 0.66 ppm. Open field behavior was evaluated for each animal 2 hours after formaldehyde
3 exposure. The number of crossed quadrants for both controls and a 5 ppm group were generally
4 comparable with those observed in the first study, although female values were somewhat lower.
5 Horizontal movement was decreased by formaldehyde exposure in a dose-dependent manner
6 with significant reductions in motor activity as low as 0.1 ppm in males and 0.5 ppm in females
7 (see Figure 4-26). The consistency of the findings across studies and between genders provides
8 greater confidence in the effects of low-level formaldehyde exposure on this standard test of
9 neurotoxicity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
I
1_
TJ
ra
3
TJ
0)
(A
8
u
0)
^2
E
3
40
10 —
0
male
T
J_^
X
T
J_ T female
***
X
pc
**
JL
JL,
0.1 0.5 5.0 0 0.1
formaldehyde (ppm)
0.5 5.0
Figure 4-26. Motor activity was reduced in male and female LEW.1K rats
2 hours after termination of 10-minute formaldehyde exposure.
Note: Values are means ± SDs. * = different from control,/* < 0.05. ** = different
from controls,/* <0.01. *** = different from controls,/* < 0.001.
Source: Drawn from data reported in Malek et al. (2003c).
Malek et al. (2004) also assessed the capacity of formaldehyde to induce persistent
behavioral deficits in mice. Groups of 20 male AB mice received a single 2-hour exposure to 0,
1.1, 2.3, or 5.2 ppm (0, 1.3, 2.8, or 6.4 mg/m3) formaldehyde prior to being tested 2 and 24 hours
after exposure for a series of behavioral responses, including ambulation (crossed squares),
grooming, sniffing, rearing, wall climbing, and defecation. Even though there were no clinical
signs of toxicity in any of the exposed groups, a number of behavioral anomalies were apparent
This document is a draft for review purposes only and does not constitute Agency policy.
4-343 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
in response to formaldehyde exposure, some of which persisted for at least 24 hours, as indicated
in Tables 4-55 and 4-56.
Table 4-55 Fluctuation of behavioral responses when male AB mice inhaled
formaldehyde in a single 2-hour exposure: effects after 2 hours
Open field parameter
No. of crossed inner squares
No. of crossed peripheral squares
Total no. of crossed squares
Air sniffing
Floor sniffing
Grooming
Rearing
Wall climbing
No. of excreted fecal boli
Formaldehyde concentration (ppm)a
0
34.10 ±7.51
56.65 ± 9.68
90.75 ±11.08
19.35 ±2.5
20.95 ±3.72
7.95 ±2.26
17.85 ±2.56
13.20 ±3.09
0.65 ±0.81
1.1
25.30 ±5.03b
59.55 ±9.75
84.85 ± 9.96
21.50 ±4.26
26.50 ± 4.64b
7.10±3.19
13.90 ± 3. 19b
14.55 ±2.74
0.75 ±0.85
2.3
21.20 ±3.41b
49.70 ± 13.24
71.10± 13.91b
16.35 ±3.84C
21.35 ±4.77
7.05 ± 2.48
11.30 ±2.30b
13.95 ±2.31
0.80 ±0.77
5.2
16.10 ±5.37b
29.15 ±7.47b
44.20 ± 7.42b
8.10±1.77b
22.80 ± 4.02
6.55 ±2.06
9.95±1.61b
13.95 ±1.82
0.90 ±1.12
7
8
9
10
11
12
13
14
15
aValues are means ± SDs.
bStatistical significance of differences from controls (p < 0.005).
Statistical significance of differences from controls (p < 0.05).
Source: Malek et al. (2004).
Table 4-56. Fluctuation of behavioral responses when male AB mice inhaled
formaldehyde in a single 2-hour exposure: effects after 24 hours
Open field parameter
No. of crossed inner squares
No. of crossed peripheral squares
Total no. of crossed squares
Air sniffing
Floor sniffing
Grooming
Rearing
Wall climbing
No. of excreted fecal boli
Formaldehyde concentration (ppm)a
0
10.40 ±2.35
42.80 ± 9.27
53.20 ± 8.67
13.65 ±2.81
21. 55 ±3.47
8.35 ±2.56
18.30 ±4.23
9.25 ±2.38
0.80 ±0.83
1.1
9.55 ±1.73
44.85 ± 14.60
54.40 ±14.77
13.30 ±3.21
15.85 ±3.94b
13.95 ±2.21b
12.40 ±2.23b
8.70 ±1.98
1.20 ±0.83
2.3
9.10 ±1.25
44.95 ±16.56
54.05 ±15. 81
12.65 ± 2.70
13.25 ±4.17b
10.20 ±3.33C
12.25 ±2.17b
8.20 ±2. 14
1.60±0.94C
5.2
9.70 ±1.13
41. 10 ±9.08
50.80 ±9.15
12.30 ±4.14
17.65 ± 3. 13b
11.90±3.26b
12.00 ±3.32b
9.90 ±2.27
1.20 ±0.89
16
17
18
19
20
aValues are means ± SDs.
bStatistical significance of differences from controls (p < 0.005).
Statistical significance of differences from controls (p < 0.05).
Source: Malek et al. (2004).
This document is a draft for review purposes only and does not constitute Agency policy.
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
Usanmaz et al. (2002) assessed spontaneous locomotor activity (SLMA) in Balb/c mice
(4-14 per group, sex unspecified) after both acute and subchronic formaldehyde exposures.
Prior to the acute exposure, mice were acclimated to the exposure chamber for 4 days but
exposed only to clean air. On the fifth day, mice (six/group, sex unspecified) were exposed for
3 hours at 0, 1.8, 3.2, 4.5, 6.4, 9.7, or 14.8 ppm (0, 2.2, 3.9, 5.5, 7.9, 11.9, or 18.2 mg/m3)
formaldehyde. Mice were removed from the exposure chamber, and SLMA behavior was
evaluated by direct observation for 5 minutes. In addition to horizontal and vertical movement,
wet dog shake (WDS) behavior was noted. In a separate trial, Balb/c mice (six/group, sex
unspecified) were exposed to 8.2 ppm formaldehyde for 1 week, 2 ppm formaldehyde for
2 weeks, or 3.3 ppm formaldehyde for 3 weeks (3 hours/day, 5 days/week) compared with
controls exposed only to air. SLMA behavior was observed for 5 minutes after the last exposure.
Mice exposed to 8.2 ppm formaldehyde for 1 week, 3.3 ppm formaldehyde for 2 weeks, and
2 ppm formaldehyde for 3 weeks lost weight over the course of the treatment (p < 0.05). All
other treatment groups had weight gain similar to control mice.
As shown in Figure 4-27, acute 3-hour formaldehyde exposures resulted in a dose-
dependent decrease in SLMA. Decreases in horizontal activity were significant for the three
highest dose groups (6.4, 9.7, and 14.8 ppm), and decreases in vertical activity were significant
for all six formaldehyde treatment groups. SLMA was similarly decreased following subchronic
exposures (data not shown here). Although the experimental protocol included longer exposures
and a slightly longer observation period (5 versus 3 minutes) than in Malek et al. (2003a, b), the
results are consistent, indicating decreased activity in formaldehyde-exposed animals several
hours after exposure was ended.
SLMA after acute exposures
D Control
D FA 1.8 ppm
B FA 3.2 ppm
CD FA 4.5 ppm
Q FA 6.4 ppm
D FA 9.7 ppm
• FA 14.8 ppm
Ambulatory Activity
Vertical Activity
Figure 4-27. The effects of the acute formaldehyde exposures on the
ambulatory and vertical components of SLMA.
Note: FA = formaldehyde exposure concentration. ** =p < 0.01 from controls.
Source: Usanmaz et al. (2001).
This document is a draft for review purposes only and does not constitute Agency policy.
4-345 DRAFT—DO NOT CITE OR QUOTE
-------
1 Usanmaz et al. (2002) noted an increase in WDS, after the acute exposures, as a possible
2 preconvulsive effect. However, the mice were only observed for 5 minutes, and it is unclear how
3 the researchers distinguished between a WDS due to an irritating odor and a preconvulsive
4 movement. No other study has noted convulsive effects from formaldehyde exposure in any
5 species. A second set of trials was reported in the same paper that sought to evaluate
6 formaldehyde effects on CNS excitability. Balb/c mice (six/group, sex unspecified) were
7 exposed to 0, 1.8, 6.4, or 14.8 ppm (0, 2.2, 7.9, or 18.2 mg/m3) formaldehyde for 3 hours.
8 Subchronic exposures were at 2 ppm (2.5 mg/m3) formaldehyde for 3 weeks or 3.3 ppm
9 (4.1 mg/m3) formaldehyde for 2 weeks. Seizures were induced by IP. injection of
10 pentilenetetrazole (PTZ), and the incidence, severity, and course of induced seizures were
11 recorded. The PTZ injection induced seizures in 83, 88, and 91% of controls, with 16, 38, and
12 67% mortality in controls in the three trials. Mortality was highly variable in treatment groups as
13 well. The authors report that PTZ-induced seizures were decreased in incidence by acute
14 formaldehyde exposure in a dose-dependent fashion with only 33% of mice exposed to 14.8 ppm
15 formaldehyde experiencing seizures versus 91% in control mice (p < 0.05 at the highest dose
16 only). However, the methodology for observing and scoring seizures is unclear. Additionally,
17 there was high mortality and high variability of results for the three similarly treated control
18 groups. Therefore, it is difficult to assess data quality and interpret these findings.
19 Boja et al. (1985) exposed male Sprague-Dawley rats to air or to formaldehyde at 5, 10,
20 or 20 ppm for 3 hours on 2 consecutive days. On the second day, half the rats received the same
21 exposure as the previous day, while half the rats were switched (e.g., half those rats receiving air
22 the first day received formaldehyde the second day, and half those receiving formaldehyde the
23 first day received air the second day), for a total of four possible exposure combinations. During
24 the exposure period, activity levels were monitored by observation, once per minute for the first
25 hour and once every 5 minutes for the second hour. At the end of the second exposure session,
26 rats were sacrificed and brains removed for neurochemical analysis (see Section 4.2.6.1.5).
27 Behavioral results were described in detail only for control and 5 ppm groups. During
28 the first exposure session, activity levels of formaldehyde-exposed animals were significantly
29 decreased (approximately 50% of control levels). On the second day of exposure, those animals
30 previously exposed to formaldehyde exhibited partial recovery, those experiencing their first
31 formaldehyde exposure behaved similarly to those initially exposed on the first day, and those
32 animals exposed to formaldehyde for a second time had a greater decrease in activity than during
33 the first exposure (to approximately 30% of control levels). The authors stated that a similar
34 effect was seen in animals exposed at 10 ppm but that results at 20 ppm were not interpretable
This document is a draft for review purposes only and does not constitute Agency policy.
4-346 DRAFT—DO NOT CITE OR QUOTE
-------
1 (data were not presented). Overall, the decreased activity seen in this study is consistent with
2 effects seen by other authors.
3 Senichenkova (1991) exposed pregnant female rats to 0 or 0.5 mg/m3 (0 or 403 ppb)
4 formaldehyde on gestation days (GDs) 1-19 for 4 hours/day. Reproductive aspects of this study
5 will be discussed in the reproduction section; however, results from behavioral assessments
6 conducted on the neonates are discussed here. The author stated that maturation of motor
7 reflexes (assessed as surface righting and pendular reflex), open field behavior, and maze
8 learning ability were assessed. Detailed descriptions of procedures and results were not provided
9 for all assessments, but it was stated that motor reflex development did not differ in treated and
10 control animals. Open field motor activity assessments in 40-day-old (juvenile) offspring
11 revealed an increase in squares visited and an increased frequency of rearing on the second and
12 third days of testing, indicating a lack of habituation in the offspring of formaldehyde-treated
13 dams; similar levels of activity by both measures were found on the first test day. Counts of
14 defecation and urination were increased on all 3 days of testing. Increased exploratory behavior,
15 described as increased impulses, was also noted in a learning task (not otherwise described), but
16 the author stated that learning rate and ability of the formaldehyde-treated group was not
17 different from controls (no data were provided).
18 Mobility and neuromuscular excitability (not otherwise described) in offspring of female
19 white rats were also evaluated by Sheveleva (1971). Dams were exposed to 0.005 or
20 0.0005 mg/L (approximately 4,000 or 400 ppb, respectively) formaldehyde on GDs 1-19.
21 Spontaneous mobility (over 15 minutes) and neuromuscular excitability were evaluated in
22 offspring at 1 or 2 months of age (other results from this study are discussed under
23 developmental toxicity, above). At 1 month, spontaneous mobility was reduced at the low dose
24 in males (52% of control levels; p < 0.01) but not at the high dose, and at both doses in females
25 (to 64 and 56% of control levels at the mid dose and high dose, respectively; p< 0.02). At two
26 months, there was a dose-related increase in activity for both sexes, statistically significant (p <
27 0.001) in high-dose females only (391% of control levels).
28
29 4.2.6.1.3. Learning and memory.
30 The effects of repeated formaldehyde exposures on learning were investigated by Malek
31 et al. (2003c), using a labyrinth swim maze. In this task, animals are required to make a series of
32 consecutive right or left turns to gain access to an escape platform (Malek et al., 2003c). Adult
33 male and female LEW. IK rats (15/sex/group) were exposed to 0, 0.1, 0.5, or 5.4 ppm (0, 0.123,
34 0.615, or 6.64 mg/m3) formaldehyde 2 hours/day for 10 consecutive days. Formaldehyde levels
35 were checked eight times throughout the 2-hour exposure periods. Mean formaldehyde levels of
This document is a draft for review purposes only and does not constitute Agency policy.
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
0.1 ± 0.02, 0.5 ± 0.1, and 5.4 ± 0.65 ppm were achieved. Body weight was measured on days 1,
5, and 10 of the experiment. Two days prior to beginning the formaldehyde exposures, all
subjects were given an acclimation trial in which they were individually placed into the water-
filled basin at the start position and allowed to navigate to the escape platform with manual
assistance to learn the correct route. Thereafter, the water labyrinth test was run on each day of
formaldehyde treatment, 2 hours after completion of each daily exposure. Time taken to
complete the test and errors made were recorded for each rat (see Table 4-57). An error was
defined as swimming toward the start position or circling in the same position without moving
forward toward the escape platform. Rats were sacrificed at the end of the experiment, and
tissues were taken from the lung, heart, thymus, kidney, liver, pancreas, skeletal muscle, and
spleen. Tissues were fixed and prepared for histologic examination by light microscope. No
differences were noted in food consumption or body weight gain for either male or female rats
(Malek et al., 2003c). No treatment-related differences in organ pathology were reported (with
the possible exception of focal microatelectasis (lung collapse at the microscopic level) seen in
two to three animals in each formaldehyde-exposed group but not control animals).
Table 4-57. Effects of formaldehyde exposure on completion of the labyrinth
test by male and female LEW.1K rats
Male rats
Control
O.lppnf
0.5 ppm
5.4 ppm
Female rats
Control
0.1 ppm
0.5 ppm
5.4 ppm
Swimming time (sec)
Dayl
105
100
97
105
Day 6
12.2
12.9
16.7 c
25.7 c
Day 10
6.33
6.07
7.60b
10.9C
Swimming time (sec)
Dayl
103
96
97
98
Day 6
12.5
12.3
14.6 c
23.5 c
Day 10
6.47
7.53
7.60b
9.73C
Error rate (mean)
Dayl
7.4
7.7
7.6
7.7
Day 6
0.5
5.0 c
4.4 c
5.0 c
Day 10
0.0
3.2C
1.8°
2.8C
Error rate (mean)
Dayl
7.9
7.1
8.0
7.9
Day 6
0
5.2 c
4.6 c
5.2 c
Day 10
0.0
3.0C
2.2C
2.6C
20
21
22
23
24
25
26
27
aRats were exposed to formaldehyde for 2 hours/day, for 10 consecutive days.
bDifferent from control, p < 0.05.
"Different from control, p < 0.005.
Source: Malek et al. (2003c).
This document is a draft for review purposes only and does not constitute Agency policy.
4-348 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
A clear learning curve was evident in control animals, with rats completing the task in
less time and with fewer errors over days (see Table 4-57). Although the number of errors
decreased with increasing experience in all groups, error rates in formaldehyde-exposed rats at
all doses were consistently higher than those observed in controls, starting on day 3 (see
Figure 4-28). All control animals performed without errors by day 6, whereas all treated animals
were still making two to three errors on day 10, the final day of testing. Time required (latency)
to complete the maze was also reduced over days. Although this measure of performance was
not as sensitive as error rate, formaldehyde-induced deficits were still evident in the 0.5 and 5.4
ppm exposure groups of both sexes.
—•— control
—A— 0.1 ppm
— •» — 0.5 ppm
•5.4 ppm
0)
(ft
^_
o
1_
^_
LJJ
23456789 10
Day of labyrinth test
Figure 4-28. Effects of formaldehyde exposure on the error rate of female
LEW.1K rats performing the water labyrinth learning test
Source: Drawn from data reported in Malek et al. (2003a).
Impaired performance on formaldehyde-treated subjects cannot be attributed to
alterations in swimming ability, since latencies to complete the maze were identical for 0 and
0.1 ppm groups, yet acquisition of the task was still impaired in the 0.1 ppm group based on
number of errors committed (see Figure 4-28). This study reports an adverse effect level of
This document is a draft for review purposes only and does not constitute Agency policy.
4-349 DRAFT—DO NOT CITE OR QUOTE
-------
1 0.1 ppm for increased error rate in the labyrinth water test, and all dose groups were equally
2 impaired across a broad range of exposures, 0.1-5.4 ppm. An independent estimate of
3 swimming speed was not included, so motor competency could not be directly evaluated.
4 However, comparable latency scores and error rates at the beginning of testing across all groups
5 and latency scores that track together over days suggest that impaired swimming ability does
6 account for the observed differences in latency, which are most likely reflective of the increased
7 number of errors in treated animals (errors usually increase the distance traveled and thus time
8 required for completion of the trial).
9 Pitten et al. (2000) evaluated the effects of very brief formaldehyde exposures
10 (10 minutes) but prolonged duration (90 days) on previously learned performance in a land
11 version of the labyrinth maze. Adult male and female Wistar rats (13/group) were trained on the
12 task for 14 days, two trials/day. Animals were required to make a series of five left or right turns
13 from the entrance of the maze to retrieve a piece of cheese placed in the goal box at the opposite
14 end. Animals were guided by the experimenter through the maze during this acclimation phase
15 until all subjects were able to retrieve the food without aid. After animals were trained (but prior
16 to formaldehyde exposure), performance was assessed once daily for 11 days, and the latency to
17 complete the maze as well as the number of errors committed when traversing from the entrance
18 to the goal box was recorded. Animals were then assigned to one of three dose groups (five to
19 eight/sex/group) such that task performance was equivalent across groups prior to
20 commencement of formaldehyde exposures. Animals were exposed to 0 ppm, 2.6 ppm (0.25%
21 formaldehyde solution to yield 3.06 ± 0.77 mg/m3), or 4.6 ppm (0.70% formaldehyde solution to
22 yield 5.55 ± 1.27 mg/m3) formaldehyde 10 minutes/day, 7 days/week for 90 days. Animals were
23 assessed for performance in the maze every seventh day, at least 22 hours after the exposure on
24 the previous day. At the end of the 90-day exposure period, monitoring of maze performance
25 continued once every 10 days for an additional 30 days. All rats were sacrificed at the end of the
26 postexposure trials and tissue sections were prepared for histologic examination by light
27 microscopy, including liver, trachea, lung, kidney, heart, spleen, pancreas, testicle, and brain.
28 No treatment-related changes in food or water consumption weight gain or in histologic samples
29 obtained at the termination of the experiment were observed.
30 Pitten et al. (2000) reported that no gender differences existed as a function of
31 formaldehyde treatment; therefore, data were presented by combining sexes. Control rats
32 showed no change in error rate but a slight decrease in running time through the maze during the
33 course of the experiment. The formaldehyde-exposed groups began with a similar performance
34 level and error rate as controls, but their performance degraded over the course of formaldehyde
35 exposure. By the fourth week of exposure, increased numbers of errors were evident in both
This document is a draft for review purposes only and does not constitute Agency policy.
4-350 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposed groups relative to controls. This trend reached statistical significance at the 12-week
2 time point, with a greater than twofold increase in number of errors (p < 0.05). Formaldehyde-
3 treated rats also tended to have increased run times through the maze (p = 0.04), but no
4 difference was seen by formaldehyde concentration. By 4 weeks after termination of exposure,
5 no statistical differences among the three groups were evident, but the tendency for the two
6 exposed groups to make more errors and have longer latencies remained. Since Pitten et al.
7 (2000) tested animals after the task was acquired, these results indicate deficits in the retention of
8 a previously learned task.
9 Lu et al. (2008) evaluated the effects of formaldehyde on performance of mice in a
10 Morris water maze. Kunming mice (five males/group) were exposed to formaldehyde at 0.2, 1,
11 or 3 mg/m3 6 hours/day for 7 days (measured concentrations: 0.2^0.01, 0.99^0.04, and
12 3.03^0.16 mg/m3). Mice were trained to locate a hidden platform in a large, circular tank
13 (106 cm diameter, 31 cm deep). Each animal received four training trials per day, beginning
14 30 minutes after the end of exposure. During training, latency to locate the platform was
15 recorded for each trial, with a maximum of 60 seconds, after which the animal was guided to the
16 platform. After the last day of training, an additional trial was conducted with the platform
17 removed (the probe trial); time spent in each maze quadrant was measured to determine the time
18 the animal spent searching for the platform in the correct area of the maze. Performance in the
19 water maze, measured as mean escape latency across the seven training trials, was significantly
20 impaired in the 3 mg/m3 group. No significant difference was seen at 1 mg/m3, although there
21 appeared to be an increased latency during the second day of testing. During the probe trial,
22 control animals spent significantly more time in the correct quadrant, but neither formaldehyde-
23 exposed group did so. Results of this study indicate deficits in learning and retention of the
24 Morris water maze following formaldehyde exposure, with greater effects seen in the higher dose
25 group.
26 Apfelbach and Reibenspies (1991) published a brief report of formaldehyde effects on
27 olfactory learning. Ferrets were exposed to 0.25 ppm (0.31 mg/m3) formaldehyde gas
28 continuously for 6 months. A Y-shaped maze was used to test odor detection, discrimination
29 between odors, and odor threshold. Ferrets were conditioned to distinguish ethyl acetate (0.1 vol
30 %) from clean air. Untreated ferrets achieved 75% success after an average of 110 trials.
31 However, formaldehyde-treated ferrets required on average 320 trials to reach a 75% success
32 rate. A 90% success rate was achieved by untreated ferrets after 420 trials. However, this level
33 of success was not reached in formaldehyde-treated ferrets.
34 The same researchers also tested olfactory function in formaldehyde-treated ferrets, as
35 summarized in Section 4.2.7 (Apfelbach et al., 1992). A decrease in olfactory discrimination and
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 51 DRAFT—DO NOT CITE OR QUOTE
-------
1 a reduction in the percentage of olfactory cells in the olfactory epithelium were reported after
2 3-12 months exposure to 0.25 or 0.5 ppm formaldehyde. Decreased olfactory sensitivity in rats
3 exposed to 0.25 or 0.5 ppm formaldehyde has also been reported by the same researchers (Weiler
4 and Apfelbach, 1992; Apfelbach and Weiler, 1991), and Weiler and Apfelbach (1992) reported
5 in an abstract that shifts in olfactory thresholds were greater when exposure was initiated at PND
6 30 than at adult ages. Given the documented changes in olfactory thresholds, observed changes
7 in olfactory learning would likely be confounded by the potential for decreased olfactory
8 function by formaldehyde exposures, and definitive conclusions regarding formaldehyde effects
9 specific to learning cannot be made based on these studies.
10
11 4.2.6.1.4. Neurosensitization.
12 Sorg et al. (1996) studied the potential for formaldehyde exposure to induce sensitization
13 in the CNS, possibly through the limbic pathways in the brain. The authors hypothesized that
14 multiple chemical sensitivity (MCS) has an onset and progression similar to CNS sensitization
15 and may, therefore, be a similar process. These experiments were conducted to test this
16 hypothesis and to determine whether formaldehyde exposure could be used as a model for MCS.
17 Behavioral sensitization can be initiated by psychostimulants (e.g., cocaine) and manifest as
18 increased locomotor activity upon subsequent challenge with the stimulant.
19 Sorg et al. (1996) evaluated cross-sensitization of cocaine-induced increases in activity
20 from an initial formaldehyde exposure. Female Sprague-Dawley rats (eight to nine) were
21 exposed to 0 or 11 ppm (0 or 13.5 mg/m3) formaldehyde 1 hour/day for 7 days. Locomotor
22 activity was measured (by photocell) after saline injection (1 day postexposure) and after cocaine
23 injection (2 days postexposure). A similar protocol was conducted on days 36 and 37
24 postexposure. Motor activity levels following saline injection were similar for controls and
25 formaldehyde-treated rats. However, formaldehyde exposure initiated sensitization to cocaine as
26 evidenced by a greater increase in locomotor activity in mice treated with formaldehyde
27 followed by cocaine (p < 0.05) with an average count of crossed grids greater than 40,000 (2
28 hours) in treated animals compared with 25,000 (2 hours) in controls. The cross-sensitization
29 was transient, with no treatment effects on cocaine-induced activity either 29 or 37 days
30 postexposure. When examining individual data, the authors suggested that the formaldehyde-
31 treated groups in both cases have a cluster of high responders (HRs), suggesting some animals
32 may have been more sensitive. A second group of similarly treated female rats was pretested for
33 locomotor activity and divided into subgroups of HRs or low responders (LRs). They were then
34 given a panel of neurobehavioral tests: anxiety (elevated plus maze, day 11); memory (passive
35 avoidance training, day 12; passive avoidance test, day 19); and nociceptive test (day 20). Trunk
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 52 DRAFT—DO NOT CITE OR QUOTE
-------
1 blood corticosterone levels were determined during stress on day 35 postexposure. No
2 significant treatment differences were found in the passive avoidance test, nociception, or
3 corticosterone levels (basal or stress induced). On the elevated plus maze, a two-way ANOVA
4 indicated no overall formaldehyde treatment effects, but the HR rats had higher open arm time
5 ratios (indicating greater anxiety) regardless of treatment. Within the treatment groups, the
6 difference in behavior between HR and LR subgroups was only significant for the formaldehyde-
7 treated rats (p < 0.05). Authors suggested that cross-sensitization to cocaine-induced locomotor
8 activity was caused by enhanced dopamine transmission within the mesolimbic system (ventral
9 tegmental area to nucleus accumbens projection) following repeated formaldehyde exposure. A
10 critical role of the hypothalamic-pituitary-adrenal (HPA) axis has also been implicated in cross-
11 sensitization.
12 Sorg et al. (1998) and Sorg and Hochstatter (1999) further explored formaldehyde-
13 induced behavioral sensitization using the cocaine model. In contrast to the results following
14 exposure to 11 ppm (Sorg et al., 1996), rats exposed to only 1 ppm for 7 days showed no cross-
15 sensitization to cocaine injection. However, animals exposed to 1 ppm formaldehyde for
16 4 weeks exhibited increased cocaine-induced vertical activity (with no difference in horizontal
17 activity) for 4-6 weeks after cessation of exposure. Activity levels of formaldehyde-exposed
18 rats were approximately threefold those of control rats 3-4 days postexposure and still 1.5-fold
19 control levels at 4-6 weeks postexposure (p < 0.05).
20 Sorg et al. (2001) examined changes in corticosterone levels in rats with and without
21 formaldehyde treatment. Basal corticosterone levels in trunk blood were established in naive
22 male Sprague-Dawley rats taken directly from their home cage immediately prior to sacrifice. In
23 an acute trial, male rats were exposed to 0, 0.7, or 2.4 ppm (0, 0.86, or 2.96 mg/m3)
24 formaldehyde for either 20 or 60 minutes, and trunk blood was collected for corticosterone
25 analysis. Therefore, these rats were challenged with a new environment (the exposure chamber)
26 in the presence or absence of formaldehyde. In a separate trial, basal and challenged
27 corticosterone levels were measured after repeated exposure (1 hour/day, 5 days/week for 2 or
28 4 weeks). Basal corticosterone levels were measured in trunk blood immediately after removing
29 the animal from its home cage. Challenged corticosterone levels were measured after rats were
30 placed into the exposure chamber for a final 20-minute exposure. Body weight was measured at
31 the beginning of each week of exposure and was unchanged by formaldehyde treatment.
32 Corticosterone levels were increased over basal levels when rats were placed in the
33 exposure chamber for 20 minutes (see Figure 4-29, panel a) but returned to basal levels after
34 60 minutes in the exposure chamber (not shown). This response may reflect the stress of the new
This document is a draft for review purposes only and does not constitute Agency policy.
4-353 DRAFT—DO NOT CITE OR QUOTE
-------
4
5
6
7
8
9
10
11
12
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
environment and acclimatization after 60 minutes in the chamber. Corticosterone levels were the
same in the presence or absence of formaldehyde, indicating no treatment effect.
Panel a
Panel b
Panel c
Corticosterone levels: with no
pre-exposure to formaldehyde
Control 0 07 24
Formaldehyde (ppm)
Corticosterone levels after 2 weeks of
repeated formaldehyde exposure
Formaldehyde (ppm)
Corticosterone levels after 4 weeks
of repeated formaldehyde exposure
Control 0 07
Formaldehyde (ppm)
Figure 4-29. Basal and stress-induced trunk blood Corticosterone levels in
male LEW.IK rats after formaldehyde inhalation exposures.
Note: Panel a: no pretreatment, Corticosterone levels after 20-minute
formaldehyde exposure. Panels b and c show both basal and induced
Corticosterone levels after a 2- or 4-week pretreatment to formaldehyde
1 hour/day. Challenge to induce Corticosterone was a 20-minute reexposure at the
formaldehyde level tested.
Source: Sorgetal. (2001).
Control animals exhibited an increase in basal Corticosterone after 2 weeks, which
returned to naive levels after 4 weeks (see Figure 4-29, panels b and c). Formaldehyde-treated
rats demonstrated a comparable increase in basal Corticosterone levels at 2 weeks, but these
levels did not return to naive levels at 4 weeks as seen with controls. Control and 0.7 ppm
exposed rats showed a similar response to challenge (the final 20-minute exposure). However,
rats exposed to 2.4 ppm were hyperresponsive, with exaggerated Corticosterone levels during this
final exposure. Differences in basal Corticosterone levels after formaldehyde exposure and the
hyperresponsiveness seen in animals exposed at 2.4 ppm provide evidence of
formaldehyde-induced perturbations of the HPA axis. Authors suggested that elevated
Corticosterone levels induced by repeated formaldehyde exposures may contribute to the cross-
sensitization to cocaine-induced motor activity.
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 54 DRAFT—DO NOT CITE OR QUOTE
-------
1 Formaldehyde-induced changes in the HPA axis may contribute to behavioral effects of
2 formaldehyde exposure reported by Sorg et al. (2004) and Sorg and Hochstatter (1999). The
3 authors also reported an enhanced conditioning to odor in animals previously exposed to
4 repeated formaldehyde. Male and female Sprague-Dawley rats (60-80 days of age) were
5 exposed at 1 ppm (1.23 mg/m3) formaldehyde 1 hour/day, 5 days/week for 4 weeks (Sorg and
6 Hochstatter, 1999). Two weeks after exposure ended, rats were trained to the conditioned fear
7 task. Rats were conditioned to a fear response by either odor only or odor associated with
8 footpad shock. Orange-oil extract was used as the odor conditioned stimulus (CS). One day
9 after conditioning, rats were reintroduced into the environment without an odor cue, and time
10 spent motionless in the freezing posture (freezing) was observed. On day 2 after conditioning,
11 rats were placed in a novel environment, and time spent in the freezing posture was evaluated in
12 the absence and then the presence of odor. This was repeated on day 12 after conditioning to
13 measure the loss of the freezing response to the conditioned odor.
14 Both treated and exposed rats showed similar responses on reintroduction into the
15 conditioning environment in the absence of an odor cue on day 1 (Sorg and Hochstatter, 1999).
16 As expected, rats conditioned with a footpad shock demonstrated greater time motionless than
17 odor-trained only rats, and there was no difference between control and formaldehyde-treated
18 rats. However, in the presence of odor on days 2 and 12, formaldehyde-exposed rats who were
19 conditioned with odor associated with foot shock spent significantly more time freezing than
20 odor-only trained rats (p < 0.05); control animals on those days showed no difference in time
21 freezing in the presence and absence of odor. The authors concluded that the formaldehyde-
22 treated rats had more difficulty than controls in extinguishing the fear response to the
23 conditioned odor, and speculated that an enhancement of the fear-conditioned response by
24 formaldehyde pretreatment supports the hypothesis that sensitization may include effects through
25 the limbic system of the brain.
26 In a second experiment, adult male and female Sprague-Dawley rats were exposed at 0 or
27 2 ppm (2.45 mg/m3) formaldehyde 1 hour/day, 5 days/week for 4 weeks (Sorg et al., 2004). Two
28 to 3 weeks after exposure ended, rats were trained to the conditioned fear task. Rats were given
29 a foot shock either associated with an odor (paired group) or unassociated with an odor (unpaired
30 group). Orange-oil extract was used as the odor CS. After training, freezing behavior was
31 assessed (1) in the same context in the absence of odor (1 day), (2) in a new context in the
32 presence and absence of the CS (5 consecutive days), and (3) in another novel context in the
33 presence and absence of the CS.
34 Formaldehyde-exposed male rats demonstrated increased conditioned fear response to an
35 odor CS (orange oil) paired with foot shock with no change in the degree of conditioning to the
This document is a draft for review purposes only and does not constitute Agency policy.
4-355 DRAFT—DO NOT CITE OR QUOTE
-------
1 context. For female rats, formaldehyde exposure did not affect the percent of time spent
2 freezing, either in the conditioning context or the novel context in the absence of the conditioned
3 odor. In contrast, male rats spent an increased time freezing in a novel context in the presence of
4 odor, indicating a greater conditioned fear response to the olfactory cue (p < 0.05). This is in
5 agreement with the previous study where formaldehyde effects were seen in the presence of the
6 conditioning odor but not the environment (Sorg and Hochstatter, 1999). However, in this study
7 female rats did not exhibit a similar enhancement of fear conditioning to the olfactory CS.
8 The authors suggested that repeated exposure to low levels of formaldehyde acts as a
9 stressor in much the same way as inescapable foot shock, with resulting sensitized responses
10 within the olfactory/limbic pathways (Sorg et al., 2004). This interpretation is consistent with
11 work described above in which augmented basal corticosterone levels following repeated
12 formaldehyde exposures were demonstrated. However, while the fear conditioning in the present
13 study and cross-sensitization to cocaine described above (Sorg and Hochstatter, 1999) occurred
14 3-4 weeks after termination of exposure to formaldehyde, the duration of corticosterone
15 elevation induced by repeated exposure to formaldehyde has not been determined. It is possible
16 that augmentation of corticosterone levels following formaldehyde exposure results from direct
17 action of formaldehyde on the HPA axis. Experiments designed to compare HPA activation
18 following standard stressors (repeated inescapable foot shock or restraint stress), stress induced
19 by other irritants (chemicals with strong irritant odors but no CNS action), and repeated
20 formaldehyde exposures are necessary to dissociate primary from secondary action of
21 formaldehyde on CNS function in this paradigm. It is also possible that enhanced conditioning
22 to an odor stimulus results from formaldehyde-induced increases in airway irritation, rendering
23 the conditioned odor stimulus a more salient cue, producing a conditioned response that is not
24 extinguished as readily as in air-exposed controls. However, damage of the nasal mucosa and
25 lesions would be expected to be minimal at 1 ppm formaldehyde exposures and most likely
26 resolved 2 weeks after exposure was ended (see Section 4.2.1.2). Therefore, a more salient cue
27 for fear conditioning to odor due to physical irritation is not likely. Alternatively, formaldehyde
28 may act to up regulate olfactory activity, producing a stronger sense of odor during conditioning.
29
30 4.2.6.1.5. Neurochemistry and neuropathology.
31 Several studies that were focused on general toxicity or carcinogen!city of formaldehyde
32 also assessed histopathology in exposed animals, including pathological evaluation of the brain.
33 In all cases, details of the pathological evaluation were not provided. Reported results stated that
34 no significant lesions were seen on unspecified tissues (Appelman et al., 1988; Maronpot et al.,
35 1986; Kerns et al., 1983) or that an increase in relative brain weight (data not provided) was
This document is a draft for review purposes only and does not constitute Agency policy.
4-356 DRAFT—DO NOT CITE OR QUOTE
-------
1 considered of no toxicological significance (Woutersen et al., 1987). The absence of procedural
2 information, or specific reported results, limits the utility of this information.
3 Boja et al. (1985) measured changes in several neurotransmitters (norepinephrine,
4 dopamine, 5-hydroxytryptamine) and their major metabolites (3,4-dihydroxyphenylacetic acid
5 [DOPAC] and 5-hydroxyindoleacetic acid [5-HIAA]) following one or two 3-hour exposures to
6 formaldehyde at 0, 5, 10, or 20 ppm. Animals were sacrificed immediately following the second
7 exposure, and brains were immediately removed, frozen, and sectioned. Regions of interest were
8 analyzed by high-pressure liquid chromatography with electrochemical detection. Authors stated
9 that neurotransmitter concentrations were measured in multiple brain regions, but results were
10 reported only for the 5 ppm exposure and only for the hypothalamus. No change was seen in
11 concentrations of norepinephrine or 5-hydroxytryptamine for any exposure paradigm. For those
12 animals exposed twice to formaldehyde, there was a slight (statistically significant) increase in
13 dopamine and a larger (approximately fourfold) increase in 5-HIAA. DOPAC was increased
14 (approximately 30%) in animals receiving formaldehyde during the second exposure only.
15 Recent work by Hayashi et al. (2004) indicates that formaldehyde exposure increases the
16 activity of periglomerular (PG) cells in the main olfactory bulb. Tyrosine hydroxylase activity
17 was measured as a marker for activity of olfactory function. The authors surmised that
18 expression levels of this enzyme are useful markers since it has been reported that the protein is
19 up regulated after sensory stimulation and is down regulated by odor deprivation or when the
20 olfactory epithelium is removed (Cho et al., 1996; Stone et al., 1991; McLean and Shipley, 1988;
21 Baker et al., 1983). Eight-week-old female C3H/HeN mice were exposed at 0, 0.08, 0.4, or
22 2 ppm (0, 0.1, 0.49, or 2.45 mg/m3) formaldehyde 16 hours/day for 1 day or 12 weeks
23 (5 days/week). Formaldehyde exposure did not affect body weight. Mice were sacrificed
24 24 hours after exposure; the brains were removed, fixed, and prepared for sectioning. One side
25 of the olfactory bulb was sliced into 40 |im-thick serial frontal sections and immuno-stained for
26 tyrosine hydroxylase activity. The number of tyrosine hydroxylase-positive PG cells was
27 determined by examining digital photomicrographs of three tissue sections, averaging the counts
28 from 10-15 glomeruli per section.
29 Neither the size of the olfactory bulb (rostrocaudal, dorsoventral, and mediolateral
30 lengths) nor the total number of PG cells was changed by formaldehyde exposure. The number
31 of tyrosine hydroxylase-positive PG cells per glomerulus was unchanged by a single
32 formaldehyde exposure but increased after 12 weeks of repeated exposures. The increases were
33 similar among treatment groups: 5.54 ± 0.31 at 0.80 ppm, 5.18 ± 0.60 at 0.4 ppm, and 6.0 ± 0.83
34 at 2 ppm or 196, 167, and 196% of controls, respectively. As an indicator of activity, it is not
35 unexpected that the enzyme was up regulated after repeated exposure to an odorous compound.
This document is a draft for review purposes only and does not constitute Agency policy.
4-357 DRAFT—DO NOT CITE OR QUOTE
-------
1 Hayashi et al. (2004) hypothesize that the increased tyrosine hydroxylase activity is an indication
2 of increased sensitivity and, therefore, may be a model for MCS. However, it is unknown if the
3 increase in enzyme activity after repeated exposures is transient or could result in sensitization.
4 Tyrosine hydroxylase is the first enzyme in the dopamine synthetic pathway, but the role of
5 dopamine in PG cells is not known. Further research would be needed to understand the
6 potential for formaldehyde to act as a sensitizing agent in this model.
7 In an abstract, Kakeyama et al. (2004) outline the results of experiments to address the
8 effects of subchronic exposure to low levels of formaldehyde on changes in neurotransmitter-
9 related mRNA expressions in mice forebrains. An unstated number of female C3H/He mice
10 were exposed 16 hours/day, 5 days/week to 400 ppb (0.49 mg/m3) formaldehyde for 12 weeks.
11 The authors used RT-PCR methodologies to quantify mRNA encoding for the glutamate receptor
12 subunits GluRl and GluR2, the dopamine receptor Dl, and the serotonin receptor 5-HT1A in the
13 neocortex, hippocampus, amygdala, and hypothalamus. Raised levels of mRNA expression were
14 observed for GluRl in the neocortex and hippocampus; GluRl, GluR2, and the dopamine
15 receptor Dl in the amygdala; and the serotonin receptor 5-HT1A in the hypothalamus. Reduced
16 mRNA expression was observed for GluR2 in the hippocampus and neocortex. When other
17 mice were subjected to a radiofrequency-induced lesion of the hippocampus then exposed to
18 formaldehyde for 12 weeks as before, the altered expression of GluRl and GluR2 in the
19 neocortex was abolished. However, the increment of mRNA expression of 5-HT1A in the
20 hypothalamus was further enhanced. In demonstrating that formaldehyde affects neocortical
21 GluRl and GluR2 mRNA expressions through a hippocampal function, Kakeyama et al. (2004)
22 concluded that subchronic exposure to low concentrations of formaldehyde can affect neural
23 transmission in the forebrain.
24 Fujimaki et al. (2004b) examined the effects of formaldehyde on NGF in the brain and
25 hippocampus. Ten female C3H/HeN mice/group were exposed to 0, 80, 400, or 2,000 ppb (0,
26 0.1, 0.5, or 2.45 mg/m3) formaldehyde 16 hours/day, 5 days/week for 12 weeks. Some groups of
27 mice received the same treatment after IP. injection of 10 jig of OVA and 2 mg alum prior to the
28 commencement of formaldehyde exposure. For this subgroup, booster injections of OVA were
29 administered on days 21, 42, 63, and 77 during the formaldehyde exposure regimen.
30 Quantitative measures of NGF and BDNF in homogenates of whole brain and hippocampus were
31 obtained by ELISA and mRNA determination. The amount of NGF protein in whole brains
32 remained unchanged in the nonimmunized mice. However, brain NGF levels were significantly
33 increased in OVA-immunized mice exposed to 80 and 400 ppb (but not 2,000 ppb)
34 formaldehyde (see Figure 4-30). This result was confirmed by parallel increases in the
35 concentrations of hippocampal NGF mRNA that were produced in immunized mice exposed to
This document is a draft for review purposes only and does not constitute Agency policy.
4-358 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
formaldehyde at the same concentrations. However, there were no comparable increases in the
amounts of brain-derived neurotrophic factor in either immunized or nonimmunized mice. In
discussing the mechanisms potentially associated with their results, Fujimaki et al. (2004b)
considered it likely that low-level exposure to formaldehyde could enhance NGF production
through the stimulation of the FIPA axis together with immunization.
Q Non-immunized mice
• Ova-immunized mice
80 400 2,000
Formaldehyde (ppb)
Figure 4-30. NGF production in the brains of formaldehyde-exposed mice.
Note: Female C3H mice were exposed to formaldehyde 16 hours/day,
5 days/week for 12 weeks. NGF in homogenates of whole brain and
hippocampus were measured by ELISA. Values are means ± SEM (n = 5-6).
* =p < 0.05 and ** =p < 0.01 versus control mice, as calculated by the authors.
Source: Redrawn from Fujimaki et al. (2004b).
The enhancement of NGF in the brains of immunized mice exposed to formaldehyde
gave rise to the suggestion that NGF may promote the survival of hippocampal neurons when
challenged with formaldehyde. To examine whether or not apoptosis plays a role in this process,
Tsukahara et al. (2006) measured the effects of formaldehyde on apoptotic mechanisms
regulating the survival and death of cells and on 7V-methyl-D-aspartate (NMDA) receptors.
Female C3H/HeN mice (13/group) were exposed to 0 or 400 (393 ± 34) ppb (0 or 490.8 |ig/m3)
formaldehyde 16 hours/day, 5 days/week for 12 weeks. Seven control and formaldehyde-treated
mice were immunized with 10 jig OVA plus 2 mg aluminum hydroxide prior to exposure.
Subsequently, these mice received OVA via aerosol as a booster during weeks 3, 6, 9, and 11.
Hippocampi were dissected from all animals 1 day after the final exposure and homogenized in
hypotonic buffer. The 12,000 rpm supernatants were analyzed by Western blotting for the
This document is a draft for review purposes only and does not constitute Agency policy.
4-359 DRAFT—DO NOT CITE OR QUOTE
-------
1 presence of the proteins BC1-2 (which inhibits apoptosis) and Bax (which opposes BC1-2 action
2 and promotes apoptosis) and the NMDA receptor subtypes 2A and 2B (NR2A and NR2B).
3 Immunohistochemical analysis was also carried out for the presence of active caspase-3, an
4 apoptosis marker.
5 The levels of NR2A and NR2B were unaffected by exposure to formaldehyde in either
6 immunized or nonimmunized mice. Likewise, the number of caspase-3 immunoreactive cells
7 did not change as a result of formaldehyde exposure. However, when measured amounts of
8 Bcl-2 and Bax were normalized to the amount of p-tubulin, the ratio Bcl-2/Bax was significantly
9 increased in immunized mice exposed to formaldehyde. Nonimmunized mice did not show this
10 apparently compound-related response. Consistent with the concept that the proportions of Bcl-2
11 and Bax are critical for the regulation of cell survival and death, the authors interpreted their data
12 as an indication that changes to the ratio of Bcl-2/Bax expressions might be an important
13 adaptive response to the effects of formaldehyde, such that the antiapoptotic changes might
14 contribute to the protection of hippocampal neurons from the pernicious effects of formaldehyde
15 exposure itself.
16 The same research group used the immunized mouse model to determine whether
17 formaldehyde exposure affected mRNA expression of genes related to synaptic plasticity
18 (Ahmed et al., 2007). Ten female C3H/HeN mice were exposed to 0 or 400 ppb formaldehyde
19 16 hours/day, 5 days/week for 12 weeks. All mice were immunized with 10 jig OVA plus 2 mg
20 aluminum hydroxide prior to initial formaldehyde exposure then treated in weeks 3, 6, 9, and 11
21 with aerosolized OVA as a booster. Five treated and control animals were IP. injected with 1
22 mg/kg MK-801, a noncompetitive NMDA receptor agonist before the last formaldehyde
23 exposure. At term, hippocampi were dissected and frozen at -80°C until processing. At that
24 point, total mRNA was extracted and first strand cDNA was synthesized by using reverse
25 transcriptase. Expression levels of various proteins/receptors, including NMDA NR2A and
26 NR2B receptor subunits, dopamine Dl and D2 receptors, cyclic AMP responsive element-
27 binding proteins (CREB-1 and CREB-2), and the transcription factors FosB and AFosB were
28 determined by using the PCR. The expression level of each mRNA species was expressed
29 relative to the sample's content of 18S rRNA. The total protein lysate was also assayed for
30 pCREB by Western blotting.
31 In the first of a sequence of histograms, Ahmed et al. (2007) demonstrated a significant
32 increase in mRNA expression of NR2A as a result of formaldehyde exposure. However, this
33 effect was abolished in animals treated with MK-801. A similar trend in the mRNA expression
34 of NR2B in response to formaldehyde exposure did not achieve statistical significance. MK-801
35 treatment significantly reduced receptor in mRNA expression in the presence of formaldehyde.
This document is a draft for review purposes only and does not constitute Agency policy.
4-360 DRAFT—DO NOT CITE OR QUOTE
-------
1 The authors provided data showing an increased expression of dopamine Dl and D2 receptor
2 mRNA response to formaldehyde, in both cases abolished by treatment with MK-801. The
3 expression of CREB-1 mRNA also conformed to the pattern of being increased as a result of
4 formaldehyde exposure but abolished by MK-801. However, the expression of CREB-2 and
5 FosB/AFosB was unaffected by formaldehyde. When normalized to the amount of p-tubulin,
6 there were no significant effects of formaldehyde exposure and MK-801 treatment on the protein
7 levels of pCREB. Finally, there was no significant difference in the expression of transient
8 receptor potential vanilloid receptor (TRPV1) between control and formaldehyde-exposed mice,
9 and MK-801 itself did not significantly alter the mRNA level of TRPV1. In seeking to explain
10 their results, the authors speculated that low-level exposure of immunized mice to formaldehyde
11 had an effect on hippocampal synaptic plasticity at the mRNA level, as evidenced by the
12 enhancement of mRNA for NR2A, the dopamine Dl and D2 receptors, and CREB1, with up
13 regulation compensating for the sustained levels of enhanced protein expression under low-level
14 formaldehyde exposure. The interpretation of these changes in NR2A mRNA, in the context of
15 the results of Tsukahara et al. (2006), showing no change in NR2A and NR2B protein
16 expression, was not discussed.
17
18 4.2.6.1.6. Neurogenesis.
19 Two papers have examined the effects of subacute exposure to formaldehyde on the
20 overall size (volume) of discrete cellular areas of the hippocampus in neonatal rats. The
21 researchers also used an optical fractionator counting method to derive a plausible estimate of
22 cell number. Asian et al. (2006) studied the effects of formaldehyde exposure on the number and
23 volume of granular cells in the hippocampal dentate gyrus. Sarsilmaz et al. (2007) examined the
24 impact of postnatal formaldehyde exposure on brain hemisphere volume and on the size and cell
25 number of pyramidal cells in the cornu ammonis region of the hippocampus. The in-life phase
26 was the same in each study, featuring the exposure of 10 neonatal male Wistar rats/group to 0, 6,
27 and 12 ppm (0, 7.36, and 14.7 mg/m3) formaldehyde 6 hours/day, 5 days/week for 30 days. Five
28 rats/group were sacrificed at that point (PND 30), while the rest were maintained without further
29 treatment until PND 90.
30 For both pyramidal and granular areas, a much lower number of cells was seen on
31 PND 90 versus PND 30 (p < 0.001). This response was evident irrespective of the amount of
32 exposure to formaldehyde and is consistent with normal brain development. Compound-specific
33 effects of formaldehyde on the volume and number of granular and pyramidal cells varied by
34 dose and over the two time points. There was a small increase in the volume of the granular cell
35 layer of the dentate gyrus in rats sacrificed on PND 30 (p < 0.001) in response to increasing
36 formaldehyde concentration (Asian et al., 2006), with no significant change in neuron number;
This document is a draft for review purposes only and does not constitute Agency policy.
4-361 DRAFT—DO NOT CITE OR QUOTE
-------
1 the increased volume (now accompanied by an increase in neuron number) was still evident at
2 the low-exposure level on PND 90 (p < 0.01) but not at the high dose. Brain hemisphere volume
3 was decreased at both concentrations on PND 30 (p < 0.01) but was increased at both
4 concentrations (p < 0.01, with a larger magnitude of effect at 6 ppm) on PND 90 (Sarsilmaz et
5 al., 2007). In the hippocampal cornu ammonis region, the volume of the pyramidal cell layer on
6 PND 30 was increased in low-dose animals (p < 0.001) but decreased in high dose animals (p <
1 0.001) as compared with control values; neither group was significantly different from controls
8 on PND 90. There was a dose-related decrease in total neuron number in the cornu ammonis on
9 PND 30 (p < 0.01 for both doses); on PND 90 the decrease in neuron number remained
10 statistically significant in both treatment groups (p < 0.01), but there was no longer any
11 difference in the magnitude of the effect between doses (Sarsilmaz et al., 2007).
12 In a third study from the same laboratory, Songur et al. (2008) used the same exposure
13 paradigm to evaluate changes in oxidant and antioxidant systems in the cerebellum of perinatally
14 exposed rats. Exposure was carried out as in Asian et al. (2006) and Sarsilmaz et al. (2007),
15 described above. On PND 30 or 90, cerebellums from seven male rats per group were evaluated
16 for levels of malondialdehyde (MDA), NO, superoxide dismutase activity (SOD), and
17 glutathione peroxidase (GPX) activity. Dose-related increases in NO (approximately 20-80%),
18 MDA (100-160%), and GPX (25-60%) and dose-related decreases in SOD (20-30%) were seen
19 on PND 30. In general, the magnitude of change from control levels was maintained on PND 90,
20 with the exception of MDA levels in 6 ppm animals, which appeared to approach control levels
21 at 90 days. The authors stated that these findings indicated that formaldehyde exposure may
22 cause neurotoxicity via the production of oxidative damage in the brain. Persistence of the effect
23 to the 90-day time point (30 days after cessation of exposure) supports the possibility that
24 formaldehyde may cause long-lasting or permanent changes in the brain following early life
25 exposure. These results are consistent with the earlier studies by Asian et al. (2006) and
26 Sarsilmaz et al. (2007), finding permanent changes in brain structure (although in a different
27 brain region) following early life exposure.
28
29 4.2.6.1.7. Summary of formaldehyde effects on neurobehavioral and neuropathological
30 measures, following exposure via inhalation.
31 As has been demonstrated in mice (Wood and Coleman, 1995), it is possible that rats
32 experience respiratory tract irritation during low-level formaldehyde exposure. Perturbations in
33 nervous system function reported with formaldehyde exposure include reductions in motor
34 activity, lack of habituation, impairment in acquisition of a new learning task, deficits in
35 retention of a previously learned task, increases in corticosterone levels, sensitization to cocaine-
36 induced locomotor activity, and enhanced fear conditioning using an olfactory CS (see
This document is a draft for review purposes only and does not constitute Agency policy.
4-362 DRAFT—DO NOT CITE OR QUOTE
-------
1 Table 4-58). Many of these effects were observed at exposure levels at or below 1 ppm, and
2 some persisted days to weeks after termination of exposure.
3 Malek et al. (2004, 2003a, b) detected behavioral changes in rats and mice tested 2 to
4 24 hours postexposure. The mechanism of these behavioral changes is unknown, and available
5 data do not allow dissociation of direct effects on the nervous system and behavioral responses to
6 the irritant effects of formaldehyde (control experiments [e.g., using a different aversive odor
7 with or without irritant properties] were not included). Given that behavioral changes were
8 observed hours to days after cessation exposure (i.e., beyond the time required for formaldehyde-
9 induced irritation to subside), it is unlikely that these behavioral changes were caused by
10 formaldehyde-induced irritation. Similarly, although it is possible that systemic effects of
11 formaldehyde exposure might cause reduced motor activity during or immediately following
12 exposure, it is unlikely that these effects can account for the differences in responses of male rats
13 24 hours after exposure (Malek et al., 2003a). Furthermore, a follow-up study demonstrated
14 reduced motor activity in animals 2 hours after a 2-hour exposure to much lower levels of
15 formaldehyde (0.1 ppm), which fall well below the levels identified by Wood and Coleman
16 (1995) as the AC50 for formaldehyde in mice (Malek et al., 2003b).
17 Two studies reported significant reductions in learning or retention following brief
18 periods of repeated exposure to low levels of formaldehyde (Malek et al., 2003c; Pitten et al.,
19 2000) (see Table 4-58). Malek et al. (2003c) reported an increased number of errors in acquiring
20 a water maze task; testing took place daily 2 hours after termination of a 2-hour exposure. The
21 work of Pitten et al. (2000) revealed that brief exposures over many weeks led to increases in
22 errors performing a previously learned task and that the magnitude of the effect increased over
23 the course of the exposure period. Testing occurred remote from the time of exposure (22 hours
24 after the previous exposure), and the deficits appeared to persist for several weeks after exposure
25 terminated, minimizing the possibility that these effects were related to irritant properties of
26 formaldehyde. Although the exposure levels were moderately high (2.6-4.6 ppm) and continued
27 over several months, the duration of a single exposure event was very brief (10 minutes).
28 Sorg and Hochstatter (1999) and Sorg et al. (2004, 2001) suggest that behavioral
29 sensitization associated with low-level formaldehyde exposure was linked to alterations in HPA
30 control of corticosterone. Cross-sensitization to the locomotor activity-enhancing properties of
31 cocaine and changes in response to a conditioned fear paradigm were observed in animals
32 exposed several weeks earlier to repeated low-level formaldehyde. Direct activation of
33 mesolimbic dopamine pathways or activation of conditioned fear response in the amygdala by
34 formaldehyde could underlie these behavioral effects; these observations were also seen by study
35 authors as consistent with a formaldehyde-induced stress response.
This document is a draft for review purposes only and does not constitute Agency policy.
4-363 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-58. Summary of neurological and neurobehavioral studies of inhaled
formaldehyde in experimental animals
Species
No./
group
Treatment"
Observations
LOAEL/
NOAEL
Reference
Irritant detection threshold
Male
Swiss
mice
8
0,1,1.8,3,
5.6, or 10 ppm
60-second
exposure
episode to
determine
irritant
response
Sensitivity of mice to acute formaldehyde levels
determines the median concentration at which 50% of
exposures were terminated by the subject (AC50)
decreased upon repeat exposure. AC50 = 3.63 for first
series, AC50 = 1.88 ppm for second series.
Time to exposure termination decreased with
increasing formaldehyde concentration.
Time to termination was decreased in repeat exposures.
NAb
Wood and
Coleman
(1995)
Motor activity and habituation
Male
and
female
LEW. IK
rats
Male
and
female
LEW. IK
rats
Male
AB mice
Balb/c
mice
Balb/c
mice
Sprague-
Dawley
rats
10/sex
10/sex
5-7/sex
6
6
8
0,1,2.5, or 5
ppm for 2
hours
0,0.1,0.5, or 5
ppm for 2
hours
0,1.1,2.3, or
5. 2 ppm for 2
hours
0,1.8,3.2,4.5,
6.4, 9.7, or
14. 8 ppm for 3
hours
3.3 ppm for 2
weeks or
2 ppm for 3
weeks
3 hours/day,
5 days/week
0, 5, 10, or 20
ppm;
3 hours/day for
1 or 2 days
Reduced horizontal activity. Number of crossed
quadrants reduced 2 hours after exposure at all doses
for males and females (p < 0.005).
Reduced habituation'. Exposed rats exhibited greater
activity than controls when reintroduced into the testing
environment 24 hours later (males and females, all
doses, p< 0.005).
Reduced horizontal activity: Number of crossed
quadrants reduced 2 hours after exposure at all doses
for males and females (p < 0.005 for males at all doses;
p < 0.005 for females at two higher doses).
Reduced horizontal activity: Number of crossed
quadrants reduced 2 hours after exposure at all doses.
Reduced horizontal and vertical activity: Dose-
dependent decreases in crossed quadrants and rearing.
Significant for males at 1.8 ppm and greater (p < 0.01).
Significant for females at 6.4 ppm or greater (p < 0.01).
Reduced horizontal and vertical activity decreases in
crossed quadrants and rearing.
3.3 ppm (2 weeks) and 2 ppm (1 week) (p < 0.01, p <
0.05).
Reduced activity levels on both days. Decreases seen at
5 and 10 ppm; data reported only for 5 ppm group
LOAEL =
1 ppm
2 hours
LOAEL =
0.1 ppm
2 hours
in males
LOAEL =
1.1 ppm
2 hours
LOAEL =
1.8 ppm
3 hours
in males
LOAEL =
2 ppm
3 weeks
LOAEL =
5 ppm,
3 hours
Malek et al.
(2003a)
Malek et al.
(2003b)
Malek et al.
(2004)
Usanmaz et
al. (2002)
Usanmaz et
al. (2002)
Boja et al.
(1985)
This document is a draft for review purposes only and does not constitute Agency policy.
4-364 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-58. Summary of neurological and neurobehavioral studies of inhaled
formaldehyde in experimental animals (continued)
Species
Rats
Rats
No./
group
15
Treatment"
0 or 0.5 mg/m3
(0.4 ppm) on
CDs 1-19, 4
hours/day
0, 0.005, or
0.0005 mg/L
(approximately
4 or 0.4 ppm),
CDs 1-19
Observations
Increased motor activity on the 2nd and 3rd days of
testing (reduced habituation) in offspring exposed in
utero. Increased number of squares entered (p < 0.0 1)
and frequency of rearing (p < 0.05).
Changes in motor activity at one and two months in
offspring exposed in utero. Decreased spontaneous
mobility at 1 month in both sexes, increased activity
at 2 months in both sexes.
LOAEL/
NOAEL
LOAEL =
0.4 ppm,
gestational
LOAEL =
0.4 ppm,
gestational
Reference
Senichenkova
(1991)
Sheveleva
(1971)
Learning and memory
Adult
male and
female
LEW. IK
rats
Adult
male and
female
Wistar
rats
Ferrets
Male
juvenile
and
adult
rats
15/sex
5-87
sex
5/group
0,0.1,0.5, or
5.4 ppm
2 hours for 10
consecutive
days
0, 2.6, 4.6 ppm
10 minutes/
day for 90
consecutive
days
0.25 ppm
0.25, 0.5 ppm
Impairment in acquisition of a new task: Male and
female rats at all formaldehyde exposures had
significantly more errors in completing a water
labyrinth (p< 0.01).
Male and female rats had longer times to completion
of the maze at 0.5 and 5.4 ppm (p < 0.05, p< 0.01).
Deficit in the retention of a learned task: Male and
female rats committed significantly more errors (p <
0.05) and took more time to complete the land maze
in across the course of the experiment (p < 0.04).
Impairment in acquisition of a new task: Exposed
ferrets only achieved a 75% success rate in training to
discriminate odors in a Y-maze versus 90% success
rate in controls.
Note: The results are confounded with other effects
on the olfactory epithelium. The same researchers
also reported a decrease in olfactory sensitivity and a
reduction in percentage of olfactory cells in similarly
treated animals.
Decreases in olfactory thresholds, in juvenile animals
but not in adults (p < 0.002).
LOAEL =
0.1 ppm
2 hours/
10 days
LOAEL =
2.6 ppm
10 minutes/
90 days
None
established
LOAEL =
250 ppm in
juveniles
Malek et al.
(2003c)
Pitten et al.
(2002)
Apfelbach
and
Reibenspies
(1991)
(abstract
only)
Weiler and
Apfelbach
(1992)
(abstract
only)
This document is a draft for review purposes only and does not constitute Agency policy.
4-365 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-58. Summary of neurological and neurobehavioral studies of inhaled
formaldehyde in experimental animals (continued)
Species
No./
group
Treatment"
Observations
LOAEL/
NOAEL
Reference
Neurosensitization endpoints
Female
Sprague-
Dawley
rats
Female
and male
Sprague-
Dawley
rats
Female
and male
Sprague-
Dawley
rats
Male
Sprague-
Dawley
rats
Female
and male
Sprague-
Dawley
rats
8-9
Various
up to
24/group
Various
up to
24/group
4-9 or
16
4-9 or
16
0 or 1 1 ppm
1 hour/day, 7
days
1 1 ppm, 1
hour/day,
7 days
1 ppm, 1
hour/day,
7 days
1 ppm, 1
hour/day,
5 days/week,
4 weeks
1 ppm, 2
hours/day,
5 days/week,
4 weeks
0,0.7, or 2.4
ppm for 20 or
60 minutes
0,0.7, or 2.4
ppm
1 hour/day,
5 days/week
for 2 or 4
weeks
0 or 2 ppm
1 hour/day,
5 days/week
for 4 weeks
Increase in cocaine-induced activity. Increased
quadrants crossed after cocaine injection 1 and 2 days
after exposure (p = 0.05 and;? < 0.04, respectively).
No change in corticosterone levels 28 days
postexposure.
No change in nociceptive or passive avoidance test or
plus-maze (21, 20, and 13 days postexposure,
respectively) (21 days).
Increase in cocaine-induced activity: Increase in
rearing after
cocaine injection 1 day after exposure but not 4-6
weeks after exposure; 1 1 ppm for 7 days or 1 ppm for
4 weeks.
No change in rats exposed at 1 ppm for 1 week.
Increased conditioned fear response in formaldehyde-
treated, foot-shock-conditioned rats, twofold (p <
0.05).
No change in corticosterone in acute (20- and 60-
minute) exposures.
Increase in basal corticosterone'. 0.7 ppm for 2 or 4
weeks.
Hyperresponsive corticosterone response to
environment'.
2.4 ppm for 2 or 4 weeks.
Increased conditioned fear response to an olfactory
cue in formaldehyde-treated, foot-shock-conditioned
male rats. Measured as increased time freezing when
presented with a novel environment (p < 0.05).
No effect in female rats.
LOAEL =
11 ppm/
7 days
(unbounded)
LOAEL =
1 ppm
4 weeks
NOAEL =
1 ppm
7 days
LOAEL =
1 ppm
4 weeks
LOAEL =
0.7 ppm/
2 weeks
NOAEL =
0.7 ppm/
20 minutes
LOAEL =
2 ppm/
4 weeks
NOAEL =
2 ppm/
4 weeks
Sorg et al.
(1996)
Sorg and
Hochstatter
(1999)
Sorg and
Hochstatter
(1999)
Sorg et al.
(2001)
Sorg et al.
(2004)
This document is a draft for review purposes only and does not constitute Agency policy.
4-366 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-58. Summary of neurological and neurobehavioral studies of inhaled
formaldehyde in experimental animals (continued)
Species
No./
group
Treatment"
Observations
LOAEL/
NOAEL
Reference
Neurochemistry and neuropathology
8-week-
old
female
C3H/
HeN
mice
Adult
male
Sprague-
Dawley
rats
5
8
0, 0.08, 0.4, or
1 ppm
16 hours/day,
5 days/week
for 1 day or
12 weeks
0, 5, 10, 20
ppm
3 hours/day, 1
or 2 days
No change in size of main olfactory bulb by several
measures.
No change in numbers of PG cells.
No change in tyrosine hydroxylase immunopositive
PG cells after 1 day.
Increase in tyrosine hydroxylase-immunopositive PG
cells after 12 weeks to 196, 167, and 196% of
controls at 0.08, 0.40, and 1 ppm, respectively.
No change in norepinephrine or 5-hydroxytryptamine
in hypothalamus.
Increase in DOPAC in hypothalamus after one
exposure.
Increase in dopamine and 5-HIAA in hypothalamus
after two exposures.
LOAEL =
0.08 ppm/
12 weeks
LOAEL =
5 ppm/
3 hours
Hayashi et al.
(2004)
Boja et al.
(1985)
Neut 'agenesis
Neonatal
Wistar
rats
Neonatal
Wistar
rats
Neonatal
Wistar
rats
5
5
7
0, 6, or 12
ppm,
6 hours/day,
5 days/week
for 30 days
0, 6, or 12
ppm,
6 hours/day,
5 days/week
for 30 days
0, 6, or 12
ppm,
6 hours/day,
5 days/week
for 30 days
Changes in volume of granular cell layer of the
dentate gyrus in the hippocampus at postnatal days
30and90(p
-------
1 Limited data regarding possible neurochemical changes in the brains of formaldehyde-
2 exposed, immunized mice (Ahmed et al., 2007; Fujimaki et al., 2004b; Hayashi et al., 2004;
3 Kakeyama et al., 2004) and rats (Boja et al., 1985) provided conflicting information, and the
4 implications of these data regarding possible formaldehyde neurotoxicity are difficult to
5 determine.
6 In developmental exposure paradigms, changes in brain structure (Sarsilmaz et al., 2007;
7 Asian et al., 2006), brain chemistry (Songur et al., 2008), and motor activity (Senichenkova,
8 1991; Sheveleva, 1971) were seen following neonatal or in utero exposure to formaldehyde. In
9 addition, Weiler and Apfelbach (1992) found juvenile animals to be more sensitive to
10 formaldehyde-induced changes in olfactory thresholds when compared with adult animals.
11 These studies raise concern about possible long-lasting neurological effects of early exposure to
12 formaldehyde. It is important to note, however, that exposure levels in these studies were higher
13 (250-6,000 ppb) than those producing the behavioral effects in adults described above.
14 Overall, available data provide substantial evidence of behavioral effects, including
15 motor activity changes and changes in learning and retention, following repeated exposure to
16 relatively low levels of formaldehyde. These effects were seen in multiple laboratories, in
17 studies conducted by different authors, and using different behavioral paradigms. These
18 conclusions are also supported by more limited data, indicating possible developmental effects
19 on the nervous system, including changes in brain structure and in the behavior of offspring; the
20 developmental findings are less robust since they were seen only in individual laboratories and
21 occurred following exposure to higher concentrations of formaldehyde. Studies evaluating
22 developmental neurotoxicity at lower doses, comparable to those used in the adult studies, were
23 not available. None of the available data provide sufficient information to allow a determination
24 of the mechanism for these behavioral changes, although it is unlikely that they are attributable
25 to the irritant properties of formaldehyde. The data regarding behavioral sensitization provide
26 some support for a stress-related mechanism for those findings, but the applicability of this
27 mechanism to the behavioral changes seen in the other studies, including the learning deficits,
28 has not been evaluated.
29
30 4.2.6.2. Oral Exposure
31 Available data regarding neurotoxic effects of formaldehyde exposure following oral
32 exposure are very limited. Several chronic or subchronic oral toxicity studies evaluated changes
33 in brain weight or histopathology in rats or dogs following repeated oral exposures to
34 formaldehyde at doses as high as 300 mg/kg-day, administered in drinking water (Til et al.,
35 1989, 1988; Tobe et al., 1989; Johannsen et al., 1986). Although data were not presented in the
This document is a draft for review purposes only and does not constitute Agency policy.
4-368 DRAFT—DO NOT CITE OR QUOTE
-------
1 publications, all stated that no changes in brain weight or pathology were seen in the standard
2 evaluations performed in these studies.
3 Two studies evaluated changes in behavior following exposure to formaldehyde in
4 drinking water (Venkatakrishna-Bhatt et al., 1997; Venkatakrishna-Bhatt and Panchal, 1992).
5 Venkatakrishna-Bhatt and Panchal (1992) evaluated changes in performance on a conditioned
6 avoidance task in adult male albino rats (five/group). Animals were exposed to formaldehyde in
7 drinking water (10 mg/mL) or by IP. injection (10 mg/kg) for 60 days. Although it was stated
8 that water consumption was recorded, the data were not presented, and thus actual exposure
9 levels cannot be documented. Prior to the initiation of exposure, rats were trained on the
10 conditioned avoidance task (climbing a wooden pole in response to a warning buzzer, thus
11 avoiding electric shock from a floor grid). Rats were trained to a predetermined performance
12 criterion (not described); animals not achieving the criterion were removed from study. Training
13 and testing conditions (e.g., retest interval and duration of sessions) were not well described.
14 Data were presented as percent response in behavioral performance (apparently separately for the
15 escape or avoidance aspects of the task) or percentage decrease in response. No control data
16 were presented, and pretreatment performance was not described. Figures presented
17 performance at 10-day intervals, starting with day 0, with each data point stated to represent the
18 mean for five experimental sets; again, the interval between experimental sets and the number of
19 trials per set was not specified. Although the authors concluded that a deficit in performance
20 was demonstrated, the data as presented were difficult to interpret and the conclusion could not
21 be verified based on the data as presented.
22 Venkatakrishna Bhatt and Panchal (1992) examined changes in performance on a
23 conditioned avoidance response, presumably using a procedure similar to the one described
24 above. Albino rats (sex not specified, five/group) were exposed to formaldehyde in drinking
25 water at 0.2 or 0.5 mg/mL for 90 days. As described above, rats were trained in performing the
26 task prior to the start of exposure. Venkatakrishna-Bhatt Bhatr and Panchal (1992) stated that
27 there was a dose-related deterioration of performance, but no data were presented to support
28 these conclusions.
29 In summary, available data are insufficient to conduct a reliable assessment of neurotoxic
30 effects of formaldehyde following oral exposure. Limited data suggest a lack of overt
31 neuropathological changes at doses up to 300 mg/kg-day (Til et al., 1989, 1988; Tobe et al.,
32 1989; Johannsen et al., 1986), but detailed information regarding the types of neuropathological
33 evaluations performed in those studies is not available, and thus no firm conclusions can be
34 drawn regarding the potential for neuropathological effects. The two available studies evaluating
This document is a draft for review purposes only and does not constitute Agency policy.
4-369 DRAFT—DO NOT CITE OR QUOTE
-------
1 behavioral changes are not considered to provide useful information, and thus effects on nervous
2 system function could not be evaluated.
O
4 4.2.6.3. Summary
5 Overall, there is strong evidence that formaldehyde exposure via inhalation may cause
6 adverse effects on nervous system function in experimental animals at relatively low levels of
7 exposure (LOAELs as low as 100 ppb). Although human data regarding neurotoxicity following
8 formaldehyde inhalation are limited, available data provide support that the types of effects seen
9 in humans are similar to those found in animal studies. Evidence from available human
10 controlled inhalation exposure studies indicates that humans may be affected at doses similar to
11 those used in animal studies; however, the human data are extremely limited.
12 There are insufficient data to evaluate the potential for neurotoxicity following oral
13 exposure to formaldehyde. Limited evaluations of brain weight or histopathology in available
14 chronic or subchronic oral studies found no evidence of formaldehyde-induced changes (Til et
15 al., 1989, 1988; Tobe et al., 1989; Johannsen et al., 1986). However, reliable studies examining
16 nervous system function or focused studies of neuropathology following oral exposure to
17 formaldehyde are not available.
18
19 4.2.6.4. Other Considerations
20 Major data gaps were found regarding the evaluation of changes in nervous system
21 structure or function following formaldehyde exposure by both the inhalation or oral routes.
22 With respect to inhalation exposure, none of the available human studies resulted in data
23 sufficient to conduct a reliable dose-response assessment for changes in nervous system function.
24 Most of the available animal inhalation studies used short exposure durations (acute or
25 short-term), precluding a reliable evaluation of neurotoxicity following chronic exposure.
26 Available data for neurodevelopmental exposures are also quite limited, consisting of evaluation
27 of neuropathology in only one brain region and functional evaluations focused only on changes
28 in motor activity.
29 Major data gaps also exist regarding neurotoxicity following oral exposure, with no
30 relevant human data and extremely limited animal data. Available oral exposure studies were
31 insufficient to permit a reliable evaluation of the potential for neurotoxicity following oral
32 exposure to formaldehyde.
33
This document is a draft for review purposes only and does not constitute Agency policy.
4-370 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.2.7. Reproductive and Developmental Toxicity
2 The potential for developmental and reproductive effects after formaldehyde exposure by
3 the inhalation route has generally been considered low, since formaldehyde, as a reactive gas, is
4 not expected to penetrate past the POE (NEG, 2003; IPCS, 2002; Collins et al., 2001).
5 Nevertheless, a number of animal studies have demonstrated effects of formaldehyde on pre- and
6 postnatal development and on the reproductive system. For example, developmental toxicity
7 was observed in two studies that evaluated a standard battery of developmental endpoints
8 resulting from inhalation exposure on GDs 6-10 (Martin, 1990; Saillenfait et al., 1989).
9 Similarly, oral exposures resulted in developmental effects when administered during
10 comparable gestational windows (Marks et al., 1980; Hurni and Ohder, 1973). There have also
11 been reports that identified developmental effects at lower-level formaldehyde exposures that
12 were administered throughout gestation (Senichenkova and Chebotar, 1996; Senichenkova,
13 1991; Kitaev et al., 1984; Sheveleva, 1971; Gofmekler and Bonashevskaya, 1969; Gofmekler,
14 1968; Pushkina et al., 1968). Postnatal functional consequences of developmental exposures
15 have also been identified (Sarsilmaz et al., 2007; Asian et al., 2006; Weiler and Apfelbach, 1992;
16 Senichenkova, 1991; Sheveleva, 1971). Additionally, a number of studies suggest that
17 formaldehyde adversely affects the male reproductive system after both inhalation and oral
18 exposures (Xing et al., 2007; Zhou et al., 2006; Ozen et al., 2005, 2002; Sarsilmaz et al., 1999;
19 Chowdhury et al., 1992; Cassidy et al., 1983; Guseva, 1972). This section reviews the available
20 published studies assessing reproductive and developmental endpoints of formaldehyde.
21
22 4.2.7.1. Inhalation Studies Addressing Developmental and Reproductive Toxicity
23 Saillenfait et al. (1989) reported a comprehensive and well-documented developmental
24 study in Sprague-Dawley rats. Pregnant rats were exposed beginning on GD 6 in order to cover
25 critical stages of development (e.g., implantation and major organogenesis). Female Sprague-
26 Dawley rats (25/group) were exposed to 0, 5, 10, 20, or 40 ppm (0, 6.15, 12.3, 24.6, or
27 49.2 mg/m3) formaldehyde 6 hours/day on GDs 6-20. The onset of pregnancy was determined
28 by the presence of sperm in a vaginal smear. Dams were exposed to formaldehyde in a dynamic
29 flow chamber, and formaldehyde concentrations were determined to be 0, 5.17 ± 0.51, 9.92 ±
30 0.88, 20.04 ± 0.88, and 38.96 ± 3.70 ppm. Dams were weighed on GDs 0, 6, and 21 and
31 sacrificed on day 21. Upon examination, uterine weights, fetal weights, sex ratio, number of
32 implantation and resorption sites, and live and dead fetuses were recorded. Fetuses were
33 examined for external malformations and cleft palate. One-half of viable fetuses were sectioned
34 to assess soft-tissue alterations. The other half were fixed, stained with alizarin red S, and
35 examined for skeletal alterations.
This document is a draft for review purposes only and does not constitute Agency policy.
4-371 DRAFT—DO NOT CITE OR QUOTE
-------
1 Body weight gain of dams and body weight of male and female fetuses were reduced by
2 exposure to 40 ppm formaldehyde to 49, 78, and 81% of control values, respectively (p < 0.01)
3 (Saillenfait et al., 1989). Reduced weight gain in dams remained significantly decreased when
4 uterine weight was accounted for (p < 0.01). Mean fetal weight of male pups was reduced at
5 maternal exposures of 20 and 40 ppm formaldehyde (5.53 and 4.42 g versus 5.61 g in controls).
6 Decreased fetal body weight in females was only seen at 40 ppm (4.27 g versus 5.24 g in
7 controls). All other pregnancy endpoints were unchanged by formaldehyde exposure (e.g.,
8 uterine weight, implantation and resorption sites, live fetuses, dead fetuses, and sex ratios). No
9 major malformations were noted in fetuses. Some minor soft tissue and skeletal anomalies, such
10 as dilated ureter, missing sternebrae, extra fourteenth rib, and rudimentary thirteenth rib
11 (statistics not given), were reported. However, these effects occurred at similar frequencies in
12 control and treatment groups. The incidence of delayed ossification of the thoracic vertebrae
13 was 8.7% in fetuses from the 40 ppm exposure group versus 1.8% in controls. However, this
14 difference was not statistically significant. Overall, from these results formaldehyde was neither
15 lethal to embryos nor teratogenic, only exhibiting fetotoxic effects at exposures of 20 ppm and
16 above. These are levels where there was a significant decrease in fetal body weight.
17 Martin (1990) conducted a similar study exposing pregnant rats during similar stages of
18 development. Mated female Sprague-Dawley rats (25/group) were exposed to 2, 5, or 10 ppm
19 (2.46, 6.15, or 12.3 mg/m3) formaldehyde 6 hours/day on GDs 6-15. The study included two
20 control groups: dams placed in the exposure chambers once a day but exposed only to clean air
21 and dams fed and housed similarly to the experimental groups but never put into the inhalation
22 chambers. The method of formaldehyde vapor generation and details of the exposure chamber
23 were not described. Mean formaldehyde exposure concentrations were reported as 1.88, 4.88,
24 and 9.45 ppm (variability not given, analytical method not discussed). Food consumption and
25 body weight were recorded. On GD 20, rats were sacrificed, and the following pregnancy
26 parameters were recorded: live fetuses, dead fetuses and resorptions, number of corpora lutea,
27 fetal weights, sex ratios, and preimplantation and postimplantation losses. Fetuses were
28 examined for major malformations, minor external and visceral anomalies, and minor skeletal
29 anomalies (details not given). Weight gain and food consumption in dams were said to be
30 reduced at 10 ppm (p < 0.05). Formaldehyde exposure of the dams at 5 and 10 ppm led to an
31 increased incidence of reduced ossification of the pubic and ischial bones in fetuses on GD 20
32 (p < 0.05). Reduced ossification correlated with lower fetal weights, and the author considered
33 both of these findings a result of larger litter size and, therefore, not related to formaldehyde
34 exposure. However, no tables presenting the data or statistical analysis were provided. All other
35 pregnancy parameters and fetal anomalies were described as unaffected by formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-372 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure. However, without data presented for the assessed endpoints, background rates of
2 malformations, trends in the data, and variability, it is difficult to evaluate the Martin (1990)
3 comparisons. However, the author's observations of reduced fetal weight and increased
4 incidence of reduced ossification are consistent with the results of Saillenfait et al. (1989).
5 Kilburn and Moro (1985) studied similar endpoints but included formaldehyde exposure
6 during earlier gestational windows. The study report, only available in abstract form and not
7 found as a subsequent published article, does not provide many methodological details. Female
8 rats (number and strain not reported) were exposed to 0 or 30 ppm (0 or 37.2 mg/m3)
9 formaldehyde 8 hours/day during GDs 3-17, 3-12, 8-12, or 9-11. A second experiment
10 included pair-fed controls for dams exposed to 30 ppm formaldehyde during GDs 3-17, 3-12,
11 and 8-12. The authors reported reductions in fetal and maternal weight gain that were greater
12 than decreases in pair-fed controls. Fetal anomalies were noted after 15 days of gestational
13 exposure (e.g., altered organ size and undescended testes). Although the report indicates some
14 maternal toxicity and fetotoxic effects (for example, stunted growth), lack of study details and
15 clear reporting make this report of negligible utility in human health risk assessment.
16 There are several early studies that examined developmental effects of formaldehyde
17 exposure administered throughout gestation (Gofmekler and Bonashevskaya, 1969; Gofmekler,
18 1968; Pushkina et al., 1968). It is unclear if these reports represent the same or overlapping
19 experimental groups. They were performed in the same laboratory and are reported with a
20 similar level of detail. The source of formaldehyde, method of vapor generation, exposure
21 conditions (dynamic versus static), confirmation of exposure concentrations, study design, and
22 data presentation details were not provided. Absence of such critical information detracts from
23 the quality of these studies as a coherent record of experimental information, and, thus, these
24 findings can only be utilized qualitatively in the formaldehyde risk assessment.
25 In the Gofmekler (1968) study, female rats (36, strain not specified) were continuously
26 exposed at 0, low, or high formaldehyde concentrations beginning 10-15 days prior to mating
27 (target concentrations of 0, 0.01, or 0.81 ppm formaldehyde [0, 0.01, or 1 mg/m3]). The author
28 reported a 14-15% increase in pregnancy duration and a decrease in litter size (data not shown).
29 However, males and females were mated 6-10 days, and no information was provided on how
30 mating and conception were confirmed. No external malformations were attributed to
31 formaldehyde exposure. Concentration-dependent increases in pup body weight and decreases in
32 lung and liver weight were attributed to formaldehyde exposure. Pup weights were increased
33 from 5.6 g in controls to 6.0 and 6.3 g in groups 1 and 2, respectively (p < 0.01 andp < 0.001).
34 Formaldehyde exposure increased pup adrenal weight in both groups and pup thymus and kidney
35 weight in group 2 only (see Table 4-59).
36
This document is a draft for review purposes only and does not constitute Agency policy.
4-373 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-59. Effects of formaldehyde on body and organ weights in rat pups
from dams exposed via inhalation from mating through gestation
4
5
6
7
8
9
10
11
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
(ppm)a
0
0.01
0.81
Body weight
(g)
5.6
6.0b
6.3C
Relative organ weights (mg/10 g body weight)
Thymus
26
25.1
31.7C
Heart
61.4
61.5
64.5
Lung
287.1
230.2C
223. 2C
Liver
587.7
557.7d
550.8b
Adrenals
3.2
4.2C
3.8d
Kidney
51.4
53.4
55.7b
aDams were exposed to formaldehyde continuously from 10-15 days prior to mating. Exposure concentrations
were not validated.
'^Different from controls, p < 0.01.
"Different from controls, p < 0.001.
dDifferent from controls, p < 0.05.
Source: Gofmekler(1968).
In a study by Gofmekler and Bonashevskaya (1969), the researchers evaluated organ
histopathology in pups from similarly treated dams. Pregnant female albino rats were
continuously exposed at two formaldehyde concentrations (groups 1 and 2, as described above).
Adult males were similarly exposed during mating. Offspring were examined for malformations,
and the organs were fixed and sectioned for histopathologic examination, including hematoxylin
and eosin staining, Brachet stain for RNA, and Feulgen stain for DNA. Liver and kidney
sections were also stained with Schiff s reagent (which reacts with aldehydes), with Sudan III for
lipids, and Pearl's stain for iron. Placenta, uterus, and ovaries from the dams and testes of the
males were sectioned, stained, and evaluated. The authors stated that formaldehyde induced no
external anomalies (reported elsewhere, but no reference given). The authors also noted
involution of lymphoid tissue and changes in liver, mild hypertrophy of Kupffer cells, and
numerous extramedullary myelopoietic centers in pups from dams in group 2. Pups from both
treatment groups showed reduced glycogen content in the myocardium and the presence of iron
in Kupffer cells. There was a localized increase in positive reaction to Schiff s reagent in the
basement membrane and intertubular connective tissue of the kidneys. The authors suggested
that this was an indication of functional alterations in the renal tubular apparatus. All other
tissues examined and histochemical staining indicated no differences due to formaldehyde
exposure.
Researchers in the same laboratory (Pushkina et al., 1968) studied the effects of
formaldehyde exposure on vitamin C (ascorbic acid, an antioxidant) and nucleic acid levels in
dams and fetuses as general measures of toxicity. Female white rats (n = 160) were continuously
exposed at two formaldehyde concentrations (groups 1 and 2, as described above) from 20 days
This document is a draft for review purposes only and does not constitute Agency policy.
4-374 DRAFT—DO NOT CITE OR QUOTE
-------
1 prior to mating (6-10 days) and then throughout gestation. Dams were sacrificed and ascorbic
2 acid and nucleic acid levels determined in harvested organs (methods referenced but not
3 described). No visible malformations in pups were noted. Formaldehyde exposure increased
4 fetal body weight and organ weight in both groups (data not given). There was an average of
5 11.3 fetuses per litter for control dams versus 9.8 and 8.6 for groups 1 and 2, respectively. The
6 authors reported that formaldehyde exposure decreased DNA levels and increased RNA levels in
7 organs (further details not provided). Formaldehyde exposure resulted in lower vitamin C levels
8 in the whole fetus (76 and 75% of controls) and in maternal liver specimens (82 and 88% of
9 controls) for exposure groups 1 and 2, respectively (p < 0.05). In contrast, vitamin C was higher
10 in fetal liver (127% of controls) in group 1 (p < 0.05). The significance of these differences is
11 unknown. The authors considered the results as general measures of biochemical changes and
12 therefore toxic.
13 The reports of Gofmekler and Bonashevskaya (1969), Gofmekler (1968), and Pushkina et
14 al. (1968) lack key methodological details. As discussed above, exposure conditions and actual
15 formaldehyde concentrations cannot be validated. Although methods were not thoroughly
16 detailed, results were reported in data tables with statistics and detailed descriptions of observed
17 pathological changes. However, without validation of exposure concentrations, these findings
18 can only be considered qualitatively.
19 In another study, Sheveleva (1971) exposed female mongrel (i.e., not a homogeneous
20 genetic strain) white rats to 0, 0.0005, or 0.005 mg/L (0, 0.5, or 5 mg/m3) (0, 0.4, or 4 ppm)
21 formaldehyde on GDs 1-19 (where GD 1 was defined as the day that spermatozoa were detected
22 in vaginal smears) for 4 hours/day. In each group, 15 dams were terminated on GD 20 for
23 evaluation of ovarian corpora lutea, uterine implantation sites, pre- and postimplantation loss,
24 number of live fetuses, fetal length and weight, and external examination for malformations.
25 Additionally, in each group, six dams were allowed to deliver their litters. Developmental
26 landmarks were monitored (i.e., ear and eye opening, incisor eruption, emergence of hair coat),
27 and the pups were further evaluated at 1 and 2 months of age for body weight, threshold of
28 neuromuscular excitability, total oxygen consumption in 1 hour per 100 g of weight, and
29 spontaneous mobility over 10 minutes. Maternal toxicity (recorded on GD 17) included
30 significantly (p < 0.05) decreased leukocyte counts in both treated groups and a number of
31 additional findings at 0.005 mg/L (i.e., significant reductions in the threshold of neuromuscular
32 excitability, rectal temperature, and blood hemoglobin level) as well as an increase in
33 spontaneous mobility over 15 minutes. Fetal examinations on GD 20 identified a 50-70%
34 increase in mean preimplantation loss in both formaldehyde-exposed groups. When pups were 1
35 month of age, a reduction in spontaneous mobility was noted in both treated groups; in pups at 2
This document is a draft for review purposes only and does not constitute Agency policy.
4-375 DRAFT—DO NOT CITE OR QUOTE
-------
1 months of age, an increase in mobility was observed in the 0.005 mg/L group. Also, when pups
2 were 2 months of age, there were alterations in hemoglobin levels and leukocyte counts in both
3 treated groups. Detailed descriptions of some study methodologies (particularly in regard to
4 neurological and behavioral assessments) were not provided in the published paper.
5 Based on a review of the work by Gofmekler (1968) and various epidemiologic studies
6 available at the time, Kitaev et al. (1984) hypothesized that formaldehyde may exert toxic effects
7 in the early days of gestation. To study embryotoxic effects of formaldehyde inhalation
8 exposures, mature female Wistar rats (five to nine per group) were exposed to 0.41 or 1.22 ppm
9 (0.5 or 1.5 mg/m3) formaldehyde 4 hours/day, 5 days/week for 4 months (Kitaev et al., 1984).
10 Rats were exposed in dynamic flow chambers and formaldehyde levels measured gravimetrically
11 (but not reported). Females were mated on day 120 of exposure and mating confirmed by the
12 presence of sperm in a vaginal smear. Embryos were harvested on the second or third day of
13 pregnancy (GD 2 or 3) and examined by both light and phase contrast microscopy for changes in
14 morphology (i.e., evidence of embryonic degeneration). Additionally, maternal weight gain and
15 organ weights (ovaries, uterus, and adrenal glands) and blood samples (HCT, Hb, and TP) were
16 monitored as indicators of general toxicity. These parameters were unchanged by formaldehyde
17 exposure. Formaldehyde exposure at 1.22 ppm for 4 months resulted in an increased number of
18 degenerating embryos on GD 3 (14.9 versus 4.4% in controls) and a smaller increase of 10.2%
19 (versus 5.1% in controls; statistical significance not assessed) on GD 2. Indications of
20 degeneration included reduced size and changes in appearance (granulation of the ooplasm,
21 wrinkling and degradation of nuclear material). However, it is unclear if litter effects were
22 accounted for in the statistical analyses, and it is unknown how the affected embryos were
23 distributed between litters. For dams exposed to 0.41 ppm formaldehyde, the number of
24 degenerated embryos was not increased on day 2 (3.8 versus 5.1% in controls) but was increased
25 on day 3 (9.1 versus 4.4% in controls; again, unknown if statistically significant) after maternal
26 exposure to 1.22 ppm formaldehyde. This observation may be coincidental since it was seen in
27 dams sacrificed on GD 2 but not in those sacrificed on GD 3. Kitaev et al. (1984) considered
28 these findings to indicate that repeated exposure to formaldehyde over a 4-month period can
29 disturb reproductive function, resulting in adverse effects early in embryonic development.
30 To further explore the effects of inhalation exposures to formaldehyde on reproductive
31 function, Kitaev et al. (1984) conducted a second series of experiments on 200 similarly treated
32 female rats. After 4 months of repeated formaldehyde exposure at 0.41 or 1.22 ppm as described
33 above, organ weights (ovaries and uterus) and blood levels of gonadotropic hormones and
34 progesterone were determined. However, the day and time of hormone measurement were not
35 given in the report, and normal diurnal variations in these hormones could affect the reported
This document is a draft for review purposes only and does not constitute Agency policy.
4-376 DRAFT—DO NOT CITE OR QUOTE
-------
1 findings if time of day was not accounted for. The length of the estrous cycle was unchanged
2 during exposure. Formaldehyde exposure modulated gonadotropin levels and relative ovarian
3 weight, suggesting low-level effects on the female rat reproductive system prior to mating
4 (Kitaev et al., 1984). Ovarian weight and blood levels of luteinizing hormone (LH) were both
5 significantly increased after exposures at 0.41 ppm formaldehyde but remained at control levels
6 in rats exposed at 1.22 ppm. Blood levels of follicle-stimulating hormone (FSH) were increased
7 approximately 66% from control after 1.22 ppm formaldehyde exposure (p < 0.05).
8 Progesterone levels were unchanged by formaldehyde treatment. Kitaev et al. (1984) suggested
9 a role of the hypothalamus-pituitary system based on increased ovary weight, a greater number
10 of degenerated embryos, and increased LH in rats exposed at 0.41 ppm. They postulated that
11 these effects were not seen at 1.22 ppm due to a toxic effect exhibited as embryonic
12 degeneration, thus the absence of a dose-response did not alter the interpretation of the validity
13 of the adverse response. The study NOAEL was not determined, and the study LOAEL was 0.4
14 ppm (0.5 mg/m3), based upon increased early embryo loss and on maternal outcomes (increased
15 ovarian weight and increased blood LH levels) following 4 months of formaldehyde treatment.
16 For the finding of increase blood FSH levels, the endpoint NOAEL was 0.4 ppm (0.5 mg/m3)
17 and the LOAEL was 1.2 ppm (1.5 mg/m3).
18 Senichenkova and Chebotar (1996) and Senichenkova (1991) examined reproductive and
19 developmental effects of daily formaldehyde exposure on GDs 1-19 of pregnancy, including the
20 potential effect of formaldehyde exposure on development early in gestation. Additionally, since
21 anemia adversely affects fetal development, Senichenkova and Chebotar (1996) also examined
22 formaldehyde effects in iron-deficient dams to determine whether coexposure further
23 compromises fetal development. In both studies, female white rats were exposed to 0 or
24 0.41 ppm formaldehyde (0 or 0.5 mg/m3), 4 hours/day on GDs 1-19. Formaldehyde
25 concentrations in the dynamic exposure chambers were measured gravimetrically to confirm the
26 exposure concentration but were not reported (methods not provided). It is unclear if gravimetric
27 measurements would be sensitive or accurate enough to validate these low-exposure
28 concentrations without a better understanding of the methodology. This uncertainty in exposure
29 conditions should be considered in evaluating the reported results.
30 Mongrel female white rats were exposed at a target concentration of 0 or 0.41 ppm (0 or
31 0.5 mg/m3) formaldehyde 4 hours/day on GDs 1-19 (Senichenkova, 1991). On GD 20, a subset
32 of the dams was sacrificed and examined for number of corpora lutea, implantation and
33 resorption sites, live/dead fetuses, and fetal weights. Fetuses were examined for gross pathology
34 of the internal organs and skeleton (details not given). Blood pH, partial pressure of CO2, and
35 partial pressure of oxygen were measured in both dams and embryos. The remaining dams were
This document is a draft for review purposes only and does not constitute Agency policy.
4-377 DRAFT—DO NOT CITE OR QUOTE
-------
1 brought to term to study postnatal effects of formaldehyde exposure. Rat pups were observed on
2 PNDs 1-25 for viability, physical development, and maturation rate of motor reflexes. Behavior
3 of juvenile offspring (PND 40) was studied in an open field test, and maze learning was tested at
4 sexual maturity. In a follow-up report, Senichenkova and Chebotar (1996) present blood
5 chemistry data, pregnancy outcome, and developmental data for similarly treated dams and their
6 pups in a chemical model of iron deficiency. Intraperitoneal injections of the iron-chelating
7 agent bipyridyl were given on GDs 12-15 at the threshold embryotoxic dose (1 mL, 25%
8 solution). On day 20, the dams were sacrificed and dams and fetuses examined as described
9 above. In addition to blood pH, partial pressure of carbon dioxide and partial pressure of
10 oxygen, acid metabolic products (not detailed), and true bicarbonates were reported for maternal
11 and fetal blood. A review of the data from these reports indicates there may be an overlap of the
12 study groups. Neither paper presents the entire data set; thus, for transparency and brevity, the
13 following text discusses the combined findings from both studies as if they were a single study
14 Formaldehyde exposure did not affect such indicators of pregnancy outcome as number
15 of corpora lutea, implantation and resorption sites, and live and dead fetuses, all of which were
16 unchanged (Senichenkova and Chebotar, 1996; Senichenkova, 1991). Although fetal weight was
17 slightly increased by formaldehyde exposure, 2.35 versus 2.24 g in controls (p < 0.001), neither
18 fetal length nor bone length were changed (femur and humerus) (Senichenkova and Chebotar,
19 1996; Senichenkova, 1991). Often, increased fetal weight is the result of early physical
20 development, and other signs of development, such as ossification, would be expected to be
21 enhanced as well. The average number of bone centers per limb was increased by formaldehyde
22 exposure from 2.45 and 2.66 to 2.78 and 2.91 in controls for metacarpal and metatarsal bone
23 centers, respectively (p < 0.05) (Senichenkova, 1991); these findings were consistent with
24 increased growth and weight. In contrast, Senichenkova (1991) reported a decrease in the
25 number of embryos with ossification centers in the hyoid bone (100% in controls versus 91% for
26 formaldehyde exposure, p < 0.05), consistent with the results of Saillenfait et al. (1989) and
27 Martin (1990). However, litter size, a factor influencing fetal weight, was not provided, and it is
28 unclear if Senichenkova (1991) took litter size into account in the analysis.
29 Senichenkova and Chebotar (1996) reported increased blood acidosis and decreased
30 blood alkaline reserves (bicarbonates and total CO2) in formaldehyde-treated dams and their
31 embryos (p < 0.05). However, this finding should be considered in light of the fact that chronic
32 blood acidosis may increase bone remodeling and decrease bone density in adults. It is unknown
33 if the reported blood acidosis could reduce ossification rates in developing embryos. A better
34 understanding of exposure conditions and the acid metabolic products measured is needed to
35 determine the biological relevance of the reported changes in blood acid balance.
This document is a draft for review purposes only and does not constitute Agency policy.
4-378 DRAFT—DO NOT CITE OR QUOTE
-------
1 Iron deficiency, induced by injections of bipyridyl (an iron-chelating agent), was found to
2 be fetotoxic. Iron-deficient dams with no formaldehyde exposure had higher rates of
3 postimplantation death than controls (12.6 ± 5.5 versus 4.8 ± 1.3%). Formaldehyde exposure in
4 conjunction with iron deficiency increased postimplantation death to 23.1 ± 5.9%. Fetal weight
5 and litter size were not reported for the bipyridyl treatment groups, but bipyridyl treatment in
6 conjunction with formaldehyde resulted in a decreased number of metatarsal bone centers (2.21
7 ±0.12 versus 2.72 ± 0.08 in controls;/? < 0.001). This decrease was also significant when
8 compared with formaldehyde or bipyridyl alone (p < 0.02). However, all pregnancy outcome
9 parameters were not reported for the bipyridyl treatment.
10 Fetal anomalies were reported after formaldehyde exposure and were increased by iron
11 deficiency. The incidence of litters with internal organ anomalies was increased from 1.4% in
12 controls to 14.2% in formaldehyde-treated dams (Senichenkova, 1991). Undescended testes
13 were the predominant anomaly described: 20.8% in litters from formaldehyde-treated dams
14 versus 1.2% in controls (p < 0.05) (Senichenkova, 1991). Similar findings were reported by
15 Senichenkova and Chebotar (1996). Bipyridyl treatment in conjunction with formaldehyde
16 exposure increased the overall incidence of fetal anomalies (13.8 ± 2.1% in controls versus 6.6 ±
17 1.8% with iron deficiency alone) (Senichenkova and Chebotar, 1996). However, there are
18 discrepancies between the two papers in the reporting of the anomalies, and it is unclear whether
19 the experimental groups overlap between papers, where some parameters are identical (which
20 would lead to double counting of the same animal, including identical SDs) and others are
21 different. Additionally, the reporting is unclear with respect to the basis of the incidence rates
22 reported (for example, overall incidence versus incidence within litter or incidence of litters with
23 anomalies). Unclear reporting, together with some of the uncertainties regarding exposure
24 conditions, suggests that the data may be of limited quality to support risk assessment.
25 In the second phase of the studies, pups were delivered and postnatal development was
26 assessed (Senichenkova, 1991). Eruption of the upper and lower incisors was delayed in pups
27 from formaldehyde-treated dams, occurring on PND 14 versus PND 12 in controls (p < 0.01).
28 All other measures of physical postnatal development were unchanged by formaldehyde. To
29 evaluate postnatal functional outcomes following in utero exposure to formaldehyde, an open
30 field test was conducted in juvenile rats on 3 consecutive days (PNDs 40-42). Juvenile rats from
31 formaldehyde-treated dams exhibited increased mobility (crossed squares), rearings, and
32 defecations/urinations compared with control rats on the second and third open field tests
33 (p < 0.05). There were no differences seen in the maze-learning test assessed in mature offspring
34 of formaldehyde-treated dams (Senichenkova, 1991).
This document is a draft for review purposes only and does not constitute Agency policy.
4-379 DRAFT—DO NOT CITE OR QUOTE
-------
1 Additional assessments of formaldehyde exposure on neurological development are
2 described above in the section on neurological and behavioral toxicity in animal studies (see
3 Section 4.2.6). In brief, studies conducted by Sarsilmaz et al. (2007) and Asian et al. (2006)
4 exposed 10 neonatal male Wistar rats/group to 0, 6, or 12 ppm (0, 7.36, or 14.7 mg/m3)
5 formaldehyde 6 hours/day, 5 days/week for 30 days. At that time, five rats/group were killed
6 and subjected to neuropathological examination; the remaining rats were maintained until
7 PND 90, at which time they were killed and evaluated. Asian et al. (2006) examined the number
8 and volume of granular cells in the hippocampal dentate gyrus, while Sarsilmaz et al. (2007)
9 examined the size and number of the pyramidal cells in the cornu ammonis of the hippocampus.
10 In both studies, lower numbers of cells were observed in both treated groups at PND 90 as
11 compared with PND 30. Although the effects of treatment on the volume and number of
12 granular and pyramidal cells were somewhat inconsistent, a significant decrease in the number of
13 neurons in the pyramidal cell layer of the hippocampal cornu ammonis was observed at both
14 PNDs 30 and 90 (Sarsilmaz et al., 2007).
15 One other study reported effects on nervous system function following exposure to
16 formaldehyde during postnatal development. An abstract by Weiler and Apfelbach (1992)
17 described a study in which juvenile rats (strain not specified) were exposed to 0.25 ppm
18 (0.31 mg/m3) formaldehyde from PNDs 30-160 or adult rats were exposed to 0.5 ppm
19 (0.62 mg/m3) formaldehyde for 90 days. Decreased olfactory sensitivity (i.e., increased olfactory
20 thresholds) was observed and was greater when the exposure was initiated in the young rats, as
21 compared with the adults.
22 Evaluation of offspring following prenatal, perinatal, and/or juvenile inhalation exposures
23 to formaldehyde have also been reported by Kum et al. (2007), Sandikci et al. (2007), and
24 Songur et al. (2005). Kum et al. (2007) exposed female Sprague-Dawley rats (six dams/group)
25 and their offspring to 0 or 6 ppm (0 or 7.4 mg/m3) formaldehyde for 8 hours/day in separate
26 groups with exposures starting on GD 1, on postparturition day 1, or in offspring at 4 weeks of
27 age and continuing for 6 weeks. In another group, exposures were initiated in adult rats. Mean
28 body and liver weights were significantly decreased in the offspring exposed in utero and in
29 early postnatal life, and mean liver weights were also significantly decreased in rats with juvenile
30 exposures. However, neither body weight nor liver weight was affected in the group with
31 exposure initiating at an adult age, suggesting a life-stage-related susceptibility to formaldehyde-
32 induced hepatic toxicity. Evaluation of biomarkers of oxidative stress revealed significantly
33 increased catalase (CAT) and MDA in the livers of offspring that were exposed prenatally,
34 significantly decreased GSH levels in the livers of offspring that were exposed in early postnatal
35 life, and significantly decreased SOD levels in the livers of offspring that were exposed starting
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 80 DRAFT—DO NOT CITE OR QUOTE
-------
1 at 4 weeks of age. No biomarkers were altered in the livers of rats exposed to formaldehyde only
2 as adults.
3 A similar study design was used by Sandikci et al. (2007) to examine the effects of 0 or
4 6 ppm (0 or 7.4 mg/m3) formaldehyde on bronchus associated lymphoid tissue (BALT)
5 following pre- and perinatal, juvenile, or adult exposures of 6 weeks duration in Sprague-Dawley
6 rats (six/group). The presence of the lysosomal enzyme alpha-naphthylacetate esterase (ANAE)
7 served as a marker of T-lymphocytes in peripheral blood and tissue sections. Significant
8 increases in ANAE-positive T-lymphocytes were found in BALT in all but the in utero exposed
9 groups as compared with control; this outcome is consistent with the postnatal development of
10 BALT in the rat. In peripheral blood, ANAE-positive lymphocyte ratios were significantly (p <
11 0.001) increased as compared with controls at all life stages tested.
12 Songur et al. (2005) examined the effect (and reversibility) of formaldehyde exposures
13 during the early postnatal period on zinc, copper, and iron levels and activity of SOD in the lung
14 tissue of Wistar rats. Litters (12-14/group) were exposed to 0, 6, or 12 ppm (0, 7.4, or
15 14.9 mg/m3) formaldehyde 6 hours/day, 5 days/week for 30 days. Trace element and
16 biochemical analyses were conducted on PND 30 or 90. Decreased SOD activity, decreased
17 levels of copper and iron levels, and increased zinc levels were observed in the lungs of treated
18 groups following 30 days of treatment and at 90 days (i.e., 60 days post-treatment). Survival was
19 not affected in neonatal rats. Clinical observations during treatment included evidence of
20 respiratory irritation and toxicity. Body weight and food and water consumption were also
21 nonsignificantly decreased as compared with controls.
22 There are several reports in the literature regarding formaldehyde effects after inhalation
23 exposure on the male reproductive system in animals (Galalipour et al., 2007; Zhou et al., 2006;
24 Ozen et al., 2005, 2002; Sarsilmaz et al., 1999; Woutersen et al., 1987; Maronpot et al., 1986;
25 Guseva, 1972). The earliest report examined the effect of simultaneous exposures to
26 formaldehyde from air and water (Guseva, 1972). Male rats (n = 12, strain not specified) were
27 coexposed to formaldehyde in air and drinking water 4 hours/day, 5 days/week for 6 months.
28 There were three exposure levels in the experiment of different air and drinking water
29 concentrations: (1) 0.41 ppm (0.5 mg/m3) formaldehyde in air and 0.1 mg/L in water; (2) 0.20
30 ppm (0.25 mg/m3) formaldehyde in air and 0.01 mg/L in water; or (3) 0.10 ppm (0.12 mg/m3)
31 formaldehyde in air and 0.005 mg/L in water. Reproductive function was assessed by mating
32 two females per male. The time for the onset of pregnancy and the number of pregnancies per
33 treatment group were recorded. A subset of dams was sacrificed on GD 20 of pregnancy, and
34 the number and weight of fetuses was determined. Postnatal development of the remaining dams
35 was tracked (e.g., times of eye opening and development of hair coat). Nucleic acid levels were
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 81 DRAFT—DO NOT CITE OR QUOTE
-------
1 determined in the testes of formaldehyde-exposed rats. Gonadotropic response was assessed by
2 injecting an emulsion of pituitaries from exposed male rats into unexposed infantile females and
3 measuring the weight ratios of the uterus and ovaries. Formaldehyde exposure reduced nucleic
4 acid levels in testes to 88 and 92% of controls in groups 1 and 2, respectively. No other
5 formaldehyde-induced differences were found.
6 Woutersen et al. (1987) and Maronpot et al. (1986) examined tissue sections from testes
7 and ovaries of exposed animals as part of studies primarily addressing respiratory tract toxicity
8 (see Section 4.2.1.2.2.4 for complete study details). Maronpot et al. (1986) exposed female and
9 male B6C3F1 mice to 0, 2, 4, 10, 20, and 40 ppm (0, 2.45, 4.91, 12.3, 24.5, and 49.1 mg/m3)
10 formaldehyde 6 hours/day, 5 days/week for 13 weeks. Decreased weight gain due to
11 formaldehyde exposure was seen in both male and female mice. Additionally, there was 80%
12 mortality at the highest exposure (40 ppm). The authors reported endometrial hypoplasia and
13 lack of ovarian luteal tissue in females exposed to 40 ppm, but no compound-related changes
14 were observed in testes sectioned and viewed by light microscopy.
15 In a study by Appleman et al. (1988), male Wistar rats (40/group) with undamaged or
16 damaged (via bilateral intranasal electrocoagulation) nasal mucosa were exposed for 13 or
17 52 weeks to 0, 0.1, 1, or 10 ppm (0, 0.124, 1.24, or 12.4 mg/m3) formaldehyde 6 hours/day,
18 5 days/week. At study termination, mean body weight was decreased, but relative testes weight
19 was increased in the 10 ppm group (interpreted as a nonadverse outcome that was associated
20 with the decreased body weight). No treatment-related histopathologic findings were reported
21 for male reproductive organs (although it is not clear to what extent they were evaluated since
22 the primary focus of the study was on the nasal epithelium).
23 Following up on earlier reports of decreased Leydig cell quality in rats administered IP.
24 injections of formaldehyde (Chowdhury et al. [1992], described in Section 5.2.1.8.3), Sarsilmaz
25 et al. (1999) studied the effects of formaldehyde inhalation on Leydig cells. Adult male Wistar
26 rats (30) were exposed to 0, 10, or 20 ppm (0, 12.3, or 24.6 mg/m3) formaldehyde 8 hours/day,
27 5 days/week for 4 weeks. Animals were observed daily and weighed weekly. Rats were
28 sacrificed on day 29 and autopsied, and testes were weighed, fixed, and sectioned for histologic
29 examination. Signs of irritation from formaldehyde exposure were noted (frequent eye blinking,
30 excessive licking, increased frequency of nose cleaning, interrupted breathing, and sneezing).
31 Body weight gain was reduced by formaldehyde exposure from 17.7% gain in control rats to
32 4.66 and 2.63% in rats exposed at 10 and 20 ppm, respectively (p < 0.001). As shown in
33 Table 4-60, relative testes weights were unaffected (reported as proportions but more likely to be
34 percentages), although trends and numerical differences were similar to those reported by Ozen
35 et al. (2002). Sarsilmaz et al. (1999) found that both Leydig cell quantity and the percentage of
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 82 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
cells with normal nuclei were reduced by formaldehyde treatment. Although the dose-dependent
reduction in Leydig cell quantity was statistically significant at both exposure levels, the study
authors considered the data to be within the normal range.
Table 4-60. Formaldehyde effects on Leydig cell quantity and nuclear
damage in adult male Wistar rats
Inhalation
exposure"
Control
lOppm
20ppm
Relative testes
weightb'c'd
0.93 (0.03)
0.92 (0.06)
0.89 (0.06)
Leydig cell
quantity c'e'f
47.27 (7.8)
45.04 (7.8)h
44.36 (7.5)1
Appearance of nucleus6'8
Normal
98
92
76
Pyknotic
2
2
9
Karyorectic
0
4
10
Karyolytic
0
2
5
aRats were exposed 8 hours/day, 5 days/week for 4 weeks.
bStated to represent the ratio of the last day's testicle weight to the body weight but more likely to be the percent
of body weight.
°Cells within 100 defined areas.
dn=10.
Tor each exposure group, 100 defined locations were assessed.
fn=100.
8Values presented as percentage of cells.
hDifferent from control (p < 0.05), as reported by the authors.
'Different from control (p < 0.01), as reported by the authors.
Source: Sarsilmaz et al. (1999).
It was hypothesized that decreased Leydig cell quality may have been the result of
oxidative stress induced by formaldehyde exposure. Ozen et al. (2002), in the same laboratory,
investigated changes in testicular iron, copper, and zinc levels as measures of oxidative stress
and damage. Adult male albino Wistar rats (seven/group) were exposed at 0, 10, or 20 ppm (0,
12.2, or 24.4 mg/m3) formaldehyde 8 hours/day, 5 days/week for either 4 or 13 weeks. Rats
were observed daily and weighed once a week. Rats were sacrificed at the end of the exposure
period and autopsied, and the testes were removed and weighed. Zinc, copper, and iron levels
were determined in testes tissue by using atomic absorption spectrophotometry. Both weight
gain and relative testes weight were decreased in a concentration-dependent and duration-
dependent manner (see Table 4-61). Both zinc and copper levels in rat testes were reduced in a
concentration- and duration-dependent manner by formaldehyde exposure. For example, zinc
was reduced from 277 to 107 mg/kg after a 4-week x 20 ppm exposure and from 260 to
95 mg/kg after a 12-week x 20 ppm exposure (p < 0.001) (see Table 4-61). Iron levels in testes
were increased from 30 to 39 mg/kg after a 4-week 20 ppm exposure (p < 0.01) and from 33 to
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 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
58 mg/kg after 12 weeks at 20 ppm (p < 0.05). The authors suggested that alterations in trace
element levels in the testes were consistent with oxidative damage and may have contributed to
changes in Leydig cell function. These researchers also reported alterations in trace metals in
lung tissue from Wistar rats exposed to formaldehyde 8 hours/day, 5 days/week for 4 or
13 weeks. Iron levels were increased at 5 ppm for 13 weeks and 10 ppm for either 4 or
13 weeks. Zinc levels decreased for all formaldehyde exposures. In both cases, the authors
attributed elevated iron levels to oxidative stress. In addition to citing the role of zinc as a
cofactor of cytoplasmic Cu-Zn-SOD, the authors suggested that zinc may have been consumed
by increased FALDH activity. Although oxidative stress and increased FALDH activity may be
relevant to the POE and therefore impact the lung, it is less clear how these changes would occur
in the testes. Taken together, the reports of Ozen et al. (2002) and Sarsilmaz et al. (1999)
suggested a LOAEL of 10 ppm 8 hours/day for 4 weeks for changes in Leydig cell quantity and
quality, decreased testes weight, and changes in trace metal content (zinc, copper, and iron).
Table 4-61. Formaldehyde effects on adult male albino Wistar rats
Inhalation
exposure"
Weight gain
(%)
Testes weight
(%)
Zinc
(mg/kg)
Copper
(mg/kg)
Iron
(mg/kg)
4 Weeks
Control
10 ppm
20 ppm
19.1 (2.7)
5.8 (2.4)b
2.4 (0.6)b
0.94 (0.03)
0.92 (0.02)c
0.91 (0.01)c
277 (16)
132 (8.9)b
107 (6.9)b
6.4 (0.42)
4.2 (0.33)b
3.3 (0.27)b
30 (2.7)
35 (2.8)d
39(3.1)d
13 Weeks
Control
10 ppm
20 ppm
55.9 (2.3)
34.7 (3.5)b
20.8 (1.4)b
0.91 (0.01)
0.84 (0.03)b
0.82 (0.03)b
269 (15)
112(8.1)b
95 (6.4)b
6.0 (0.34)
3.6 (0.30)b
1.9(0.17)b
33 (2.6)
52 (3.5)b
58 (3.0)b
""Formaldehyde exposure was 8 hours/day, 5 days/week for either 4 or 13 weeks. Values are means ± SDs of
seven animals..
seven animals..
'Different from control, p < 0.001, as calculated by the authors.
Different from control, p < 0.05, as calculated by the authors.
Different from control,/) < 0.01, as calculated by the authors.
Source: Ozen et al. (2002).
In another study that assessed testicular toxicity (Ozen et al., 2005), male Wistar rats
(18/group) were exposed by inhalation to 0, 5, or 10 ppm (0, 6.2, or 12.4 mg/m3) formaldehyde
8 hours/day, 5 days/week for 91 days. In-life observations in exposed rats included clinical signs
of respiratory irritation and decreased mean food and water consumption. At study termination,
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 84 DRAFT— DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
serum testosterone levels and mean seminiferous tubule diameters were significantly decreased
from control in a dose-responsive manner (see Table 4-62). Immunohistochemical staining of
testis tissues showed increased localization of heat shock protein (Hsp) 70 in the cytoplasm of
spermatogonia, spermatocytes, and spermatids of treated animals compared with controls (not
shown here).
Table 4-62. Formaldehyde effects on testosterone levels and seminiferous
tubule diameters in Wistar rats following 91 days of exposure
Inhalation
exposure"
Control
lOppm
20ppm
Testosterone levels
(ng/dL)
n = 6
406.54 ± 16.82
244.01 ± 23. 86b
141.30 ±08.56b
Seminiferous tubule
diameters
(um)
w = 100
259.22 ±16. 18
236.17 ±13.09C
233.24 ±10.13C
"Formaldehyde exposure was 8 hours/day, 5 days/week for 91 weeks. Values are means
± SEMs.
bDifferent from control, p < 0.0001, by one-way ANOVA, as calculated by the authors.
'Different from control,/) < 0.001, by one-way ANOVA, as calculated by the authors.
Source: Ozenetal. (2005).
Zhou et al. (2006) investigated the effect of formaldehyde on the testes and the protective
effect of vitamin E against oxidative damage by formaldehyde in adult male rats. In this study,
adult male Sprague-Dawley rats (10/group) were treated for 2 weeks in the following groups:
(1) control rats were administered physiological saline by oral gavage, (2) rats were administered
physiological saline by gavage and exposed to 10 mg/m3 (8.05 ppm) formaldehyde by inhalation
for 12 hours/day, and (3) rats were administered daily gavage doses of 30 mg/kg vitamin E and
exposed to 10 mg/m3 (8.05 ppm) formaldehyde by inhalation for 12 hours/day. Formaldehyde
treatment resulted in significantly decreased (p < 0.05) mean testis weight. Histopathologic
findings in treated rats included atrophy of seminiferous tubules, decreased spermatogenic cells,
and seminiferous cells that were "disintegrated" and shed into the lumina, which was
azoospermic. Interstitial tissue was edematous with vascular dilatation and hyperemia. In the
formaldehyde-treated group, epididymal sperm count and percentage of motile sperm were
significantly decreased, and the percentage of abnormal sperm was increased (p < 0.05), as
compared with control. Evaluation of biochemical markers in testes tissue showed the activities
of testicular SOD, GPX, and GSH were decreased; MDA levels were significantly increased as
This document is a draft for review purposes only and does not constitute Agency policy.
4-385 DRAFT—DO NOT CITE OR QUOTE
-------
1 compared with control. All observed effects of formaldehyde treatment (i.e., decreased testes
2 weight, biochemical alterations, histopathologic effects, and sperm count, motility, and
3 morphology findings) were attenuated by administration of 30 mg/kg-day vitamin E.
4 In a study by Golalipour et al. (2007), testicular changes of increasing severity with
5 increasing duration were reported. A total of 28 Wistar rats, aged 6-7 weeks old, were divided
6 into four groups including three formaldehyde treatment groups (4 hours of exposure/day, 4
7 days/week for 18 weeks; 2 hours of exposure/day, 4 days/week for 18 weeks; 4 hours of
8 exposure/day, 2 days/week for 18 weeks) and one untreated control. The three formaldehyde-
9 treated groups were exposed via inhalation to formaldehyde. The mean concentration of
10 formaldehyde vapor, based on three measurements during the study (stated as the beginning,
11 during, and end of the study period), was reported as 1.5 ppm. At the end of the study period,
12 the rats were sacrificed by ether anesthesia and subsequent cervical dislocation. The left testis
13 was dissected from each rat and a specimen was taken from each testis. Tissues were fixed,
14 embedded, sectioned (at 4 um), and stained with hematoxylin and eosin. Using a light
15 microscope, a histopathological examination was performed on the testes tissues, including
16 morphometric evaluation of the diameter and height of 20 seminiferous tubules/testis.
17 Golalipour et al. (2007) reported an formaldehyde exposure frequency (or duration)-dependent
18 increase in testicular germ cells and seminiferous tubule defects. In the most frequent duration
19 treatment group, a severe decrease in germ cells in >85% of the seminiferous tubules and
20 arrested spermatogenesis were observed. In the mid-level frequency of duration treatment group,
21 a decrease in the number of germ cells and an increased thickness of the basement membrane of
22 75% of the tubules was observed. In the lowest level duration treatment group, a disruption in
23 the Sertoli and germinal cell arrangement, and increased spacing between germ cells was
24 observed. Further, the seminiferous tubule diameter (STD) and seminiferous epithelial height
25 (SEH) was most decreased among the treatment groups (exhibiting the greatest decrease in the
26 group with the greatest hours and days of exposure) compared to the control (see Table 4-63).
27 The results of this study are consistent with the findings of other studies of male reproductive
28 system outcomes with inhalation formaldehyde exposure (e.g., Ozen et al., 2005 and Zhou et al.,
29 2006).
30 Xing et al. (2007) also studied the effects of formaldehyde on sperm development and
31 reproductive capacity in adult male mice. In this study, male mice (12/group, strain not
32 specified) were exposed to 0, 21, 42, or 84 mg/m3 (0, 16.9, 33.8, or 67.6 ppm) formaldehyde via
33 inhalation for 13 weeks at 2 hours/day, 6 days/week. The males were mated to untreated females
34 in a dominant lethal protocol, and sperm morphology was assessed at study termination. The
35 percent abnormal sperm was increased significantly (p < 0.05) in all treated groups, as was the
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 86 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
10
11
12
13
14
15
16
17
18
19
rate of resorptions (p < 0.01) (see Table 4-64). The mean number of live fetuses/litter was
decreased in all treated groups, with statistical significance achieved at 84 mg/m3. Although this
study did not assess the number of corpora lutea per dam, thereby precluding the calculation of
preimplantation loss, it is nevertheless indicative of formaldehyde-induced sperm morphology
changes and dominant lethal effects in male mice.
Table 4-63. Effects of formaldehyde exposure on seminiferous tubule
diameter and epithelial height in Wistar rats following 18 weeks of exposure
Inhalation
exposure"
Control
1.5ppm, 4h/d, 4 d/w
1.5ppm, 2h/d, 4 h/w
1.5ppm, 2h/d, 2 d/w
Seminiferous tubule
diameters
(urn)
M = 7
252.12 ±4.82
204.55 ±3.29b
232.45 ±2.42b
238.94 ±4.37b
Seminiferous tubule
height
(urn)
M = 7
82.77 ± 2.00
65.26 ± 1.43b
69.46 ± 1.78b
72.80 ±2.03b
a Values are means ± SEMs.
b Different from control,/) < 0.05, as calculated by the authors.
Source: Golalipour et al. (2007).
Table 4-64. Incidence of sperm abnormalities and dominant lethal effects in
formaldehyde-treated mice
Dose
(mg/m3)
0
21
42
84
Sperm abnormalities
Total abnormal
sperm heads
391
568
849
974
Aberration rate
(%)
3.53 ±0.98
5.48 ±1.45
6.15 ±1.36
9.24±2.13a
Reproductive capacity
Mean live fetuses/litter
11.00 ±1.01
10.67 ±1.16
9.63 ±2.83
9.04 ± 2.98 a
Resorption rate (%)
2.273
9.380b
10.390b
12.440 b
20
21
22
23
24
25
26
""Significantly different from controls (p < 0.05), as calculated by the authors.
bSignificantly different from controls (p < 0.01), as calculated by the authors.
Source: Xing etal. (2007).
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 87 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.2.7.2. Oral Exposure Studies Addressing Developmental and Reproductive Toxicity
2 No contemporary testing guideline studies, such as a prenatal developmental toxicity
3 study or two-generation reproductive toxicity study, have been performed by the oral route for
4 formaldehyde. However, a number of studies have evaluated developmental toxicity and
5 reproductive parameters in rats, mice, and dogs.
6 Hurni and Ohder (1973) tested the developmental toxicity of formaldehyde administered
7 as a 40% w/v solution containing 11-14% w/w methanol in 9-10 pregnant beagle dogs that
8 received the compound in their diet on GDs 4-56. Commercial grade formaldehyde (as a 40%
9 solution) was sprayed on the pellets prior to feeding. Each animal was allotted a diet of 300 g of
10 chow (reduced to 200 g 1 week prior to term) that was promptly consumed (within
11 5-10 minutes) before the formaldehyde volatilized appreciably. The concentrations of
12 formaldehyde in the chow were 0, 125, or 375 ppm, equivalent to doses of 0, 3.1, or
13 9.3 mg/kg-day, respectively. Dams were allowed to deliver normally and weight gain, gestation
14 length, number of litters, litter size, number of live pups, number of pups surviving through
15 weaning, and pup weights weekly for the first 8 weeks were monitored as indices of the potential
16 reproductive/developmental toxicity of formaldehyde. There were no formaldehyde-related
17 effects in any of the parameters other than progressive pup weights, which were lower by group
18 in litters of dams exposed to formaldehyde (see Table 4-65). A developmental impact of
19 formaldehyde was evident in this strain of dog under the conditions of the experiment. At birth,
20 mean pup body weights were 4 and 8.4% less than control for the low- and high-dose groups,
21 respectively; at 8 weeks of age, the pup weight decrements were 8.3% for the low dose and
22 12.5% for the high dose, as compared with control, and established a LOAEL of 125 ppm. The
23 contribution of methanol, which is a developmental toxin (Deglitz et al., 2004; Rogers et al.,
24 2004) to these outcomes is not known. No internal or skeletal malformations were observed in
25 any of the 264 live-born and 20 still-born pups.
26 Marks et al. (1980) conducted a developmental toxicity study of formaldehyde in CD-I
27 mice in which 29-35 pregnant animals were gavaged on GDs 6-15 with aqueous formaldehyde
28 (containing 10-15% methanol) at 74, 148, and 185 mg/kg-day. Seventy-six controls were
29 gavaged with water alone. All dams were sacrificed on GD 18, and the numbers of implantation
30 sites in each uterine horn were counted. The high dose of formaldehyde was toxic to the dams,
31 as indicated by the deaths of 22 of 34 mice before GD 18. Thus, the dose of 148 mg/kg-day was
32 a NOAEL for maternal toxicity in this study. However, it is unclear to what extent an estimated
33 concurrent dose of up to 75 mg/kg-day methanol may have contributed to this toxic response.
34 To assess the developmental toxicity of formaldehyde, live fetuses were weighed individually,
35 sexed, and examined for external, visceral, and skeletal malformations. Fetuses of surviving
This document is a draft for review purposes only and does not constitute Agency policy.
4-388 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
Table 4-65. Body weights of pups born to beagles exposed to formaldehyde
during gestation
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Time (weeks)
0
1
2
3
4
5
6
7
8
Formaldehyde concentration in chow (ppm)
0
125
375
Average body weight (g)
321
547
818
1,078
1,264
1,601
2,020
2,449
2,957
308
526
755
987
1,247
1,512
1,816
2,263
2,712
294
467
706
944
1,166
1,429
1,741
2,145
2,587
Source: Hurni and Ohder (1973).
high-dose dams and of those of other groups did not show an increased incidence of
malformations. Therefore, Marks et al. (1980) concluded that formaldehyde did not induce fetal
abnormalities and that the 185 mg/kg-day dose level was a NOAEL for the developmental
toxicity of formaldehyde, nor were fetotoxic effects of methanol apparent under the study
experimental conditions.
Seidenberg and Becker (1987) and Seidenberg et al. (1986) included formaldehyde
(purity not indicated) in a survey of the behavior of potential toxicants in a developmental
toxicity screening assay (Chernoff and Kavlock, 1982). The protocol featured the administration
of a borderline toxic dose to 26-30 pregnant ICR/SIM mice on GDs 8-12. Dams were allowed
to deliver, and the neonates were examined, counted, and weighed on PNDs 1 and 3. The
selected formaldehyde dose of 540 mg/kg-day was fatal for 1 1/30 dams, but the average weight
gain among surviving dams was little changed compared with controls (3.9 ± 2.3 versus
4.0 ±1.0 g). Similarly, there were no changes in perinatal responses in the neonates of exposed
dams compared with controls. For example, the average values for number of neonates/litter,
percent survival, and fetal weights on PNDs 1 and 3 were closely similar to those of controls.
Evidence of toxicity to the male reproductive system has been observed following oral
administration of formaldehyde in a 40% w/v solution containing 1 1-14% w/w methanol.
Cassidy et al. (1983) administered single oral doses of 100 or 200 mg/kg to five male Wistar
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 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
rats/group. Testes from these animals and 20 controls were excised and examined for
spermatogenic abnormalities 11 days after dosing. A significant (19%) increase in testicular
sperm head counts was observed in rats exposed to 200 mg/kg-day formaldehyde as compared
with controls (see Table 4-66). The percentage of abnormal sperm heads was also significantly
increased (5%) in the 200 mg/kg-day dose group compared with controls. These data suggest
that formaldehyde can induce morphologic abnormalities in the germ cells of male experimental
animals at dose levels that did not significantly affect testis weights. The contribution of
methanol to these outcomes is unknown.
Table 4-66. Testicular weights, sperm head counts, and percentage incidence
of abnormal sperm after oral administration of formaldehyde to male Wistar
rats
Dose (mg/kg)
0
100
200
Mean testes weight (g)
3.30
3.27
3.16
Mean sperm heads x
106/g testis
175
166
209a
Abnormal sperm heads (%)
4.76
5.22
9.77a
""Significantly different from controls (p < 0.001), as calculated by the authors.
Source: Cassidy et al. (1983).
Postmortem evaluation of the reproductive organs was conducted in a number of oral
studies that ranged between 4 weeks and 2 years in duration (Til et al., 1989, 1988; Tobe et al.,
1989; Johanssen et al., 1986). Johannsen et al. (1986) administered 0, 50, 100, or 150 mg/kg-day
formaldehyde in the drinking water to Sprague-Dawley rats (15/sex/group) for 91 days and 0, 50,
75, or 100 mg/kg-day formaldehyde in basal diet to beagle dogs (4/sex/group) for 91 days; the
study reported no treatment-related effects on absolute or relative gonad weights or
histopathology for either species. In a 4-week drinking water study conducted by Til et al.
(1988), formaldehyde was administered to Wistar rats (10/sex/treated group) at nominal levels of
0, 25, and 125 mg/kg-day; gonad organ weights and histopathology were not affected by
treatment. Tobe et al. (1989) conducted a chronic (24-month) study in Wistar rats
(20/sex/group), with drinking water concentrations of 0, 0.02, 0.1, or 0.5%. According to the
study report, gonad weights were measured and histopathology was conducted, but no treatment-
related findings were noted. In a chronic (105-week) study (Til et al., 1989) in Wistar rats
(70/sex/group), formaldehyde was administered in the drinking water at mean actual levels of 0,
1.2, 15, or 82 mg/kg-day to males and 0, 1.8, 21, or 109 mg/kg-day to females; serial sacrifices
This document is a draft for review purposes only and does not constitute Agency policy.
4-390 DRAFT—DO NOT CITE OR QUOTE
-------
1 were conducted at 53, 79, and 105 weeks of study. At study termination (105 weeks), mean
2 testes weights were 30% increased (p < 0.01) in high-dose males as compared with controls, and
3 histopathology evaluation revealed Leydig cell tumors in treated males (incidences of 0/50, 3/50,
4 3/50, and 2/50 for the control through high-dose groups, respectively; historical control tumor
5 incidence data were not provided). The study authors did not judge these findings to be
6 treatment related. By design, none of the subacute to chronic studies included measures of
7 reproductive function (e.g., estrous cyclicity, sperm measures, or reproductive performance).
8 With the exception of Til et al. (1989), detailed mean organ weight and histopathology incidence
9 data were not provided in the published reports, and Til et al. (1988) only included tumor (not
10 nontumor) data.
11
12 4.2.7.3. Intraperitoneal Studies Addressing Developmental and Reproductive Toxicity
13 Other studies in which formaldehyde was administered by IP. injection have confirmed
14 the potential effects on the male reproductive system.
15 Chowdhury et al. (1992) administered IP. injections of 0, 5, 10, or 15 mg/kg-day
16 formaldehyde to Charles foster adult male rats (10/group) for 30 days. On study day 31, blood
17 was collected for serum testosterone measurements and the rats were sacrificed. The testes were
18 removed, weighed, fixed in Bouin's solution, and processed for histopathology. The study
19 authors reported adverse findings in all treated groups, including significant (p < 0.01) mean
20 body weight gains, serum testosterone levels, and testes weights. Histopathologic evaluation
21 revealed normal spermatogenic processes and Leydig cells in control animals. However, in
22 treated rats, gradual cellular degeneration in seminiferous tubules and in Leydig cells was
23 observed. Marked nuclear damage was noted in the 10 and 15 mg/kg-day groups, with
24 significantly (p < 0.001) decreased Leydig cell populations and nuclear diameters observed in all
25 treated groups. Additionally, a decrease in 3p-A5-hydroxy steroid dehydrogenase was noted in
26 the Leydig cell region of treated rat testes.
27 In a 30-day study performed by Majumder and Kumar (1995), 10 mg/kg-day
28 formaldehyde was administered IP. to eight male Wistar rats. All animals were sacrificed at
29 term, and the testes, prostate, seminal vesicles, and epididymides were excised and weighed.
30 With the use of methodologies that were not described in the report other than by reference to the
31 Laboratory Manual for the Examination of Human Semen and Sperm-Cervical Mucus
32 Interaction (WHO, 1987), sperm counts, motility, and viability were compared with those of 10
33 controls (injected IP. with water alone). As shown in Table 4-67, striking reductions in sperm
34 count and motility were noted in formaldehyde-treated rats compared with controls. Sperm
This document is a draft for review purposes only and does not constitute Agency policy.
4-391 DRAFT—DO NOT CITE OR QUOTE
-------
1 viability was also significantly reduced by formaldehyde treatment, though to a lesser overall
2 extent than sperm count and motility.
4
5
6
Table 4-67. Effect of formaldehyde on spermatogenic parameters in male
Wistar rats exposed intraperitoneally
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Parameters
Sperm count (106/mL)
Sperm viability (%)
Sperm motility (%)
Control (n = 10)
46.30 ±5.01
87. 10 ±0.83
75.00 ± 10.90
Treated (n = 8)
20.40 ±2.01a
72.60 ±2.32a
22.00 ± 6.40a
""Significantly different from controls (p < 0.0001), as calculated by the authors.
Source: Majumder and Kumar (1995).
Majumder and Kumar (1995) also carried out an in vitro experiment in which sperm from
normal rats were incubated with different concentrations of aqueous formaldehyde at
concentrations ranging from 125 pg/mL to 2.5 |ig/mL. Viability of control sperm remained close
to 80% for a period of 1 hour, whereas the presence of formaldehyde dose-dependently reduced
viability. Thus, only 50% spermatozoa were viable for 30 minutes in the presence of 5 ng/mL
formaldehyde compared with 6 minutes in the presence of 500 ng/mL. Sperm motility also was
sensitive to the presence of formaldehyde. Less than 10% of sperm was motile for 10 minutes in
the presence of 125 pg/mL formaldehyde. The authors of the study considered their data to be
good evidence that functional parameters of spermatozoa, such as viability and motility, can be
adversely affected by exposure to formaldehyde. Moreover, they suggested that the cumulative
effects of IP. administration of formaldehyde on the male rat reproductive system raise an alert
that formaldehyde might impair the reproductive health of males who are occupationally exposed
to the compound.
Odeigah (1997) conducted two short-term in vivo assays to examine sperm head
abnormalities and dominant lethal mutations. In the sperm assessment, five daily IP. injections
of 0, 0.125, 0.25, or 0.5 mg/kg formaldehyde were administered to male albino rats (six/group;
strain not specified). The rats were killed 3 weeks after the last injection, and epididymal sperm
counts and abnormalities were assessed. A dose-related decrease in sperm count was observed,
and significantly increased incidences of sperm head abnormalities were found at all treatment
levels (see Table 4-68).
This document is a draft for review purposes only and does not constitute Agency policy.
4-392 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-68. Incidence of sperm head abnormalities in formaldehyde-treated
rats
4
5
6
7
8
9
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 (mg/kg)
0
0.125
0.25
0.5
Total abnormal sperm
heads
90
184
436
514
Frequency (%) ± SEM
1.50±0.11
3.09 ± 0.16 a
7.27 ± 0.30 b
8.57 ± 0.33 b
""Significantly different from controls (p < 0.05), as calculated by the authors.
bSignificantly different from controls (p < 0.001), as calculated by the authors.
Source: Odeigah (1997).
In the dominant lethal assay (Odeigah, 1997), five daily IP. injections of 0, 0.125, 0.25,
or 0.5 mg/kg formaldehyde were administered to male rats (5 control rats, 12/treated group).
Subsequently, each male was mated with two untreated virgin females per week for
3 consecutive weeks. The females were killed 13 days after the midpoint of the mating period
and evaluated for live and dead uterine implants. In general, the number of live embryos was
decreased with treatment, and the number of dead implants was increased (see Table 4-69).
Additionally, there was a reduction in fertile matings in females mated 1-7 days after the males
had been treated. This study did not assess the number of corpora lutea and therefore precluded
the determination of preimplantation loss. Nevertheless, it is indicative of dominant lethal
effects on the male germ cells.
4.2.7.4. Dermal Exposure Studies Addressing Developmental and Reproductive Toxicity
In a study designed to assess the embryotoxic effects of dermal exposure to
formaldehyde, Overman (1986) applied 0.5 mL of a 37% formaldehyde solution directly to the
dorsal skin of female Syrian golden hamsters (four-six/group) on GDs 8, 9, 10, or 11 for 2
hours. To prevent grooming during the treatment period, the animals were anesthetized with
Nembutal. At the end of the 2-hour treatment period, the application site was washed thoroughly
to remove any remaining formaldehyde. The dams were terminated on GD 15 (i.e., one day
prior to expected delivery, since the typical gestation period for the Syrian golden hamster is
16-18 days). The fetuses were removed and fixed in either Bouin's solution or 95% ethanol for
visceral or skeletal evaluation, respectively. The uteri were examined for implantation sites.
Fixed fetuses were weighed, measured (crown-rump), and examined for external abnormalities;
fetuses that had been placed in Bouin's fixative were evaluated for visceral anomalies by using a
This document is a draft for review purposes only and does not constitute Agency policy.
4-393 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-69. Dominant lethal mutations after exposure of male rats to
formaldehyde
Dose
(mg/kg)
0
0.125
0.25
0.5
Time of
mating
(days)
0-21
1-7
8-14
15-21
1-7
8-14
15-21
1-7
8-14
15-21
Fertile
matings3 (%)
96.67 (29)
75.0(18)
79.17(19)
91.67 (22)
33.33 (8)
50.0 (12)
87.5 (21)
25.0 (6)
29.17(7)
83.33 (20)
Implants per
female3
(mean ± SE)
7.86 ± 0.2 (29)
7.18 ±0.3 (18)
7.38 ±0.5 (19)
7.68 ± 0.2 (22)
5.75 ± 0.3 (8)
6.60 ±0.2 (12)
7.25 ±0.4 (21)
5.05 ± 0.03 (6)
5.27 ±0.01 (7)
7.08 ± 0.04 (20)
Live embryos
per female
(mean ± SE)
7.43 ±0.3
5.95 ±0.2
6.30 ±0.5
6.89 ±0.3
2.05 ±0.3
3.91 ±0.2
6.63 ±0.3
1.10 ±0.5
1.50 ±0.6
5.79 ±0.4
Dead implants
per female
(mean ± SE)
0.43 ±0.8
1.23 ±0.5
1.08 ±0.3
0.79 ±0.5
3.70 ±0.4
2.69 ±0.2
0.62 ±0.5
3.95 ±0.22
3.77 ±0.28
1.29 ±0.17
Dominant
lethal
mutation
indexb
0
19.92
15.21
7.27
72.41
47.38
10.77
85.20
79.81
22.07
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
"Number of females with implants presented in parentheses.
bDominant lethal mutation index:
Index = 1 - (Live implants experiment group per female) x 100
(Live implants of control group per female)
Source: Odeigah (1997).
free-hand sectioning technique, and those that were placed in ethanol were macerated, stained,
and cleared for skeletal examination. In this study, the dams exhibited signs of dermal irritation
and irritability, but the author reported no treatment-related effects on maternal body weight
gain. The percent of resorption sites was increased (although not significantly) in treated litters
as compared with control (0, 4.2, 8.1, 4.6, and 3.2% resorbed implantation sites for control and
GDs 8, 9, 10, and 11 treatment groups, respectively). No treatment-related effects on fetal
weight, length, or visceral or skeletal development were observed.
4.2.7.5. Summary of Reproductive and Developmental Effects
Formaldehyde exposures up to 40 ppm 6 hours/day from GDs 6-15 or 6-20 did not
result in external or internal malformations (Martin, 1990; Saillenfait et al., 1989). Martin
(1990) reported delayed skeletal ossification and dose-dependent decreases in fetal body weight
at 5 ppm. Formaldehyde exposure at 40 ppm to pregnant female Sprague-Dawley rats reduced
fetal body weights in male and female progeny and in male pups of dams exposed to 20 ppm
This document is a draft for review purposes only and does not constitute Agency policy.
4-394 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde (Saillenfait et al., 1989). Based on these studies (see Table 4-70), the LOAEL for
2 developmental effects in rats is 5 ppm, with a NOAEL of 2 ppm for decreased fetal weight and
3 delayed skeletal ossification, based on inhalation exposures during GDs 6-20.
4 Developmental studies during earlier gestational windows of inhalation exposure to
5 formaldehyde have reported additional adverse health effects, including delayed ossification,
6 changes in relative organ weight, undescended testes, biochemical changes (e.g., ascorbic acid
7 and nucleic acids), and blood acidosis (Senichenkova and Chebotar, 1996; Senichenkova, 1991;
8 Kilburn and Moro, 1985; Gofmekler and Bonashevskaya, 1969; Gofmekler, 1968; Pushkina et
9 al., 1968). Kitaev et al. (1984) hypothesized that formaldehyde may affect reproductive function
10 by stimulating the hypothalamo-pituitary-gonadal (HPG) axis based on their observations of
11 increased ovary weight, increased number of ovulating cells, and changes in blood levels of
12 gonadotropins (LH and FSH). Evidence of preimplantation loss, which may be related to HPG
13 disruption, was observed in this study and by Sheveleva (1971). Additional studies are needed to
14 better understand developmental effects of formaldehyde exposure during early gestational
15 windows.
16 The prenatal developmental toxicity of oral and dermal exposures to formaldehyde has
17 not been thoroughly studied. Reductions in postnatal growth in beagle pups was observed by
18 Hurni and Ohder (1973) following in utero exposure to 125 ppm maternal dietary formaldehyde
19 during GDs 4-56 in beagle dogs. However, gavage dosing during gestation of mice to overtly
20 maternally toxic doses (Seidenberg and Becker, 1987; Marks et al., 1980) (see Table 4-71) and
21 dermal application during gestation to hamsters at a dose that caused dermal irritation and
22 irritability (Overman, 1986) did not result in any observed fetal toxicity (see Table 4-72).
23 Few studies identified effects on maternal toxicity or female reproductive capacity. As
24 summarized in Table 4-70, exposure of rat dams at 10-40 ppm formaldehyde during pregnancy
25 has been shown to result in significantly decreased weight gain (Martin, 1990; Saillenfait et al.,
26 1989; Kilburn and Moro, 1985). Maronpot et al. (1986) reported endometrial hypoplasia with a
27 lack of ovarian luteal tissue in female rats exposed at 40 ppm but not at 20 ppm. Changes in LH
28 and FSH levels were reported in dams exposed to 0.41 ppm formaldehyde by Kitaev et al.
29 (1984), establishing an unbounded LOAEL for maternal toxicity.
30 Studies designed to assess male reproductive system endpoints in rats following repeated
31 inhalation exposures to formaldehyde have shown concentration-dependent decreases in Leydig
32 cell number and quality, effects on seminiferous tubules, decreases in testes weight, alterations in
33 sperm measures, decreased testosterone levels, alterations in trace metals in the testes, and/or
34 dominant lethal effects (Zhou et al., 2006; Ozen et al., 2005, 2002; Zhou et al., 2006Sarsilmaz et
35 al., 1999) (see Table 4-73). Based on available studies, the LOAEL for changes in the male
This document is a draft for review purposes only and does not constitute Agency policy.
4-395 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-70. Summary of reported developmental effects in formaldehyde inhalation exposure studies
Species,
strain, sex
Rat, strain
NR, female
Rat, "albino"
strain NR,
female
Rat, strain
NR, male
«/Group
12
12
4
Dose; time of
treatment"
0,0.01, or 0.81 ppm
(reported as 0.0 12,
and 1 mg/m3)d;
continuous dosing
10-15 days prior to
mating and during
gestation
0,0.01, or 0.81 ppm
(reported as 0.0 12,
and 1 mg/m3)d;
continuous dosing
10-15 days prior to
mating and during
gestation
Inhalation and
drinking water
coexposure: 0;
0.10 ppm plus 0.005
mg/L water; 0.20
ppm plus 0.01 mg/L
water; or 0.41 ppm
plus 0.1 mg/L water;
all treatments
4 hours/day,
5 days/week for
6 months
Reported study findings'"
Maternal
At 0.01 and 0.81 ppm:
t pregnancy duration
(dose-dependent data not
shown)6
NE
No effects
Offspring
Fetuses:
At 0.01 and 0.81 ppm:
J, fetuses/dam (dose dep., data not
shown)e
t body wt (dose dep., stat. sig.)
I lung and liver wt (dose dep.,
stat. sig.)
t adrenal wt (dose dep., stat. sig.)
At 0.81 ppm: f thymus and
kidney wt (stat. sig.)
Age of assessment NR.
At 0.81 ppm:
histologic effects in liver (e.g., t
extramedullary hematopoietic
centers), kidney (e.g., t
polymorphism of renal epithelial
cell nuclei)
and thymus
No effects
LOAEL/NOAELC
Maternal
L: 0.01 ppm
N:ND
NEe
NDg
Offspring
L: 0.01 ppm
N:ND
N: 0.01 ppm
L: 0.08 ppm
ND
Reference
Gofmekler(1968)
Gofmekler and
Bonashevskaya
(1969)f'h
Guseva (1972)
oo
VO
fe
H
o
o
-------
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
Table 4-70. Summary of reported developmental effects in formaldehyde inhalation exposure studies (continued)
Species,
strain, sex
Rat, strain
NR, female
Rat, Wistar,
female
Rat, Sprague-
Dawley,
female and
offspring of
both sexes
w/Group
NR
Embryo
dev expt:
5-9/group
(42 adult
animals);
maternal
effects: NR
(200 adult
females
total) '
25
Dose; time of
treatment"
Exptl:
0 or 30 ppmd
Expt 2: 0, pair-fed
control (15, 10, or
5 days), or 30 ppm; 8
hours/day for 15 days
(CDs 3-17), 10 days
(CDs 3-12), 5 days
(CDs 8-12), or 3
days (CDs 9-11)
0,0.4, or 1.2 ppm
(converted from
reported 0, 0.5 or 1.5
mg/m3); 4 hours/day,
5 days/week for
4 months; exposed
females mated to
unexposed males on
120th day exp.
0 (air control group),
0 (room control
group), 2, 5, or 10
ppm;
6 hours/day
CDs 6-15.
Exposed females
mated to unexposed
males
Reported study findings1"
Maternal
At 30 ppm
50% mortality (10 and
15 day exp)
I wt gain (duration dep.;
3, 5, 10, and 15 day exp.)
I wt of liver, kidney,
spleen and thymus
t wt of lung and adrenal6
At 0.4 ppm:
t wt of ovaries (stat.
sige)
t LH level (stat. sig.e)
At 1.2 ppm:
tFSH level in blood
(stat. sig.; nonsig. at
0.4 ppme)
At 10 ppm:
J, food consumption
(stat. sig.)
1 wt gain (stat. sig.)
Offspring
Fetuses:
At 30 ppm
1 fetal wt and growth (duration
dep., 10 and 15 day exp.)
t dev. defects (undescended
testes, large hearts, small
thymuses, small lungs)6
At 0.4 ppm:
t no. of embryos and 2
blastomere stage embryos (stat.
sig. in 2nd day preg.)
At 1.2 ppm:
t no. degenerating embryos (stat.
sig. in 3rd day preg.)
At 5 and 10 ppm:
Fetuses:
t incidence of reduced
ossification of pubic and ischial
bones (stat. sig. compared with air
control group)
I fetal wts (nonsig.)
t litter size (nonsig.)
LOAEL/NOAELC
Maternal
N:ND
L: 30 ppme
L: 0.4 ppm
N:ND
L: 10 ppm
TV: 5 ppm
Offspring
jV:ND
L: 30 ppme
L: 0.4 ppm
A^:ND
L: 5 ppm
N: 2 ppm
Reference
Kilburn and
Moro (1985)f Ab
Kitaev et al.
(1984)8
Martin (1990)e'f
oo
VO
fe
H
o
O
-------
Table 4-70. Summary of reported developmental effects in formaldehyde inhalation exposure studies (continued)
Species,
strain, sex
Rat, "white"
strain NR,
female and
offspring of
both sexes
Rat, Sprague-
Dawley,
female and
offspring of
both sexes
«/Group
5-12 (NR
for
formalde-
hyde only)
25
Dose; time of
treatment"
0,0.01, or 0.81 ppm
(reported as 0.0 12
and 1 mg/m3)d;
continuous
10-15 days prior to
mating through
gestation
0 (air control), 5, 10,
20, 40 ppm; 6
hours/day,
CDs 6-20.
Exposed females
mated to unexposed
males.
Reported study findings1"
Maternal
At 0.01 and 0.81:
I vit. C level in liver
(stat. sig.)
I vit. C level in placenta
(nonsig.)
GD 21 dams:
At 5 ppm:
t absolute body wt gain
(5 ppm only)
At 40 ppm:
J, body wt gain GDs
6-21 (stat. sig.)
J, absolute body wt gain
(stat. sig., dose-
dependent trend 20 and
40 ppm)
Offspring
Fetuses:
At 0.01 and 0.81 ppm:
I fetuses/female6
t body wt and organ wt (data not
shown6)
1 vit. C level in whole fetus (stat.
sig.)
At 0.01 ppm:
1 vit. C level in fetal liver (stat.
sig.)
GD 21 fetuses:
At 20 and 40 ppm:
1 fetal body wt, male (stat. sig.)
At 40 ppm:
Delayed ossification of thoracic
vertebrae (stat. sig., trend 20 ppm)
t unossified sternebrae (nonsig. at
40 ppm)
J, fetal body wt, female (stat. sig.)
LOAEL/NOAELC
Maternal
L: 0.01 ppm
N:ND
L: 40 ppm
N: 20 ppm
Offspring
L:0.01
A^ND
L: 20 ppm
N: 10 ppm
Reference
Pushkina et al.
(1968)f
Saillenfait et al.
(1989)
s ,
oo
VO
oo
fe
H
o
o
-------
Table 4-70. Summary of reported developmental effects in formaldehyde inhalation exposure studies (continued)
Species,
strain, sex
Rat, mongrel
white, female
and offspring
of both sexes
Mouse,
mongrel,
female and
offspring of
both sexes
«/Group
NR1
NR(254
dams) '
Dose; time of
treatment"
0 or 0.41 ppm
(reported as 0.5
mg/m3) formaldehyde
(also a 3rd group of
gasoline exposure,
not described in this
table); 4 hours/day
CDs 1-19
0 + ethyl alcohol;
0.41 ppm
formaldehyde;
0.41 ppm
formaldehyde +
bipyridyl;
4 hours/day
CDs 1-19.
Induced maternal iron
deficiency anemia by
IP. bipyridyl
injections on GDs
12-15; controls
injected w/25% ethyl
alcohol.
Reported study findings1"
Maternal
Dams GD 20:
I corpora lutea (nonsig.),
embryos dead before
implantation (not stat.
sig.), and implanted
embryos (nonsig.)
t blood pCO2 (stat. sig.)
Dams GD 20:
formaldehyde alone:
t blood pCC>2 (stat. sig.)
formaldehyde +
bipyridyl:
I blood acid metabolic
products (stat. sig.)
I blood true bicarbonates
and CO2 cone. (stat. sig.)
Offspring
Fetuses (GD 20):
Stat. sig. findings include
t fetal wt
t litters w/internal organ
anomalies
I fetuses w/ossification centers in
hyoid bone
t metacarpal bone centers
t metatarsal bone centers
t developmental defects
t blood pCO2 and pO2
Pups: I pup wt
Dev. delays (data not shown)
Fetuses (GD 20):
Formaldehyde alone:
t cryptorchidism
Formaldehyde + bipyridyl:
t birth defects (stat. sig.)
I dev. delay (stat. sig.)
I blood acid-base measures of
embryos (stat. sig.)
LOAEL/NOAELC
Maternal
L: 0.41 ppm
N:ND
L: 0.41 ppm
N:ND
Offspring
L: 0.41 ppm
A^ND
L: 0.41 ppm
jV:ND
Reference
Senichenkova
(1991)
Senichenkova
and Chebotar
(1996)
& S n
5'
s ,
oo
VO
VO
fe
H
o
o
-------
Table 4-70. Summary of reported developmental effects in formaldehyde inhalation exposure studies (continued)
Species,
strain, sex
Rat, mongrel,
white, female
and offspring
of both sexes
Rat, Sprague-
Dawley,
female and
offspring of
both sexes
Rat, Sprague-
Dawley,
female and
offspring of
both sexes
«/Group
15/group
terminated
GD20,
6/group
littered
6 dams
6 dams
Dose; time of
treatment"
0, 0.0005, or
0.005 mg/L (0, 0.4, or
4 ppm), CDs 1-19, 4
hours/day
0 or 6 ppm
8 hours/day, 6 weeks,
starting at GD 1,
PND 1, 4 weeks of
age, or adult age
0 or 6 ppm
8 hours/day; 6 weeks,
starting at GD 1,
PND 1, 4 weeks of
age, or adult age
Reported study findings1"
Maternal
At 0.4 ppm: J, leukocyte
counts
At 4 ppm: J, leukocyte
counts; reduced
threshold of
neuromuscular
excitability, J, rectal
temperature, J, blood
hemoglobin; f
spontaneous mobility
NE
NE
Offspring
At 0.4 ppm: f preimplantation
loss; at 1 mo. of age, J,
spontaneous mobility; at 2 mo. of
age, I hemoglobin levels and
leukocyte counts
At 4 ppm: f preimplantation loss;
at 1 and 2 mo. of age, J,
spontaneous mobility; at 2 mo. of
age, I hemoglobin levels and
leukocyte counts
In offspring exposed in utero and
during early postnatal life: J, mean
B W and liver weight; t markers
of oxidative stress
In offspring exposed as juveniles:
jmean liver weight; t markers of
oxidative stress
In offspring exposed only as
adults: no effect
In offspring exposed in early
postnatal life, as juveniles, or as
adults, t ANAE-positive T-
lymphocytes in B ALT
In all exposure initiation groups, t
ANAE-positive lymphocyte ratios
LOAEL/NOAELC
Maternal
L: 0.4 ppm
N:ND
NE
NE
Offspring
L: 0.4 ppm
A^ND
L: 6 ppm
jV:ND
L: 6 ppm
N:ND
Reference
Sheveleva(1971)
Kum et al. (2007)
Sandikci et al.
(2007)
& S n
5'
o
o
fe
H
O
o
-------
TO K.
3 ?
5'
3 ,
Table 4-70. Summary of reported developmental effects in formaldehyde inhalation exposure studies (continued)
Species,
strain, sex
Rat, Wistar,
female and
offspring of
both sexes
«/Group
12-14
dams
Dose; time of
treatment"
0, 6, or 12 ppm,
6 hours/day;
5 days/week, 30 days
Reported study findings'"
Maternal
NE
Offspring
At 6 and 12 ppm, at postnatal
days 30 and 90: respiratory
irritation and toxicity; deer. BW,
FC, WC; I SOD activity, \, levels
of copper and iron levels in lungs,
1 zinc levels in lungs
LOAEL/NOAELC
Maternal
NE
Offspring
L: 6 ppm
A^ND
Reference
Songur et al.
(2005)
-(^
O
fe
H
ND: not determined; NE: not evaluated; NR: not reported; Ab: abstract only; wt: weight; stat. sig.: statistically significant; p: pressure; 1: length;
To convert concentrations in air (at 25°C) from mg/m3 to ppm: 1 ppm = 1.23 mg/m3; 1 mg/m3 = 0.813 ppm.
""Treatment is given as the formaldehyde concentration in air (ppm) with the length of exposure each day and the duration of treatment in days, as available.
bStudies with negative findings are included.
CL: LOAEL; N: NOAEL.
dExposure concentrations not validated; details of formaldehyde vapor generation not reported; exposure during gestation not well characterized in study report.
eNo statistics provided.
fLack of study details.
gSee Table 4-73 for reproductive effects.
hGofmekler and Bonashevskaya (1969) seem to report on different findings from the same study (i.e., same animals) as Gofmekler (1968).
'Number/group not clear from study report.
O
O
-------
i
& s
II
Table 4-71. Summary of reported developmental effects in formaldehyde oral exposure studies
5'
s ,
Species,
strain, sex
Dog, beagle,
female and
pups
Mouse, CD-
1, female
Mouse,
ICR/SIM,
female
nl
Group
9-10
29-35
total
26-30
total
Dose; time of
treatment
0, 125, or
375 ppmc
(corresponding to
0, -3.1, or- 9.3
mg/kg-dd),
dietary, CDs 4-56
0, 74, 148, or 185
mg/kg-day, GDs
6-15
(aqueous
formaldehyde
solution contained
10-15%
methanol)
0 or 540 mg/kg-
day, GDs 8-12
Reported developmental effects a
Maternal
No effects
At 185 mg/kg-day:
Mortality
At 540 mg/kg-day:
Mortality
Offspring
At 125 (3.1 mg/kg-day), and
375 ppm (9.3 mg/kg-day):
I birth wt and wt gain through
postnatal week 8
No effects at GD 18
No effects in pups on PND 1
and 3
LOAEL/NOAELb
Maternal
L:ND
TV: 9.3 mg/kg-day (375
ppm)
L: 185 mg/kg-day
TV: 148 mg/kg-day
L: 540 mg/kg-day
N:ND
Offspring
L: 3.1 mg/kg-day
(125 ppm)
N:ND
L:ND
TV: 185 mg/kg-day
L:ND
TV: 540 mg/kg-day
Reference
Hurni and Ohder
(1973)
Marks et al.
(1980)
Seidenberg and
Becker (1987)
o
to
ND: not determined.
aStudies with negative findings are included.
bL: LOAEL; TV: NOAEL.
'Reported dose units.
dmg/kg-day doses are based on a conversion presented in Hurni and Ohder (1973).
fe
H
O
o
-------
its
& s
I- 1
2J I
5'
3 ,
Table 4-72. Summary of reported developmental effects in formaldehyde dermal exposure studies
Species,
strain, sex
Hamster,
Syrian
golden ,
female
»/
Group
4-6
Dose; time of treatment"
0 or 37%; 0.5 mL applied
to dorsal skin (hair
clipped) for 2 hours then
washed; CDs 8, 9, 10, or
11
Reported developmental effects
Maternal
Signs of dermal irritation and
irritability
Offspring
At all CDs of treatment, t
percent resorptions (not sig.)
LOAEL/NOAEL3
Maternal
L: 37%
jV:ND
Offspring
L: 37%
N: ND
Reference
Overman (1986)
J^.
o
fe
H
aL: LOAEL; jV: NOAEL.
ND = not determined.
O
O
-------
Table 4-73. Summary of reported reproductive effects in formaldehyde inhalation studies
Species,
strain, sex
Rat, strain
MR, male
Rat, Wistar,
female
Mouse,
B6C3F1,
male and
female
Rat, albino
Wistar,
male
Rat, Wistar,
male
nl
Group
4
NR (200
female)
10
7
18
Dose; time of treatment"
Inhalation plus drinking water
coexposure: 0; 0.10 ppmplus
0.005 mg/L water; 0.20 ppm
plus 0.01 mg/L water; or 0.41
ppm plus 0. 1 mg/L water; all
treatments; 4 hours/day,
5 days/week for 6 months.
Exposed males mated to
unexposed females
0, 0.4, or 1.2 ppm ; 4 hours/day,
5 days/week for 4 months.
Exposed females mated to
unexposed males on 120th day
of exposure.
0, 2, 4, 10, 20, or 40 ppm;
6 hours/day, 5 days/week
for 13 weeks
6 groups: 0, 10, or 20 ppm;
8 hours/day, 5 days/week for
4 weeks (subacute) or 13 weeks
(subchronic)
0, 5, or 10 ppm; 8 hours/day,
5 days/week, 91 days
Reported reproductive effects'3
At 0.20 ppm + 0.01 mg/L water and 0.41 ppm + 0.1 mg/L water:
[ nucleic acid in testes (dose dep.; data not shown; stat. sig.)
At 0.4 ppm:
t wt of ovaries (stat. sig.e)
t LH level in blood (stat. sig. at 0.4 ppmf)
At 1.2 ppm:
t FSH level in blood (stat. sig., dose dep. trendf)
Males and females:
At 20 ppm: J, wt gain
At 40 ppm: f mortality (13 weeks exp.); J, wt loss
Females:
At 40 ppm: f Uterine endometrial and ovarian hypoplasia
At 10 and 20 ppm (both durations):
I wt gain (stat. sig., dose dep.)
I relative testes wt (stat. sig., dose and cone, dep.)
I zinc and copper in testes (stat. sig., dose and cone, dep.)
t iron in testes (stat. sig., dose and cone, dep.)
No effect on testes wt.
At 5 and 10 ppm:
clinical signs of respiratory irritation, J, BW, FC, WC; J, serum
testosterone; J, mean seminiferous tubule diameters; t localization of
heat shock protein 70 in cytoplasm of spermatogonia,
spermatocytes, and spermatids
LOAEL/
NOAELC
L: 0.20 ppm
+ 0.01 mg/L
water
N:ND
L: 0.4 ppm h
jV:NDh
L: 20 ppm
jV:ND
L: 10 ppm
jV:ND
L: 5 ppm
A^ND
Reference
Guseva
(1972)d'e
Kitaev et al.
(1984)e
Maronpot et al.
(1986)
Ozen et al.
(2002)f
Ozen et al.
(2005)
J^.
o
fe
H
O
O
-------
Table 4-73. Summary of reported reproductive effects in formaldehyde inhalation studies (continued)
Species,
strain, sex
nl
Group
Dose; time of treatment"
Reported reproductive effects'3
LOAEL/
NOAELC
Reference
I
Rat, albino
Wistar,
male
10
0, 10, or 20 ppm
8 hours/day, 5 days/week for
4 weeks
Dose NR:
Irritation: standing hair, interrupted breathing, f eye blinking,
licking, nose cleaning, and sneezing.
At 10 and 20 ppm:
[ Body wt gain (dose dep.; stat. sig.)
[ Leydig cell quantity (stat. sig.)
t Nuclear damage of Leydig cells (dose dep.; stat. sig.)
L: 10 ppm
jV:ND
Sarsilmaz et al.
(1999)
Mouse,
strain not
specified,
male
12
0, 21, 42, or 84 mg/m3 (0, 16.9,
33.8, or 67.6 ppm); 2 hours/day.
6 days/week; 13 weeks.
Exposed males mated to
unexposed females.
In all treated groups: t percentage abnormal sperm, f resorption
rate, and [ live fetuses
L: 16.9 ppm
A^:ND
Xing et al.
(2007)
Rat, Wistar,
male
0 or 1.5 ppm, 18 weeks
formaldehyde exposures:
(1) 4 hours/day, 4 days/week
(2) 2 hours/day; 4 days/week
(3) 2 hours/day, 4 days/week
In all treated groups: sig. { seminiferous tubular diameter and
epithelial height. Other effects in exposure groups:
(1) sig. I germ cells; arrested spermatogenesis
(2) I cells, increased thickness in basal membrane
(3) t spaces between germ cells; disrupted association between
Sertoli and germinal cells
L: 1.5 ppm
jV:ND
Golalipour et al.
(2007)
Rat,
Sprague-
Dawley,
male
10
(1) 0 (gavage saline);
(2) 10 mg/m3 (8.05 ppm),
12 hours/day, 2 weeks; or
(3) 10 mg/m3 (8.05 ppm),
12 hours/day, 2 weeks, plus
30 mg/kg-day vitamin E orally
At 10 mg/m3: sig. \. testis weight, atrophy of seminiferous tubules, \.
spermatogenic cells, disintegrated and sloughed seminiferous cells;
edematous interstitial tissue with vascular dilatation and hyperemia;
I epididymal sperm count and percentage motile sperm, f
percentage abnormal sperm; I SOD, GSH-Px, GSH and t MDA in
testes; vitamin E attenuated all effects
L: 8.05 ppm
jV:ND
Zhou et al.
(2006)
Rat, Wistar,
male
40
0,0.1, 1, or 10 ppm;
6 hours/day, 5 days/week, 13 or
52 weeks
No effects: testis weight; histopathologic findings8
L:ND
TV: 10 ppm
Appleman et al.
(1986)
O
o
""Treatment is given as the formaldehyde concentration in air (ppm) with the length of exposure each day and the duration of treatment in days.
bStudies with negative findings are included.
CL: LOAEL; N: NOAEL.
dGuseva (1972) was a drinking water and inhalation study.
-------
a TO KO Tab\e 4-73. Summary of reported reproductive effects in formaldehyde inhalation studies (continued)
§ ^3 §- Developmental effects shown in Table 4-70.
j| J S Statistics not provided in study report.
!• 8 1 8Focus of study was not the reproductive syste
>- ^ ^ reproductive system findings.
a 5. a h F°r increased FSH, the LOAEL was 1.2 ppm (1.5 mg/m3) and the NOAEL was 0.4 ppm (0.5 mg/m3).
& s 9
|- g I gFocus of study was not the reproductive system; only reproductive system findings are addressed in the table; NOAEL and LOAEL in table are based only on
>- ^ ^ reproductive system findings.
~ ^ ^ hT^_.-__ _1 -r^CTT-r ,I__T/"VA T^T __ 1 --» ____ /I f / _^3\ f 1 ,1,^ M/^ A T7T ,,,^^ A A ™ /A C ^,™/*«3\
X ~
'o a, "§> ND: not determined; NR: not reported.
^ §- ^ To convert concentrations in air (at 25°C) from mg/m3 to ppm: 1 ppm = 1.23 mg/m3; 1 mg/m3 = 0.813 ppm.
-i^
o
fe
H
O
O
-------
reproductive system in rats following 5 days/week of inhalation exposures is 5 ppm for 3 months
of daily exposures and 10 ppm for 4 weeks of daily exposures; these dose levels are unbounded.
Abnormal sperm were also noted in mice at an inhalation dose of 16.9 ppm 2 hours/day, 6
days/week for 13 weeks (Xing et al., 2007), but, in contrast, Maronpot et al. (1986) reported no
histologic abnormalities in male mice after formaldehyde exposures at 40 ppm 6 hours/day, 5
days/week for 13 weeks. Varied results among studies may be due to species differences or
differences in methods. Although several oral subchronic and chronic studies with formaldehyde
did not identify effects on the testes (Tobe et al., 1989; Til et al., 1988; Johanssen et al., 1986),
Cassidy et al. (1983) observed spermatogenic abnormalities after a single oral dose of 200 mg/kg
to rats, and a chronic drinking water study in rats (Til et al., 1989) reported low incidences of
Leydig cell tumors in all treated groups, compared with none in control (see Table 4-74).
Additionally, studies utilizing IP. injection of formaldehyde in rats have demonstrated testes and
sperm anomalies (Majumder and Kumar, 1995; Chowdhury et al., 1992) and dominant lethal
effects (Odeigah, 1997) (see Table 4-75).
4.3. GENOTOXICITY
Formaldehyde has been extensively studied for its mutagenic and genotoxic activity in a
variety of assay systems. The first reported mutagenic activity of formaldehyde was when
Rapoport (1946) described the induction of sex-linked recessive lethals inDrosophila larvae fed
on a medium containing formalin. A variety of genotoxic and mutagenic effects have been
subsequently demonstrated in both in vitro and in vivo test systems, including the formation of
DNA-protein crosslinks (DPXs or DPCs), point mutations, DNA strand breaks, increased
micronuclei (MN), and chromosomal aberrations (CAs) (Auerbach et al 1977; Ma and Harris,
1988; Conaway et al 1996; IARC 1995; 2006).
In this section, reactions of formaldehyde with cellular macromolecules, such as DNA
and proteins, and formaldehyde-induced clastogenicity are described. In addition, the evidence
for formaldehyde-induced mutations is considered in the context of the current EPA cancer
guidelines (U.S. EPA, 2005a). Particular emphasis is given to the genotoxic effects of
formaldehyde in humans, described in Section 4.3.4.2.
4.3.1. Formaldehyde-DNA Reactions
Formaldehyde is a reactive chemical and interacts with DNA in several ways, forming
DPXs or DPCs, DNA adducts, and DNA-DNA cross-links (DDXs) (Fennell, 1994; Casanova et
al., 1989; Heck and Casanova, 1987; Casanova-Schmitz et al., 1984a, b; Casanova-Schmitz and
Heck, 1983; Ohba et al., 1979; Donecke, 1978; Brutlag et al., 1969). Formaldehyde also may
This document is a draft for review purposes only and does not constitute Agency policy.
4-407 DRAFT—DO NOT CITE OR QUOTE
-------
ts
^ 8
1!
f^ S
2 5'
s ,
Table 4-74. Summary of reported reproductive effects in formaldehyde oral studies
Species,
strain, sex
Rat, Wistar,
male
Rat,
Sprague-
Dawley,
both sexes
Dog, beagle,
both sexes
Rat, Wistar,
both sexes
Rat, Wistar,
both sexes
Rat, Wistar,
both sexes
nl
Group
5(20
control)
15
4
10
70
20
Dose; time of treatment
0, 100, or 200 mg/kg, single
gavage dose
0, 50, 100, or 150 mg/kg-day,
drinking water; 91 days
0, 50, 75, or 100 mg/kg-day,
dietary; 91 days
0, 25, or 120 mg/kg-day,
drinking water, 4 weeks
0, 1.2, 15, or 82 mg/kg-day
(males), 0, 1.8, 21, or
109 mg/kg-day (females),d
drinking water; 105 weeks
0, 0.02, 0.1, or 0.5% in drinking
water; 24 months
Reported reproductive effects3
At 200 mg/kg:
t (19%) testicular sperm head counts (stat. sig.)
t (5%) abnormal sperm head (stat. sig.)
No effects: absolute or relative gonad weights;
histopathologic findings of reproductive organs0
No effects: absolute or relative gonad weights;
histopathologic findings of reproductive organs0
No effects: gonad weights; histopathologic findings of
reproductive organs0
In all treated groups:
Leydig cell tumors observed at 105 weeks of study0
At 82 mg/kg-day:
t mean testes weights
No effects: gonad weights; histopathologic findings of
reproductive organs0
LOAEL/
NOAELb
L: 200 mg/kg
N: 100 mg/kg
L:ND
N: 150 mg/kg-day
L:ND
N: 100 mg/kg-day
L:ND
N: 120 mg/kg-day
L: 1.2 mg/kg-day
N:ND
L:ND
N: 0.5%
Reference
Cassidy et al.
(1983)
Johanssen et al.
(1986)
Til etal. (1988)
Til etal. (1989)
Tobe et al.
(1989)
o
oo
fe
H
"Studies with negative findings are included.
bL: LOAEL; N: NOAEL.
°Focus of study was not the reproductive system; only reproductive system findings are addressed in the table; NOAEL and LOAEL in table are based only on
reproductive system findings.
dActual concentrations.
ND = not determined.
O
O
-------
it
& s
II
Table 4-75. Summary of reported reproductive effects in formaldehyde intraperitoneal studies
5'
s ,
Species,
strain, sex
Rat, Charles
foster, male
Rat, Wistar,
male
Rat,
"albino"
strain NR,
male
Rat,
"albino"
strain NR,
male
nl
Group
10
8
6
12
Dose; time of treatment"
0 or 5 mg/kg-day; 30 days
0 or 10 mg/kg-day; 30 days
0, 0.125, 0.250, or 0.60 mg/kg-
day; 5 days
0, 0.125, 0.250, or 0.60 mg/kg-
day; 5 days. Exposed males
mated to unexposed females.
Reported reproductive effects3
[ body weight gain
[ Leydig cell population and cell nuclear diameter
[ serum T levels
[ testes weights
cellular degeneration of seminiferous tubules
[ sperm count, motility and sperm viability
At all treatment levels:
[ sperm count and t sperm head abnormalities (3 weeks
after the last injection)
At all treatment levels (at GD 13):
Delayed time to mating
[ mean no. implants and live embryos
t dead implants and dominant lethal index following mating
to untreated females
LOAEL/
NOAELb
L: 5 mg/kg-day
N:ND
L: 10 mg/kg-day
N:ND
L: 0.125 mg/kg-day
N:ND
L: 0.125 mg/kg-day
N:ND
Reference
Chowdhury et
al. (1992)
Majumder and
Kumar (1995)
Odeigah(1997)
o
VO
fe
H
aStudies with negative findings are included.
bL: LOAEL; A^: NOAEL.
ND = not determined; NR = not reported; T = testosterone.
O
O
-------
1 facilitate the formation of adducts between other chemicals or drugs (endogenous or xenobiotic)
2 and DNA (Fennell, 1994; Koppel et al., 1991; Casanova et al., 1989; Heck and Casanova, 1987;
3 Lam et al., 1985; Casanova-Schmitz et al., 1984a; Casanova-Schmitz and Heck, 1983; Ohba et
4 al., 1979; Donecke, 1978; Brutlag et al., 1969).
5 The high reactivity of formaldehyde results in little specificity in reaction sites, indicating
6 that a range of adducts and cross-links might be expected. However, the spectrum of
7 formaldehyde-DNA reaction products is difficult to quantify in vivo as many of these products
8 are labile and difficult to measure directly (Fennell, 1994; Casanova et al., 1989). Additionally,
9 formaldehyde is metabolically incorporated into nucleic acids, and therefore DNA and RNA
10 assays incorporating radiolabeled formaldehyde need careful interpretation to distinguish
11 between covalently bound and metabolically incorporated formaldehyde (Casanova et al., 1989;
12 Heck and Casanova, 1987; Casanova-Schmitz et al., 1984a, b; Casanova-Schmitz and Heck,
13 1983). Hence, reports of formaldehyde-DNA reactivity in cell-free system results may not
14 provide a useful measure of exposure (Fennell, 1994). Besides, the question of biological
15 relevance must also be considered. On the other hand, methods used to extract and measure
16 formaldehyde-DNA reaction products after in vivo exposures should be evaluated to ensure that
17 formaldehyde reaction products are neither created nor removed during sample preparation
18 (Fennell, 1994; Casanova et al., 1989).
19
20 4.3.1.1. DNA-Protein Cross-Links (DPXs)
21 Formaldehyde forms DPX by reacting with the amino or imino groups of proteins (e.g.,
22 lysine and histidine side chains) or of nucleic acids (e.g., cytosine) resulting in a Schiff base
23 formation which then reacts with another amino group (McGhee and von Hippel 1975a, b).
24 Evidence from numerous experimental models, ranging from cell-free systems to single cells and
25 in vivo animal and human exposures, suggests that formaldehyde reacts readily with DNA
26 forming DPXs (Reviewed in Conaway et al 1996; IARC 2006, 1995). As shown in Table 4-76,
27 cross-links between histones and DNA have been demonstrated in isolated chromatin samples on
28 exposure to formaldehyde from earlier studies (Ohba et al., 1979; Donecke, 1978; Brutlag et al.,
29 1969). Several in vitro studies demonstrated induction of DPX by formaldehyde exposure in
30 bacteria (Wilkins and McLeod 1976), yeast (Magana-Schwencke and Ekert, 1978) and
31 mammalian cells including animal cells (Chinese hamster ovary [CHO] cells, Chinese hamster
32 V79 lung epithelial cells, mouse leukemia L1210 cells, mouse hepatocytes, rat Yoshida
33 lymphosarcoma cells, rat CIS tracheal epithelial cells, hepatocytes, nasal, tracheal, buccal
34 epithelial cells, kidney cells and aortic endothelial cells) and human cells (lung and bronchial
35 epithelial cells, fibroblasts, white blood cells, peripheral blood lymphocytes, Epstein-Barr Virus
36 (EBV)-Burkitt's lymphoma cells, Jurkat E6-1 cells, HeLa cells, lymphoblastoid cells, gastric
This document is a draft for review purposes only and does not constitute Agency policy.
4-410 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-76. Formaldehyde-DNA reactions (DPX formation)
Species/Strain
Cell/Strain
Result
References
DNA Interaction
DPX formation in vitro
In vitro
Bacteria
Yeast
Hamster/Chinese
Hamster/Chinese
Mouse
Rat/unspecified
Rat/F344
Rat/Wistar
Human
Nucleohistone
Escherichia coli
Saccharomyces cerevisiae
Ovary cells
V79 lung epithelial cells
Leukemia L 12 10 cells
Hepatocytes
Yoshida lymphosarcoma cells
CIS tracheal epithelial cell line
Hepatocytes
Nasal mucosa
Nasal epithelium
Primary tracheal epithelial cells
Aortic endothelial cells
Lung/bronchial epithelial cells
Bronchial epithelial cells
Lung/bronchial epithelial cells,
fibroblasts, skin fibroblasts
Foreskin fibroblasts
Bronchial/Skin fibroblasts
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Brutlagetal., 1969
Doenecke, 1978
Ohbaetal., 1979
Wilkins andMcLeod, 1976
Magana-Schwencke andEkert, 1978
Marinari et al, 1984
Zhitkovich and Costa, 1992
Olinetal., 1996
Garcia et al., 2009a
Swenberg et al., 1983
Merk and Speit 1998, 1999
Speit et al., 2007a
Ross and Shipley, 1980
Rossetal., 1981
Casanova and Heck, 1997
Casanova et al., 1997
O'Connor and Fox, 1987
Cosma and Marchok, 1988
Casanova and Heck, 1997
Casanova-Schmitz and Heck, 1983
Bermudez and Delehanty, 1986
Cosma etal., 1988
Lin etal., 2005
Fornaceetal., 1982
Saladino et al., 1985
Graf strom etal., 1986, 1990
Graf strom etal., 1983
Graf strom etal., 1984
Snyder and Van Houten, 1986
Olinetal., 1996
This document is a draft for review purposes only and does not constitute Agency policy.
4-411 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-76. Formaldehyde-DNA reactions (DPX formation) (continued)
Species/Strain
Human (continued)
Cell/Strain
White blood cells
EBV-Burkitt's lymphoma cells
Gastric mucosa cells
Peripheral blood lymphocytes
Fibroblasts
Primary skin fibroblasts and
keratinocytes
Buccal epithelial cells
JurkatE6-l cells
HeLa cells
Whole blood cultures
Lung/bronchial epithelial cells
Lymphoblastoid/TK6
Lung carcinoma (A549) cells
Result
+
+a
+
+
+
+
+
+
+
+
+
+
+
+
+
References
Shahametal., 1996
Costa etal., 1997
Blasiak et al 2000
Quievryn and Zhitkovich 2000
Andersson et al., 2003
Liu et al 2006
Speit et al 2000
Emri et al 2004
Li et al 2004
Saito et al 2005
Liu et al 2006
Schmid and Speit 2007
Speit et al 2008
Craft etal., 1987
Speit etal., 2010
DPX formation in vivo
Rat
Rhesus monkeys
Human
Nasal mucosa
Tracheal implants
Peripheral blood cells
Bronchoalveolar lavage cells
Nasal, larynx, trachea, and carina
White blood cells
Peripheral blood lymphocytes
+
+
+
+
+
+
+
-
-
+
+
+
Casanova-Schmitz and Heck 1983b
Casanova-Schmitz et al 1984b
Lam etal 1985
Heck et al 1986
Heck Hd and Casanova 1987
Casanova etal 1989, 1994
Cosmaetal., 1988
Speit et al 2009
Neussetal2010
Casanova etal 1991
Shahametal., 1996
Shahametal., 1997,2003
2
3
4
5
6
a indicates that DNA-protein cross-links formed at cytotoxic concentrations.
b used homogenates of nasal mucosa.
'+' indicates a positive test result.
'-' indicates a negative test result.
This document is a draft for review purposes only and does not constitute Agency policy.
4-412 DRAFT—DO NOT CITE OR QUOTE
-------
1 mucosal cells and whole blood cultures) and in vivo studies involving experimental animals and
2 occupationally exposed human populations as summarized in Table 4-76. Some relevant studies
3 are described here.
4 Ross and Shipley (1980) showed that formaldehyde induces single strand breaks (SSBs)
5 and DPXs in mouse leukemia cells; SSBs are formed at concentrations >200 jiM and a reduction
6 of radiation-induced breaks (indirect measure of DPXs) at 50 jiM. The authors used a [14C]-
7 thymidine-incorporated mouse L1210 cell line to monitor formaldehyde-induced DNA strand
8 breaks and DPXs. They exposed cells to varying concentrations of formaldehyde for 2.5 hours.
9 An alkaline-elution technique in the presence or absence of proteinase K was used to measure
10 strand breaks. In order to detect DPXs, some cells were exposed to 300 R of X-rays immediately
11 after formaldehyde treatment. Formaldehyde-induced DPXs were repaired 24 hours after the
12 compound was removed from the culture (Ross and Shipley, 1980).
13 Casanova-Schmitz and Heck (1983) have shown that homogenates of rat nasal mucosa
14 incubated with formaldehyde in vitro followed by extraction with a strong aqueous-immiscible
15 organic solvent demonstrated increased DPX formation in DNA obtained after enzymatic
16 proteolysis from the aqueous-organic interface (IF), termed as "interfacial DNA" (IF-DNA).
17 Bermudez and Delehanty (1986) observed the formation of DPXs, scheduled (S) and
18 unscheduled DNA synthesis (UDS), and synthesis of RNA when cultured F344 rat nasal
19 epithelial cells from the naso- and maxillary turbinates were incubated with formaldehyde.
20 Unscheduled and scheduled DNA synthesis was stimulated (0.05-0.1 mM) and then inhibited
21 (0.1-1 mM), depending on the formaldehyde concentration. Experiments by Cosma et al. (1988)
22 and Cosma and Marchok (1988) showed the induction of DPXs and DNA SSBs in cultured CIS
23 rat tracheal epithelial cells exposed to 200 |jM formaldehyde for 90 minutes (Cosma et al., 1988;
24 Cosma and Marchok, 1988).
25 Several human cells (epithelial cells, fibroblasts, buccal cells) or cell lines
26 (lymphoblastoid cells) exposed to formaldehyde have been shown to form DPX (Craft et al
27 1987; Costa et al 1997; Emri et al 2004; Li et al 2004).
28 Craft et al. (1987) detected DPXs by alkaline elution in TK6 human lymphoblastoid cells
29 immediately after a 2-hour exposure (zero time) to 0, 15, 50, 75, 100, 150, 300, and 600 uM
30 formaldehyde with a significant nonlinear increase in DPXs above 50 uM concentration, which
31 correlated with the onset of cytotoxicity, but DPXs were completely removed in cultures held for
32 24 hours before processing. In the zero-time sample, significant increases in DPXs were first
33 observed at 50 uM and increased linearly up to 150 uM. In cells held for 24 hours, there was no
34 detectable increase in DPXs.
This document is a draft for review purposes only and does not constitute Agency policy.
4-413 DRAFT—DO NOT CITE OR QUOTE
-------
1 However, Costa et al. (1997) detected DPXs with paraformaldehyde (which dissociates to
2 release formaldehyde) at doses that were cytotoxic (>0.003%) but could not discriminate
3 between the DPX-inducing and cytotoxic effects of this chemical in EBV-human Burkitt's
4 lymphoma cells (Costa et al., 1997). Grafstrom et al. (1983) reported that the number of DPXs
5 induced by 100 |jM formaldehyde in vitro in human epithelial cells and fibroblasts of bronchial
6 origin was similar and that the frequency of these cross-links was proportional to the
7 concentration of formaldehyde. Besides the bronchial epithelial cells and fibroblasts, the authors
8 also noted that formaldehyde exposure resulted in DPXs and DNA SSBs in skin fibroblasts and
9 DNA excision repair-deficient skin fibroblasts. However, formaldehyde was only moderately
10 cytotoxic to normal bronchial epithelial cells and fibroblasts at concentrations that induced
11 substantial DNA damage. Repair of the formaldehyde-induced DNA SSBs and DPXs appeared
12 to be inhibited by the continued presence of formaldehyde in the culture medium (Grafstrom et
13 al., 1984).
14 Emri et al. (2004) detected a significant increase in DPX formation in primary human
15 skin fibroblasts and keratinocytes at 8 hours of exposure in vitro to formaldehyde at 25 |jM with
16 an approximately linear increase up to 100 |jM. These cells were exposed to 0, 12.5, 25, 50, and
17 100 (jM formaldehyde for 8 hours and then exposed to 250 |jM methyl methane sulfonate
18 (MMS) for 2.5 hours. The induction of DPX formation was measured by the ability of
19 formaldehyde to reduce DNA migration in the comet assay induced by MMS in this study (Emri
20 et al., 2004).
21 Li et al. (2004) measured DNA damage in primary human buccal cells by using the
22 comet assay. The appearance of SSBs, suggesting compound-induced fragmentation of DNA,
23 occurred at formaldehyde concentrations of 5 and 7.5 |jM. At higher concentrations, the
24 response slope decreased, indicating formation of DPXs or DDXs (Li et al., 2004). The same
25 laboratory reported similar findings in primary human peripheral blood lymphocytes and HeLa
26 cells (Liu et al., 2006). Peak response for SSBs was seen at 10 |jM in both cells, with higher
27 concentrations resulting in cross-link formation. SSBs in HeLa cells induced by 10 |jM
28 formaldehyde were repaired by 60 minutes after cells were washed to remove formaldehyde.
29 Schmid and Speit (2007) tested formaldehyde for its ability to induce DPXs in blood
30 cultures. They used an indirect method to monitor DPX formation in which the extent of DNA
31 migration in the comet assay in response to y radiation was compared in formaldehyde-treated
32 cultures versus controls. A concentration of 25 jiM was required for DPX formation, and repair
33 of these lesions was rapid, with DPXs induced by concentrations of formaldehyde up to 100 jiM
34 and completely removed after 8 hours. Overall, several in vitro studies in cultured mammalian
This document is a draft for review purposes only and does not constitute Agency policy.
4-414 DRAFT—DO NOT CITE OR QUOTE
-------
1 cells demonstrated the formation of DPX and SSB formation confirming the genotoxicity of
2 formaldehyde.
3 Several in vivo studies involving rodents, monkeys and humans have demonstrated DPX
4 formation following formaldehyde exposure (see Table 4-76). As mentioned earlier, in vitro
5 studies by Casanova-Schmitz and Heck (1983) have shown that reaction of formaldehyde with
6 cellular macromolecules in tissue homogenates causes a decrease in the extractability of DNA
7 into aqueous phase during solvent extraction and the formaldehyde-bound DNA-protein complex
8 migrates into the interface. In the same study Casanova-Schmitz and Heck (1983) have also
9 shown that DNA isolated from the nasal, but not olfactory, mucosa of rats exposed to
10 formaldehyde (2, 6, 15, and 30 ppm 6 hours/day for 2 days) via inhalation showed significant
11 increase in DPXs in the interfacial DNA >6 ppm, which was shown to be linear in the exposure
12 range of 2-30 ppm (2.45-36.8 mg/m3). However, DNA in the aqueous phase did not show DPX
13 formation. Thus, the cross-linked DNA that could be extracted from the interface after
14 proteolysis was considered to be supporting evidence of chemically induced DPX formation.
15 The inability of this study to detect DPXs at lower levels of formaldehyde exposure is likely be
16 due to the protective mechanism of GSH, which catalyzes the conversion of formaldehyde to
17 formate. Later, Lam et al., (1985) have shown that coexposure of rats with 2 ppm acrolein and 6
18 ppm formaldehyde for 6 hours resulted in higher DPX in the nasal mucosa of rats compared to
19 the rats given formaldehyde alone, suggesting that GSH depletion by acrolein enhanced the DNA
20 binding of formaldehyde. In this study inhalation exposure of rats to acetaldehyde alone at 0,
21 0.1, 0.5, 1 and 2.5 ppm for 6 hours induced a concentration-dependent depletion of GSH in rat
22 nasal mucosa with no detectable DPX levels at 2 ppm dose. However, acetaldehyde forms DPX
23 (as interfacial DNA) in rat nasal mucosa at much higher concentrations (100-1000 ppm)
24 compared to formaldehyde.
25 So, in a later study, Casanova and Heck (1987) reported that GSH depletion caused an
26 increase in DPX formation in the IF-DNA in the nasal mucosa of F344 rats when a dual-isotope
27 (3H/14C) method was used. The dual isotope method helps in making the distinction between
28 metabolic incorporation and covalent binding of formaldehyde. Oxidation by removal of one
29 hydrogen atom is required for metabolic incorporation of formaldehyde into cellular
30 macromolecules, but not in the formation of DNA adducts or DNA-protein crosslinks. Thus, the
31 ratio of 3H/14C of DNA containing DPX will be higher than the macromolecules where
32 formaldehyde is metabolically incorporated. However, the authors further demonstrated that,
33 when the double isotope method was used, the 3HCHO is oxidized significantly more slowly
34 than H14CHO under these conditions, resulting in an overestimate of the concentration of cross-
35 links due to an isotopic effect on the oxidation of 3HCHO catalyzed by formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-415 DRAFT—DO NOT CITE OR QUOTE
-------
1 dehydrogenase (FDH). Besides, this method leaves residual formaldehyde that is likely to form
2 DNA adducts by reacting with deoxyribonucleosides in the DNA hydrolysates (Heck and
3 Casanova, 1987).
4 To overcome this, Casanova et al. (1989) used an improved method, which is based not
5 on the analysis of residual formaldehyde bound to deoxyribonucleosides in DNA hydrolysates
6 but on the determination of the total 14C-formaldehyde bound to DNA. This study showed that
7 formaldehyde was exclusively bound to interfacial DNA, indicating the formation of DPXs.
8 Hydrolysis of DPXs in different samples quantitatively released formaldehyde. Besides, DPX
9 formation was detectable at all concentrations of exposure to formaldehyde (0.3-10 ppm for
10 6 hours). Overall, these studies clearly show that formaldehyde induces DPXs in nasal epithelial
11 cells of rodents. However, there are no published rodent studies that assess DPXs beyond the
12 nasal passages of the upper respiratory tract.
13 Formaldehyde-induced DPXs were also found in the nasal mucosa and extranasal tissues
14 of rhesus monkeys exposed to 0, 0.71, 2, or 6 ppm (0, 0.86, 2.45, or 7.36 mg/m3) formaldehyde
15 6 hours/day for 3 days (Casanova et al., 1991). These data were used as a basis for cross-species
16 prediction of formaldehyde-induced DPXs in humans. The presence of DPXs in rhesus monkeys
17 confirms formaldehyde's DNA reactivity as a general effect. Additionally, DPXs were detected
18 in the larynx/trachea/carina (pooled sample) and in intrapulmonary airways of monkeys exposed
19 to 2 or 6 ppm formaldehyde. These data demonstrate direct effects of formaldehyde on DNA in
20 tissues that correspond to observed tumor sites in humans (nasal and nasopharynx).
21
22 4.3.1.2. DNA Adducts
23 In addition to the formation of DPX, there is evidence that formaldehyde forms
24 hydroxymethyl (hm) DNA adducts in vitro in a variety of cell-free systems (Zhong and Que Hee,
25 2004a; Cheng et al., 2003; Kennedy et al., 1996; Fennell, 1994; Beland et al., 1984) and nasal
26 epithelial cells (Zhong and Que Hee 2004b). In cell-free systems, formaldehyde directly reacts
27 with DNA forming hmDNA adducts (Cheng et al., 2003; Kennedy et al., 1996; Fennell, 1994;
28 McGhee and von Hippel, 1977a, b, 1975a, b). The results on formaldehyde-induced DNA
29 adduct formation are summarized in Table 4-77.
30 Beland et al. (1984) first reported hmDNA adducts in Chinese hamster ovary (CHO) cells
31 incubated with 1 mM of radiolabeled formaldehyde. After a 2-hour incubation, small amounts of
32 N6-hydroxymethyldeoxyadenosine (N6-hmdA) adducts were detected with concomitant
33 metabolic incorporation of formaldehyde. Besides N6-hmdA, various other forms of hmDNA
34 adducts, including N4-hydroxymethyldeoxycytidine (N4-hmdC), and
35 N2 hydroxymethyldeoxyguanosine (N2-hmdG), were detected by high performance liquid
36
This document is a draft for review purposes only and does not constitute Agency policy.
4-416 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-77. Formaldehyde-DNA reactions (DNA adduct formation)
Species/Strain
Cell/Strain
Result
References
DNA Interaction
DNA Adduct Formation in vitro
Cell-free system
Placental DNA
Calf thymus DNA
Calf thymus DNA
Cell-free system
Rats
Hamster
Human
Deoxyribonucleosides
Guanosine
Guanosine
In vitro
In vitro
In vitro
In vitro
Nuclei
Nasal epithelial cells
Nuclei
CHO cells
Nasal epithelial cells
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Cheng et al 2008
Kennedy et all 996
Cheng et al 2003
Zhong and Que Hee 2004a
Belandetall984
Von Hippel and Wang 1971
McGhee and von Hippel 1975a
McGhee and von Hippel 1975b
McGhee and von Hippel 1977a
McGhee and von Hippel 1977b
Fennell 1994
Cheng et al 2003
Fennell 1994
Casanova et all 989
Heck Hd and Casanova 1987
Belandetall984
Zhong and Que Hee 2004b
DNA adduct formation in vivo
Drosophila
Rats
Rat/F344
Rat/F344
Human
Larvae
Indirect evidence
Nasal epithelium
White blood cells, bone marrow cells, lung,
liver, spleen, thymus
White blood cells
+
+
+
-
+
Alderson, 1985
Wang et al 2007
Luetal2010
Luetal2010
Wang et al 2009
3
4
5
6
7
8
9
10
11
'+' indicates a positive test result.
'-' indicates a negative test result.
chromatography (HPLC) following in vitro reaction between formaldehyde and calf thymus
DNA or individual deoxynucleotides.
Since the formaldehyde adducts are labile, Fennel (1994) developed a method by
derivatizing them with sodium bisulfite to their sulfomethyl form, whereby he detected
This document is a draft for review purposes only and does not constitute Agency policy.
4-417 DRAFT—DO NOT CITE OR QUOTE
-------
1 N6-sulfomethyldeoxyadenosine (SOMedA) and N2-sulfomethyldeoxyguanosine (SOMedG) by
2 using HPLC. However, the levels of SOMedA in DNA isolated following incubation of
3 radiolabeled formaldehyde with isolated rat hepatic nuclei were similar to those in control nuclei.
4 And in human TK6 lymphoblastoid cells treated with formaldehyde, detection of SOMedA
5 adducts was precluded by additional radioactive spots. These observations suggest that N6-
6 sulfomethyldeoxyadenosine adducts are formed at very low levels in formaldehyde-incubated rat
7 hepatic nuclei and that measurement of hydroxymethyldeoxyadenosine would not provide a
8 useful measure of formaldehyde exposure. Fennel (1994) also reported that 32P-postlabeling
9 studies (Gupta et al. 1982) allowed for much greater analytical sensitivity but did not confirm the
10 level of SOMedA adduct detected by HPLC. However, either estimate of adduct formation is
11 much less than the estimate of DPX formation (120 pmol/mg DNA) in similarly treated rat
12 nuclei (Heck and Casanova, 1987).
13 Casanova et al. (1989) demonstrated that detection of hmDNA adduct formation was
14 sensitive to the methodology used, particularly the buffer used for sample preparation.
15 Specifically, Tris buffer can prevent hmDNA adduct formation due to the abundance of
16 formaldehyde-reactive primary amine sites in the buffer. In contrast, the tertiary amine sites that
17 predominate in Bis-Tris buffer do not react with formaldehyde.
18 Zhong and Que Hee (2004a) observed hmDNA adducts (N6-hmdA, N2-hmdG, and
19 N4-hmdC) in placental DNA exposed to 100 ppm formaldehyde in vitro for 20 hours at 37°C
20 followed by hydrolysis of formaldehyde-reacted DNA using bis-Tris buffer. However,
21 deoxythymidine did not form hydroxymethyl derivatives in this study (Zhong and Que Hee,
22 2004a). On the other hand, the same investigators were able to detect N6-hmdA and N2-hmdG
23 adducts in human nasal epithelial cells cultured in the presence of 0, 10, 25, 50, 100, 250, 400, or
24 500 |ig/mL formaldehyde and using Tris buffer during hydrolysis of adducted DNA. The
25 toxicity threshold for <90% viability appeared to be between 100 and 250 |ig/mL initial
26 formaldehyde culture concentration, and even at 500 |j,g/mL concentration it was not toxic, with
27 a viability was 70% in this study (Zhong and Que Hee, 2004b).
28 There are only a few reports of formaldehyde-induced hmDNA adducts in vivo.
29 Alderson (1985) hypothesized that a N6-hydroxymethyl adenosine, an RNA adduct formed in
30 the reaction of formaldehyde and adenosine is likely to play a mutagenic role in Drosophila.,
31 based on the observation that flies fed on a medium containing formalin was mutagenic only
32 when adenosine or adenylic acid, but not adenine was present in the medium. An indirect
33 evidence from the study of Wang et al., (2007) has shown that N6-hydroxymethyl-
34 deoxyadenosine (N6-hmdA) adducts were formed in hepatic and pulmonary DNA from rats
35 exposed to 7V-nitrosodimethylamine (NDMA) and 4-(methylnitrosamino)-l-(3-pyridyl)-l-
This document is a draft for review purposes only and does not constitute Agency policy.
4-418 DRAFT—DO NOT CITE OR QUOTE
-------
1 butanone (NNK), suggesting that these two nitrosamines could be a source of formaldehyde-
2 DNA adducts. The same group recently has shown that formaldehyde-DNA adducts are
3 detectable in the leukocytes of smokers and nonsmokers. More recently, Lu et al (2010) have
4 shown that in rats exposed to formaldehyde by inhalation deoxyguanosine monoadducts and
5 DDX were detected in nasal epithelium, while endogenous formaldehyde monoadducts and
6 DDX were detectable in white blood cells and several internal organs including liver, lung,
7 spleen, thymus and bone marrow.
8
9 4.3.1.3. DNA-DNA Cross-Links (DDXs)
10 Formaldehyde, besides forming DPXs and DNA adducts, has also been shown to form
11 DDX in vitro. Li et al. (2004) showed that formaldehyde induces DNA strand breaks at low-
12 exposure concentration and DDXs and DPXs at higher concentrations in buccal cells. The
13 authors also showed that formaldehyde induces DDXs in human peripheral blood lymphocytes
14 exposed in vitro when the concentration was more than 25 |jM. Recently, Lu et al (2010)
15 reported DDXs in the nasal epithelium of F344 rats that were exposed by nose-only to 10 ppm
16 formaldehyde for 1 or 5 days (6 hrs/day). The relevance of these modifications in formaldehyde-
17 induced genotoxicity is not known at the moment.
18 Overall, formaldehyde forms predominantly DPXs that are detected in cell-free systems
19 and single cells in vitro and in animal and human tissues in vivo. In rodents, DPXs are formed in
20 nasal epithelia but not in extranasal passages, which are completely removed within a day after
21 formation. The DPXs are detected in nasal and extranasal tissues of monkeys, suggestive of
22 direct effects of formaldehyde in tissues that correspond to observed tumor sites (nasal and
23 nasopharynx) in humans. Besides, this is used as a basis for cross-species comparison with
24 humans. Formaldehyde-DNA adducts are labile, constituting a minor fraction of the DNA-
25 reaction products. DPXs but not DNA adducts appear to play an important role in the
26 genotoxicity of formaldehyde.
27
28 4.3.1.4. Single Strand Breaks
29 Formaldehyde has been shown to induce DNA single strand breaks in a number of
30 mammalian cell systems in vitro as well as in vivo as shown in Table 4-78. In the in vitro
31 studies, most studies were positive, a few were negative and one study was equivocal for strand
32 breaks induced by formaldehyde. Among the in vivo studies, Im et al (2006) observed a dose-
33 dependent increase in DNA damage, analyzed by the Oliver tail movements in the comet assay
34 in both blood lymphocytes as well as livers of rats exposed to 5 and 10 ppm formaldehyde. Sul
35 et al (2007) also observed a dose-dependent increase in SSBs of Sprague-Dawley rats exposed to
This document is a draft for review purposes only and does not constitute Agency policy.
4-419 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-78. Formaldehyde-DNA interactions (single strand breaks)
Species/Strain
Cell/Strain
Result
References
DNA single strand breaks (in vitro)
Hamster/Chinese
Mouse
Rat
Rat
Rat/F344 trachea
Human
Human
V79 lung epithelial cells
Leukemia L1210 cells
Hepatocytes
Yoshida lymphosarcoma cells
Trachea! epithelial cell/Primary culture
Bronchial cell/Skin fibroblast
Peripheral blood lymphocytes
HeLa cells
Skin fibroblast
Lung/bronchial epithelial cells
Lung/bronchial epithelial cells
Skin keratinocytes/fibroblasts
-
(+)
-
+
+
+
+
+
+
+
+
+
+
+
+
-
Speit et al 2007a
Ross and Shipley, 1980
Ross et all 981
Demkowicz-Dobrzanski and
Castonguay 1992
O'Connor and Fox 1987
Cosmaetal., 1988
Grafstrometal., 1984
Liu et al 2006
Liu et al 2006
Snyder and Van Houten, 1986
Saladinoetall985
Grafstrom et al 1984
Grafstrom 1990
Fornace et al 1982
Vock et al 1999
Emri et al 2004
DNA single strand breaks (in vivo)
Mouse
Rats/Sprague-Dawley
Rats/Sprague-Dawley
Liver (maternal)
Liver (fetal)
Lung epithelial cells
Peripheral blood lymphocytes and liver
+
+
+
+
Wang and Liu 2006
Wang and Liu 2006
Sul et al 2007
Im et al 2006
4
5
6
7
8
9
10
11
12
13
'no' indicates test was not done in vivo; '+' indicates a positive test result; 'yes' indicates test was done in vivo;
indicates a negative test result; (+) indicates a weak positive test result.
0, 5, and 10 ppm formaldehyde by inhalation for 2 weeks (6 hrs/day, 5 days/wk). Additionally,
there is some evidence that DNA single strand breaks (SSBs) may be induced directly by
formaldehyde reactivity (Grafstrom et al., 1984). Thus, formaldehyde has been shown to induce
dose-dependent DNA damage in vivo in rodents. Besides, formaldehyde has also been shown to
induce SSBs in cultured human cells such as lung/bronchial epithelial cells (Vock et al 1999;
Saladino et al 1985; Grafstrom et al 1984; Grafstrom 1990; Fornace et al 1982), skin fibroblasts
This document is a draft for review purposes only and does not constitute Agency policy.
4-420 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
(Grafstrom et al., 1984; Snyder and van Houten, 1986; Emri et al., 2004) and HeLa cells (Liu et
al., 2006).
4.3.1.5. Other Genetic Effects of Formaldehyde in Mammalian Cells
Formaldehyde induces several other genetic and related effects in mammalian cells which
are evaluated by in vitro assays such as unscheduled DNA synthesis (UDS), DNA repair
inhibition and cell transformation as summarized in Table 4-79.
9
10
Table 4-79. Other genetic effects of formaldehyde in mammalian cells
Species/Strain
Cell/Strain
Result
References
Unscheduled DNA synthesis (UDS)
Rat/F344
Hamster/Syrian
Human
Nasal epithelial cells
Nasal epithelial cells
Hepatocytes
Embryo cells
HeLa cells
Bronchial epithelial cells
+
+
+
+
+
-
Bermudez and Allen 1984
Bermudez and Delehanty 1986
Williams et all 989
Hamaguchi and Tsutsui 2000
Martinet all 978
Doolittle et al 1985
DNA repair inhibition
Human
Bronchial epithelial cells/skin
fibroblasts
Normal fibroblasts (MRC5CV), XPA
cell line, and FA cell line
Skin fibroblasts/keratinocytes
+
+
+
Grafstrom et all 984
Speit et al 2000
Emri et al 2004
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
XPA, xeroderma pigmentosum, complementation group A (deficient in NER pathway).
FA = Fanconi's anemia (cell line has genetic defect leading to hypersensitivity to DNA-DNA cross links; NER =
nucleotide excision repair.
UDS, which represents DNA repair activity following excision of DNA damage, has
been reported in nasal epithelial cells of F344 rats (Bermudez and Allen 1984; Bermudez and
Delahanty 1986), rat hepatocytes (Williams et al 1989) and Syrian hamster embryo cells
(Hamaguchi and Tsutsui 2000) exposure to formaldehyde. UDS was observed in HeLa cells
(Martin et al 1978), but not in human bronchial epithelial cells (Doolittle et al 1985) upon
formaldehyde exposure. These studies suggest that following formaldehyde-induced DNA
damage was followed by DNA repair.
Studies involving human bronchial epithelial cells and skin fibroblasts or keratinocytes
(Grafstrom et al 1984; Emri et al 2004), DNA repair proficient or -deficient cell lines (e.g.
This document is a draft for review purposes only and does not constitute Agency policy.
4-421 DRAFT—DO NOT CITE OR QUOTE
-------
1 xeroderma pigmentosum) or cell lines hypersensitive to DNA-DNA crosslinks (e.g., Fanconi's
2 anemia) (Speit et al 2000) have been shown that formaldehyde causes DNA repair inhibition at a
3 concentration range of 0.125 mM to 10 mM. Emri et al (2004) have shown that DNA repair was
4 inhibited in human keratinocytes and fibroblasts after irradiation with UVB and UVC, but not
5 UVA followed by treatment with low concentrations of formaldehyde (10 |jM). UVA, UVB,
6 and UVC induced wavelength- and time-specific changes of DNA migration in comet assay.
7 They observed that DNA SSB induced by UVB or UVC irradiation alone were repaired within
8 3-6 hours of exposure, while cells with UV irradiation followed by formaldehyde exposure still
9 had the strand breaks at the same timepoints suggesting that formaldehyde is likely to contribute
10 to UV-induced carcinogenesis.
11
12 4.3.2. In Vitro Clastogenicity
13 Clastogenic effects, including increased MNs, CAs, and SCEs are also reported in a range
14 of in vitro study systems as shown in Table 4-80.
15 Several studies demonstrated formaldehyde-induced chromosomal aberrations (CAs) in a
16 variety of mammalian cells, such as CHO cells (Garcia et al 2009; Natarajan et al 1983), Chinese
17 hamster lung fibroblasts (Ishidate et al., 1981), Syrian hamster embryo cells (Hagiwara et al.,
18 2006; Hikiba et al., 2005), mouse lymphoma cells (Speit and Merk 2002) and human peripheral
19 blood lymphocytes (Dresp and Bauchinger, 1988; Schmid et al 1986; Miretskaya and
20 Shvartsman 1982) and fibroblasts (Levy et al., 1983).
21 Miyachi and Tsutsui (2005) measured the induction of sister chromatid exchanges
22 (SCEs) in Syrian hamster embryo (SHE) cells. Cells were exposed to 0, 3.3, 10, and 33 |jM
23 formaldehyde for 24 hours. SCE levels after 3.3 (jM formaldehyde were not different from
24 controls, but significant increases were observed at both 10 and 33 |jM. Toxicity as measured by
25 reduced cloning efficiency was seen only at 33 (jM (Miyachi and Tsutsui, 2005). The same
26 laboratory used SHE cells to measure the induction of CAs (Hikiba et al., 2005). Cells were
27 exposed to 0, 33, 66, and 99 |jM formaldehyde for 24 hours prior to staining for analysis and the
28 percentages of aberrant metaphases were 0, 6, 6, and 71, respectively. The aberrations were
29 predominantly chromosome gaps and chromosomal breaks and exchanges. The relative colony-
30 forming efficiency remained high (at least 85%) for the concentrations of formaldehyde used in
31 the experiment (Hikiba et al., 2005).
32
33
34
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-422 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-80. In vitro clastogenicity of formaldehyde
Species
Cell/Tissue origin
Without
activation
With
activation
References
Cytogenetic Assays
Chromosomal aberrations (CA)
Hamster/Chine se
Hamster/Syrian
Mouse
Human
Ovary cells
Lung fibroblasts
Embryo cells
Lymphoma cells
Peripheral blood lymphocytes
Fibroblasts
(+)
-
+
+
+
+
+
+
+
+
+
+
(+)
ND
ND
ND
ND
ND
ND
+
+
ND
ND
ND
Galloway etal, 1985
Dresp and Bauchinger,
1988
Natarajan et al 1983
Garcia et al 2009
IshidateJr etal 1981
Hikiba etal 2005
Hagiwara et al 2006
Speit and Merk 2002
Schmid etal 1986
Miretskaya and
Shvartsman 1982
Dresp and Bauchinger,
1988
Levy etal 1983
Micronucleus (MN)
Hamster/Chine se
Human
V79 lung epithelial cells
Fibroblasts
Whole blood cultures
Human MRC5CV (normal) and
XP(Repair-deficient) and FA
(repair-deficient) cell lines
+
+
+
+
+a
ND
ND
ND
ND
ND
Speit et al 2007b
Merk and Speit 1998
Emri et al 2004
Schmid and Speit 2007
Speit et al 2000
Sister Chromatid Exchange (SCE)
Hamster/Chinese
Hamster/Chinese
Ovary cells
V79 lung epithelial cells
(+)
+
+
+
+
+
+
+
(+)
ND
ND
ND
(+)
ND
ND
ND
Galloway etal., 1985
Natarajan et al 1983
Garcia et al 2009
Obe and Beek 1979
Easier etal. 1985
Speit et al 2007b
Merk and Speit 1998, 1999
Neuss and Speit 2008
This document is a draft for review purposes only
4-423
and does not constitute Agency policy.
DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-80. In vitro clastogenicity of formaldehyde (continued)
Species
Hamster/Syrian
Human
Cell/Tissue origin
Embryo cells
A549 lung epithelial cells
A549 + V79 (cocultivated)
A549 + V79 (cocultivated)
Lymphocytes
Whole blood cultures
Without
activation
+
+
+c
d
+b
+
+
+
+
With
activation
ND
ND
ND
ND
ND
ND
ND
ND
ND
References
Miyachi and Tsutsui 2005
Neuss and Speit 2008
Neuss and Speit 2008
Neuss and Speit 2008
Garry etal., 1981
Krieger and Garry 1983
Schmid etal 1986
Obe and Beek 1979
Schmid and Speit 2007
Premature chromosome Condensation (PCC)
Hamster/Chinese
Ovary cells
+
ND
Dresp and Bauchinger,
1988
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
a MN frequency increased in repair-deficient cell lines compared to normal cell lines.
b indicates SCE with significant loss of cell viability.
0 A549 cells exposed for 1 h with formaldehyde then cocultivated with V79 cells.
d A549 cells exposed for 1 h with formaldehyde, cells washed and then cocultivated with V79 cells.
'+' indicates a positive test result.
'ND' indicates test was not done.
- indicates a negative test result.
(+) indicates a weak positive test result.
XP, xeroderma pigmentosum; FA = Fanconi's anemia.
Schmid and Speit (2007) observed that SCEs were induced in lymphocytes of whole
blood cultures at a formaldehyde concentration of 200 jiM, an effect apparently associated with
cytotoxicity. This was indicated by a concomitant reduction in the proliferative index. These
authors also observed the formation of MNs in their cultures. This effect was statistically
significant at a formaldehyde concentration of 300 jiM and above. However, MN formation was
confined to those cultures in which formaldehyde treatment commenced 44 hours after the start
of the culture. This prompted the conclusion that the level of DPX formation would need to be
high for MN formation and that the cells must be exposed after the first mitosis. In examining
MN formation more closely, Schmid and Speit (2007) used the FISH technique, employing a
"biotin-labeled pan-centromeric chromosome paint specific for all human centromeres."
This document is a draft for review purposes only and does not constitute Agency policy.
4-424 DRAFT—DO NOT CITE OR QUOTE
-------
1 Indicative that formaldehyde was inducing a clastogenic (rather than aneugenic) effect, 81% of
2 MNs in binucleated cells were centromere-negative.
3 In summary, formaldehyde forms MNs, SCEs, and CAs in isolated animal and human
4 cells following in vitro exposure (see Table 4-80).
5
6 4.3.3. In Vitro Mutagenicity
7 Mutations may occur during repair of formaldehyde-induced DNA damage (DPXs, DNA
8 adducts, SSBs, or clastogenic effects) or as a result of replication errors during mitogenesis. The
9 in vitro evidence for formaldehyde-induced mutations is strengthened by examining the
10 correlation between these genotoxic and clastogenic events and induction of mutations.
11 Therefore studies are presented with respect to relevance to one or more of the following lines of
12 evidence for mutagenicity recommended for consideration in the EPA guidance (U.S. EPA,
13 2005a): (1) that the chemical is DNA reactive and/or has the ability to bind to DNA, (2) that the
14 chemical generates positive results in in vitro mutagenic test systems (specifically gene
15 mutations and CAs), and (3) that the chemical induces indications of genetic damage in in vivo
16 tests (specifically gene mutations and CAs). Numerous studies have demonstrated
17 formaldehyde-induced DNA mutations under a variety of experimental conditions (reviewed in
18 IARC 1995, 2006; Ma and Harris 1988; Auerbach et al. 1977; Conaway et al 1996; NTP 2009).
19
20 4.3.3.1. Mutagenicity in Bacterial Systems
21 A number of research reports describe the mutagenicity of formaldehyde in bacterial test
22 systems using reverse and forward mutation assays as well as specific strains detecting deletions,
23 insertions and point mutations. Among the bacterial strains, Salmonella typhimurium TA102 and
24 the Escherichia coll strains containing an AT base pair at the primary reversion site are often
25 used to detect oxidative compounds, cross-linking agents and hydrazines. In an early National
26 Toxicology Program (NTP) collaborative study with three laboratories, formaldehyde
27 consistently tested positive for mutagenicity in Salmonella typhimurium strain TA100 in the
28 presence of a rat or hamster liver S9 activating system (Haworth et al., 1983). Formaldehyde
29 was mutagenic with and without metabolic activation in a number of other studies using reverse
30 mutation assays with S. typhimurium strains TA98, TA100, TA102, TA104, TA2638, and
31 TA2638a and E. coli strains WP2 (pkMlOl), WP2 wvrA (pkMlOl), and hrs/r30R (Ryden et al.,
32 2000; Dillon et al., 1998; Watanabe et al., 1996; Le Curieux et al., 1993; O'Donovan and Mee,
33 1993; Zielenska and Guttenplan, 1988; Schmid et al., 1986; Connor et al., 1983, 1985; Orstavik
34 and Hongslo, 1985; Takahashi et al., 1985; Fiddler et al., 1984; Frei et al., 1984; Donovan et al.,
35 1983), while other studies (Muller et al., 1993; Jung et al., 1992; Wilcox et al., 1990; Marnett et
This document is a draft for review purposes only and does not constitute Agency policy.
4-425 DRAFT—DO NOT CITE OR QUOTE
-------
1 al., 1985) show both positive and negative results. These results are summarized in Table 4-81
2 and some of the studies are described in greater detail.
3 Formaldehyde has been shown to be mutagenic in forward mutation assays using S.
4 typhimurium (Couch et al 1982; Donovan et al 1983; Temcharoen and Thilly 1983) as well as in
5 E. coli (Bosworth et al 1987). Temcharoen and Thilly (1983) examined the toxicity and
6 mutagenicity of S. typhimurium strain TM677, using forward mutation to 8-azaguanine
7 resistance, and have shown that formaldehyde induced both toxicity and mutagenicity at
8 minimum concentrations of 0.17 mM (-S9) and 0.33 mM (+S9). It has also been shown that
9 formaldehyde formed as an intermediate by oxidation at the methyl group of
10 7V-nitrodimethylamine, a biologically active 7V-nitramine of environmental significance, is
11 mutagenic to S. typhimurium TA100 strain at low concentrations and toxic above 2 (jmol/plate
12 (Freietal., 1984).
13 Bosworth et al (1987) developed a forward mutation assay in E. coli D494 uvrB strain
14 transformed with a multi-copy mutator plasmid pGW1700, in which mutations are scored by an
15 increase in ampicillin-resistant colonies after exposure of bacterial cells during the logarithmic
16 growth by the test chemicals. This assay is more sensitive to base-pair substitutions, but less
17 sensitive to frameshift mutations compared to Salmonella/microsome-based assays. In this
18 assay, the authors (Bosworth et al 1987) observed positive curvilinear response to formaldehyde
19 exposure. Crosby et al (1988) used four E. coli strains GP120, GP120A, 7-2, and 33694
20 containing the xanthine guanine phosphoribosyl transferase (gpf) gene (which detects point
21 mutations, deletions and insertions) tested the mutagenicity of formaldehyde by exposing for 1
22 hour at 4 and 40 mM concentrations. They observed 41% large insertions, 18% large deletions
23 and 41% point mutations. However, at 40 mM dose there were 92% point mutations, a majority
24 of them (62%) being transition mutations at a single AT base pair in the gpt gene. In the same
25 study they observed frameshift mutations in E. coli that was transformed with naked pSV2gpt
26 plasmid DNA exposed to 3.3 or 10 mM formaldehyde. Thus, the mutation pattern appear to
27 differ depending on the concentration of formaldehyde exposure to the bacterial strain as well as
28 the nature of DNA.
29 Formaldehyde has also been shown to induce primary DNA damage in E. coli and
30 mutagenic activity in the Ames fluctuation test in S. typhimurium TA100, TA102, or TA98
31 strains (Le Curieux et al., 1993).
32
33
34
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-426 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-81. Summary of mutagenicity of formaldehyde in bacterial systems
Species
Strain
Metabolic
activation
+S9
-S9
References
Mutagenicity Assays
Reverse Mutation
S. typhimurium
E. coli
S. typhimurium
E. coli
TA98, 100, 1535, 1537, 1538
TA100
TA100
TA1535, 1537
TA98
TA98, TA100
UTH8414,UTH8413
TA97, 98, 100
TA102
TA100
TA100
TA100
TA98
UTH8414, UTH8413
TA100
TA97, 98, 1535, 1537, 1538
TA98, 100, 102
WP2, WP2 uvrA
H/R30R, HSSORuvrA
NG30rec4 016polA
TA97, 98, 100
TA102, 104
TA98, 100
TA100
TA104
TA102
WP2Mvrv4/(pKM101)
WP2 (pKMlOl)
-
ND
+
-
(+)
+
-
+
+
+
ND
+
(+)
-
(+)
-
ND
ND
ND
ND
ND
ND
+
+
+
ND
ND
ND
-
(+)
-
-
-
+
-
+
+
ND
+
(+)
-
-
-
-
(+)
+
+
-
-
+
+
+
ND
-
+
-
De Flora, 1981
Couch etal., 1982
Haworth et al., 1983
Haworth et al., 1983
Haworth et al., 1983
Connor etal., 1983a
Connor etal., 1983 a
Donovan et al., 1983
De Flora etal., 1984
Freietal., 1984
Fiddler etal., 1984
Connor etal., 1985
Connor etal., 1985
Connor etal., 1985
Ashbyetal., 1985 b
Ashbyetal., 1985 b
Takahashietal., 1985
Takahashietal., 1985
Takahashietal., 1985
Takahashietal., 1985
Marnettetal., 1985
Marnettetal., 1985
Oerstavik and Hongslo, 1985
Schmidetal., 1986
Zielenska and Guttenplan, 1988
Wilcoxetal., 1990
Wilcoxetal., 1990
Wilcoxetal., 1990
This document is a draft for review purposes only and does not constitute Agency policy.
4-427 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-81. Summary of mutagenicity of formaldehyde in bacterial systems
(continued)
Species
S. typhimurium
E. coli
S. typhimurium
E. coli
Strain
TA102
TA102
TA102
TA98, 100, 102
TA1535, 1537, 1538
WP2 (pKMlOl),
WP2uvrA (pKMlOl)
K12(AB1157)(WT)
K12(AB1886)/(uvrA),
K12(AB2480)/(recA/uvrA)
TA102, 2638
WP2 (pKMlOl), WP2uvrA
(pKMlOl)
Metabolic
activation
+S9
+
ND
+
ND
ND
ND
ND
ND
ND
ND
-S9
ND
+
ND
+
-
+
+
-
+
+
References
Jungetal., 1992
Le Curieux et al., 1993
Muller et al., 1993
O 'Donovan and Mee, 1993
O 'Donovan and Mee, 1993
O 'Donovan and Mee, 1993
O 'Donovan and Mee, 1993
Graves etal., 1994
Graves etal., 1994
Watanabe et al., 1996
Watanabe et al., 1996
Mutagenicity Assays
S. typhimurium
E. coli (Lac+ reversion)
S. typhimurium
TA1535 c
TA1537 c
TA98, 100 c
TA97C
TA1535, 1537 d
TA98d
TA100 d
TA100 e
TA100, 104 d
WP3101P, WP3106P
TA102, 2638 d
-
+
+
ND
-
+
-
+
+
+
ND
-
+
-
+
-
+
+
+
+
+
Sarrifetal., 1997
Sarrifetal., 1997
Sarrifetal., 1997
Sarrifetal., 1997
Sarrifetal., 1997
Sarrifetal., 1997
Sarrifetal., 1997
Sarrifetal., 1997
Dillon etal., 1998
Ohta et al 1999
Ryden et al., 2000
Forward Mutation
S. typhimurium
E. coli
TM677
D494wvr5
ND
+
+
+
(+)
+
+
Couch etal., 1982
Donovan et al., 1983
Temcharoen and Thilly, 1983
Bosworthetal 1987
This document is a draft for review purposes only and does not constitute Agency policy.
4-428 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-81. Summary of mutagenicity of formaldehyde in bacterial systems
(continued)
Species
Strain
Metabolic
activation
+S9
-S9
References
Deletion Mutation
E. coli
GP120, GP120A 7-2, 33694
ND
+ f
Crosby etal., 1988
Point Mutation
E. coli
GP120, GP120A 7-2, 33694
ND
+
Crosby etal., 1988
Insertion Mutation
E. coli
GP120, GP120A 7-2, 33694
ND
+
Crosby etal., 1988
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
a Indicates the use of formalin in mutagenicity assay.
b Indicates the use of hexamethylmelamine (HEMLA), a formaldehyde-releasing compound, in mutagenicity assay.
0 Indicates use of the Standard Plate Method.
d Indicates use of the Preincubation Plate Method.
e Indicates use of the Suspension Method.
f Indicates loss of DNA.
'+' indicates a positive test result.
'ND' indicates test was not done.
'-' indicates a negative test result.
(+) indicates a weak positive test result.
O'Donovan and Mee (1993) observed clear mutagenicity by the preincubation exposure
method in S. typhimurium TA98, TA100, and TA102 strains and bothE. coli WP2(pKM101) and
WP2uvrA(pKM101) strains, while the standard plate-incorporation assays showed consistent
mutagenicity only with TA100 and WP2wvrA(pKM101) strains and no evidence of mutagenicity
in TA1535, TA1537, or TA1538 strains using either method of exposure in the absence of
metabolic activation. The S. typhimurium andE. coli strains used in this study are histidine and
tryptophan auxotrophs, with an AT base pair at the critical mutation site within the hisG and trpE
genes, respectively, with an intact excision repair system facilitating the detection of cross-
linking agents and both strains carrying the mutator plasmid, pKMlOl, which enhances error-
prone repair. These salmonella strains detect frameshift (TA98 and TA1537) and base-pair
substitutions (TA100, TA102, and TA1535), while theE. coli strains detect base-pair
substitutions (WP2z/vrA). These findings are consistent with the suggestion that formaldehyde
induces excision-repairable lesions in bacteria and indicate that the presence of the R-factor
plasmid may be required for the expression of its mutagenicity in excision repair-deficient
salmonella (O'Donovan and Mee, 1993).
This document is a draft for review purposes only and does not constitute Agency policy.
4-429 DRAFT—DO NOT CITE OR QUOTE
-------
1 Dillon et al. (1998) employed salmonella strains TA100, TA102, and TA104 because of
2 the latter two strains being more sensitive to oxidative mutagens. Formaldehyde was clearly
3 mutagenic between 6 and 50 |j,g/plate in all three strains with and without metabolic activation
4 using Aroclor-induced S9 from male F344 rats or male B6C3F1 mice, except for an equivocal
5 response in TA102 with mouse S9 (Dillon et al., 1998). Using a set of six tester strains
6 (WP3101-WP3106) of E. coli, each reversible by a mutation involving a single DNA base pair
7 substitution, Ohta et al. (1999) determined that formaldehyde preferentially induced GC to TA
8 transversion mutations. Ryden et al. (2000) demonstrated a statistically significant increase in
9 the number of revertants in S. typhimurium TA102 (2.5-fold) and TA2638a (3-fold) strains by
10 formaldehyde at >17 |j,g/plate compared with solvent controls.
11 In summary, formaldehyde induces mutations in several bacterial strains containing an
12 AT base pair at the primary reversion site that are used to detect oxidative compounds and cross-
13 linking agents without metabolic activation by exogenous enzyme-activating systems. This
14 evidence is strengthened by examining the correlation between genotoxic and clastogenic events
15 and mutation induction.
16
17 4.3.3.2. Mutagenicity in Other Nonmammalian Cell Systems
18 Formaldehyde has been shown to be mutagenic in several nonmammalian systems also.
19 It has been shown to cause gene conversion, strand breaks, crosslinks, homozygosis and related
20 damage in yeasts (Saccharomyces cervisiae), forward and reverse mutations in molds
21 (Neurospora crassa), micronuclei formation in spiderworts (Tradescantiapallida), DNA
22 damage and mutations in several plants, genetic cross-over or recombination, sex-linked
23 recessive lethal mutations, dominant lethal mutations, heritable translocations and gene
24 mutations in insects (Drosophila melanogaster) and recessive lethal mutations in nematodes
25 (Caenorhabditis elegans), but failed to show micronuclei formation in newt larvae (Pleurodeles
26 waltl) (Reviewed in Conaway 1996; IARC 2006).
27 4.3.3.3. Mutagenicity in Mammalian Cell Systems
28 Several studies demonstrated the mutagenicity of formaldehyde in mammalian cells. In
29 its report, the Federal Panel on Formaldehyde underlined the role of formaldehyde as an inducer
30 of gene mutations and CA in a variety of test systems (Report of the Federal Panel on
31 Formaldehyde, 1982). Results from several studies are summarized in Table 4-82.
32
This document is a draft for review purposes only and does not constitute Agency policy.
4-430 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-82. Mutagenicity in mammalian cell systems
Species/Strain
Cell/Strain
In Vivo
test
Metabolic
activation
-S9
+S9
References
Mutagenicity Assays
Dominant Lethal Mutation
Rat/Albino
Rat
Mouse
Spermatocyte, Live implants
Dominant lethal
Dominant lethal
yes
yes
yes
yes
yes
+
(+)
-
-
(+)
ND
Odeigah, 1997
Kitaevaetall990
Epstein and Shafner 1968
Epstein et al 1972
Fontignie-Houbrechts 1981
Deletion Mutation
Hamster/Chinese
Mouse
Human
V79 cells (Hprt locus)
V79 cells Hprt locus)
V79/HPRT
Ovaiy HPRT
Lymphoma L5 178Y cells
(Tk+A locus)
Lymphoma L5 178Y cells
Bronchial cell
Bronchial
fibroblasts/epithelial cells
(HPRT locus)
Bronchial
fibroblasts/epithelial cells
(HPRT locus)
Lymphoblast/HPRT
Lymphoblast/tk
Peripheral blood
lymphocytes
Lymphoblast (TK6)
no
no
no
no
no
no
no
no
no
no
no
yes
no
-
-
+
-
+
+
+
+
+
+a
+
+
+
ND
+
ND
ND
ND
Merk and Speit 1998
Merk and Speit 1999
Graf strom etal., 1993
Graves etal., 1996
Macerer et al 1996
Speit and Merk 2002
Graf strom etal., 1983
Grafstrom etal 1985
Grafstrom 1990
Crosby etal., 1988
Craft etal 1987
Shaham et al 2003
Goldmacher and Thilly 1983
Point Mutation
Hamster/Chinese
Mouse
Mouse
Human
Ovary HPRT
Lymphoma cell/TK+/-
Lymphoma cell/TK+/-
Lymphoblast/TK6
no
no
no
no
+
+
+
+
ND
+
ND
ND
Graves etal., 1996
Blackburn et al., 1991
Wangeheim and Bolcsfoldi, 1988
Liber etal., 1989
This document is a draft for review purposes only and does not constitute Agency policy.
4-431 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-82. Mutagenicity in mammalian cell systems (continued)
Species/Strain
Cell/Strain
In Vivo
test
Metabolic
activation
-S9
+S9
References
Insertion Mutation
Hamster/Chinese
Heritable Mutation
Mouse
Ovary HPRT
Heritable mutation
no
yes
+
+
ND
Graves etal., 1996
Liu et al 2009
DNA Repair enzyme activity
Human
Peripheral lymphocyte
yes
-
Hayes etal., 1997
Mutagenicity Assays
Cell Transformation
Mouse
Hamster
C3H10T1/2 cells
Embryo
fibroblast/C3H/10Tl/2
Embryo
fibroblast/C3H/10Tl/2
Kidney cell/BHK-21/cL13
no
no
no
-S9
+b
[+]
[+]
+
+S9
ND
ND
+
Ragan and Boreiko 1981
Boreikoetal., 1983
Frazelleetal., 1983
Plesner and Hansen, 1983
p53 mutation and/or p53 protein expression
Rats/F344
Rats/F344
Rats/F344
Human
Nasal squamous cell
carcinomas
Nasal tumor cell lines
Nasal squamous cell
carcinomas
Peripheral blood
lymphocytes
yes
No
Yes
yes
+c
+
+d
+
Recioetal 1992
Bermudez etal 1994
Wolf etal 1995
Shaham et al 2003
4
5
6
7
8
9
10
11
12
13
14
15
16
17
a indicates loss of DNA.
b Positive only in the presence of 12-O-tetradecanoylphorbol 13-acetate (TPA).
0 p53 mutations.
d p53 mutated protein detected by immunohistochemistry.
'no' indicates test was not done in vivo
'+' indicates a positive test result.
'ND' indicates test was not done.
'yes' indicates test was done in vivo.
- indicates a negative test result.
(+) indicates a weak positive test result.
[+] indicates positive test result after TPA or A^-methyl-A^-nitro-A^-nitrosoguanidine promoter treatment.
Snyder and Van Houten (1986) demonstrated that formaldehyde increases the levels of
misincorporation of bases into synthetic polynucleotides catalyzed by E. coli DNA polymerase I,
This document is a draft for review purposes only and does not constitute Agency policy.
4-432 DRAFT—DO NOT CITE OR QUOTE
-------
1 indicating that the mutagenicity of formaldehyde may be due to covalent alteration of DNA
2 bases. They have also shown that formaldehyde-induced DNA damage in human fibroblasts was
3 not susceptible to repair by the typical "long patch" excision repair mechanism.
4 Craft et al. (1987) measured the induction of mutations at the thymidine kinase (tk~) locus
5 or at the ouabain resistance (Oucf) locus in TK6 human lymphoblastoid cells. The tk mutations
6 can result from a variety of mutational events, including base pair substitution, small and large
7 deletions, and chromosome exchange events, while mutations to Oucf require specific base pair
8 substitutions. Single treatment of formaldehyde (0, 15, 30, 50, 125, and 150 uM) for 2 hours
9 resulted in a nonlinear increase in tk mutagenesis with increasing slope >125 uM (see
10 Figure 4-31). To explore a dose-response effect, cells were also exposed as follows: three
11 treatments of 5 0 uM for 2 hours or five treatments of 3 0 uM or 10 treatments of 15 uM for
12 2 hours (treatments were spaced 2-4 days apart) with multiple treatments causing an increase in
13 tk mutations, although their combined effect was less than a single treatment of equivalent C x t
14 (150 uM for 2 hours). Lymphoblasts given four treatments of 150 uM formaldehyde for 2 hours
15 failed to induce mutations at the Oucf locus. Dose-response increases were seen in all exposure
16 scenarios, with 30 uM being the level of statistical significance. There was little indication of a
17 dose-response effect until the cumulative concentration was greater than 100 uM.
18 Formaldehyde-induced DPXs were no longer evident after 24 hours of exposure; mutants
19 induced in the TK6 lymphoblast cell line showed a similar dose-response curve to the DPXs
20 measured immediately after exposure ended (Craft et al., 1987).
21 The same group also studied mutations induced at the X-linked hypoxanthine-guanine
22 phosphoribosyl transferase (FtPRT) locus by eight repetitive treatments of 150 |jM formaldehyde
23 in TK6 human lymphoblast cell line by Southern blot analysis, wherein half (14/30) of induced
24 mutants contained partial or complete deletions with most of the partial deletions showing
25 unique deletion patterns, while only a third (5/15) of spontaneous mutants had partial or
26 complete deletions, indicating that formaldehyde can induce large losses of DNA in human
27 lymphoblast cells (Crosby et al., 1988).
28 Liber et al. (1989) followed up the findings of Crosby et al. (1988) by performing
29 Southern blot, Northern blot, and DNA sequence analysis on the 16 induced and 10 spontaneous
30 human lymphoblast mutants not showing deletions. Northern blot analysis showed that the point
31 mutations fell into four categories: normal size and amount of RNA, normal size but reduced
32 amounts of RNA, reduced size and amounts of RNA, and no RNA. Sequence analysis of
33 recombinant DNAs from hprt mRNA in formaldehyde-induced mutants showed a preferential
34 AT to CG transversion at a specific site, with other changes represented to a lesser degree (Liber
35 etal., 1989).
This document is a draft for review purposes only and does not constitute Agency policy.
4-433 DRAFT—DO NOT CITE OR QUOTE
-------
— n— DPX
MF
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
80
72
64
56
48
40
32
24
16
8
0
30 60 90
Formaldehyde concentration
120
150
Figure 4-31. DNA-protein cross-links (DPX) and thymidine kinase (tk)
mutants in TK6 human lymphoblasts exposed to formaldehyde for 2 hours.
Note: D DPXs immediately after exposure, BDPXs 24 hours after exposure, A tk
mutants. Relative survival was 100% at 0, 15, 30, and 50 |iM, 30% at 125 |iM,
and 20% at 150 |iM.
Source: Adapted from Craft et al. (1987).
Even in CHO cells formaldehyde has been shown to induce hprt mutations involving
mostly single-base pair transversions mostly occurring at AT sequences, including three AT to
TA at position 548 of exon 8 and two AT to CG and one GC to TA transversion at other sites
(Graves et al., 1996). In another study, formaldehyde-induced forward mutations to
trifluorothymidine resistance in mouse lymphoma L5178Y tk* cells both in the absence and
presence of rat liver S9 (higher concentrations required for effect with S9). Both toxicity and
mutagenicity were abolished when FADH was incorporated in the exposure medium (Blackburn
etal., 1991).
Formaldehyde-induced DPXs are removed in part through spontaneous hydrolysis and in
part due to active repair processes (Quievryn and Zhitkovich, 2000). Inhibition of specific
proteosomes in XP-A cells inhibited DPX repair, thereby supporting the role of enzymatic
degradation (Quievryn and Zhitkovich, 2000). The half-life of formaldehyde-induced DPXs in
This document is a draft for review purposes only and does not constitute Agency policy.
4.434 DRAFT—DO NOT CITE OR QUOTE
-------
1 human cell lines was consistent with the findings of Craft et al. (1987), ranging from 11.6 to
2 13 hours (Quievryn and Zhitkovich, 2000). In the same report, removal of DPXs from human
3 peripheral lymphocytes was much slower, with a half-life of 18.1 hours. This difference was
4 primarily in slower active repair of DPXs, with a ti/2 of 66.6 hours for human lymphocytes
5 versus 23.3 hours for human cell lines (Quievryn and Zhitkovich, 2000).
6 Since DPX repair involves proteolytic removal of proteins from the DNA, Speit et al.
7 (2000) hypothesized that single peptides or small peptide chains cross-linked to the DNA are
8 critical to formaldehyde-induced mutation. However, these authors did not find significant
9 difference in the induction and repair of DPXs in normal and DNA repair-deficient cell lines but
10 observed increased susceptibility of the repair-deficient cell lines to formaldehyde-induced MN
11 induction. In this study, a normal human cell line (MRC5CV1), a xeroderma pigmentosum cell
12 line deficient in nucleotide excision repair (NER), and a Fanconi anemia cell line, which has a
13 genetic defect leading to hypersensitivity towards DDXs, were exposed to 125, 250, and 500 |jM
14 formaldehyde for 2 hours. The authors suggest that more than one repair pathway is involved in
15 the repair of cross-links and that the altered NER pathway has more severe consequences to
16 formation of CAs than disturbed cross-link repair (Speit et al., 2000).
17 The correlation of early DPX formation and mutation is at first counterintuitive since the
18 cross-linking of protein to DNA inhibits DNA replication. Without active DNA replication,
19 formaldehyde-DNA adducts and DPXs would not induce replication error and would be unlikely
20 to result in a change in DNA sequence or mutation. Recent evidence indicates that residual
21 peptides and short polypeptides that remain cross-linked to DNA after DPX removal may in fact
22 be the cause of DPX-associated, formaldehyde-induced mutation (Speit et al., 2000).
23 A study by Merk and Speit (1998) indicated that formaldehyde-induced DPXs did not
24 result in direct gene mutations in the hprt locus of V79 Chinese hamster cells, leading the
25 authors to speculate that formaldehyde was not mutagenic. Since, the hprt locus in the V79
26 Chinese hamster cell line is primarily sensitive to point mutations and other studies show the
27 formation of deletion mutations by formaldehyde at the same locus in human lymphoblasts
28 (Crosby et al., 1988), Merk and Speit (1998) concluded that the hprt mutation assay is insensitive
29 to deletion mutations.
30 Later, using the mouse lymphoma assay, Speit and Merk (2002) demonstrated that
31 exposure to formaldehyde for 2 hours was mutagenic in a concentration-dependent manner in the
32 L5178Y mouse lymphoma cells, which was mainly contributed by a strong increase in small
33 colony mutants, suggestive of CAs (Speit and Merk, 2002). Detailed analysis of both
34 spontaneous and formaldehyde-induced lesions indicates that recombination or deletion of DNA
35 from the tk locus was primarily responsible for the loss of heterogeneity, thereby leading to the
This document is a draft for review purposes only and does not constitute Agency policy.
4-43 5 DRAFT—DO NOT CITE OR QUOTE
-------
1 observed mutant phenotype. Therefore, it is believed that formaldehyde is mutagenic in the
2 L5178Y cell mouse lymphoma system by a clastogenic mechanism rather than through point
3 mutations. This finding is consistent with that of Craft et al. (1987), who demonstrated
4 formaldehyde mutagenicity at the tk locus of TK6 cells, and also with the findings of Grafstrom
5 et al. (1984), who demonstrated increased SSB formation in formaldehyde-exposed cell lines.
6 Formaldehyde has also been shown to induce cell transformation in mouse embryo
7 fibroblasts (Ragan and Boreiko 1981; Boreiko et al 1983; Frazelle et al 1983) . At low
8 concentrations of 0.017 mM formaldehyde has shown to cause cell transformation in C3H10T1/2
9 mouse cells (Ragan and Boreiko 1981) and hamster kidney cells in vitro (Plenser and Hansen
10 1983).
11 More recently, Shaham et al. (2003) examined the frequency of DPXs and the incidence
12 of mutant versus wild type p53 tumor suppressor genes in the peripheral blood lymphocytes of a
13 cohort of workers exposed to formaldehyde. The adjusted mean levels of DPXs were greater in
14 the lymphocytes of exposed subjects compared with those of unexposed subjects, and exposure
15 to formaldehyde increased the likelihood of their having a higher level of pantropic p53
16 (>150 pg/mL). The authors speculated on a possible causal relationship between DPXs and
17 mutations in p53. Recio et al (1992) demonstrated point mutations in the p53 tumor suppressor
18 gene in 45% (5 out of 11) of the primary nasal squamous cell carcinomas (SCCs) obtained from
19 F344rats that were chronically exposed to 15 ppm formaldehyde for 2 years (Recio et al ., 1992).
20 In summary, the results of in vitro experiments demonstrate the mutagenicity of
21 formaldehyde. Mutagenicity is observed below levels of significant cytolethality in mammalian
22 cell lines. Formaldehyde is clearly a DNA-reactive genotoxicant inducing lesions (DPXs) that
23 show clastogenicity (SSBs, MNs, etc.). The experiments by Speit and Merk (2002) explore
24 mechanistic links between DPXs, clastogenicity, and the observed locus-specific mutations in
25 the mouse lymphoma in vitro testing system.
26
27 4.3.4. In Vivo Mammalian Genotoxicity
28 4.3.4.1. Genotoxicity in Laboratory Animals
29 As discussed above, formaldehyde is clearly reactive at the POE in animal studies,
30 resulting in increased DPXs in the nasal mucosa. Despite formaldehyde's reactivity and
31 mutagenicity in isolated mammalian cells, clear evidence of mutagenicity does not emerge from
32 animal bioassays (see Table 4-83).
33
34
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-436 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-83. Genotoxicity in laboratory animals
4
5
6
7
8
9
10
11
12
13
14
15
16
Species/Strain
Cells/Organ/Tumor
Result
References
Cytogenetic Assays
Chromosomal aberrations (CA)
Mice/Q strain
Mice/CBA
Rats/F344
Rats/Sprague-Dawley
Rats/Wistar
Rats/Sprague-Dawley
Rats/Sprague-Dawley
Rats/F344
Spermatocyte
Spermatogonia
Polychromatic erythrocytes
Spleen cells
Lymphocytes
Gastric epithelial cells
Bone marrow
Bone marrow
Pulmonary lavage cells
Peripheral blood cells
-
-
-
-
-
+
+
-
+
-
Fontignie-Houbrechts et al., 1981
Fontignie-Houbrechts et al., 1982
Natarajan et al., 1983
Natarajan et al., 1983
Kligerman et al 1984
Miglioreetall989
Kitaevaetall990
Dallas et all 992
Dallas et all 992
Speit et al 2009
Micronucleus (MN)
Mouse/NMRI
Mice/CBA
Mice/B6C3Fl
Rats/Sprague-Dawley
Bone marrow
Femoral polychromatic
erythrocyte and spleen cell
Bone marrow
Gastric epithelial cells
-
-
+
+
Gockeetall981
Natarajan et al., 1983
Ward et all 983
Migliore et al 1989
Sister Chromatid Exchange (SCE)
Rats/F344
Rats/F344
Lymphocyte
Peripheral blood cells
-
-
Kligerman et al 1984
Speit et al 2009
'+' indicates a positive test result.
'-' indicates a negative test result.
In a chromosomal analysis study (Fontignie-Houbrechts, 1981), formaldehyde given IP.
at 50 mg/kg to male Q strain mice and analyzed 8-15 days after treatment did not induce any
chromosomal lesions in spermatocytes. Also, in another study from the same group (Fontignie-
Houbrechts et al., 1982), formaldehyde (30 mg/kg) given along with hydrogen peroxide (90
mg/kg) as a mixture to male Q strain mice failed to produce significant increases in
chromosomal lesions in the spermatogonia.
Ward et al. (1983) studied the cytogenetic effects of commercial formalin on the bone
marrow of B6C3F1 mice. Since commercial formalin contains 10-15% of methanol, another
group was dosed with methanol (1000 mg/kg) and also included a negative control (water) and a
This document is a draft for review purposes only and does not constitute Agency policy.
4-437 DRAFT—DO NOT CITE OR QUOTE
-------
1 positive control group dosed with 100 mg/kg cyclophosphamide (CP). Mice from all groups
2 were administered 1 mg/kg colcemid (which arrests cells in metaphase), 28 h after dosing with
3 test compounds. Two hours after colcemid treatment, mice were sacrificed; bone marrows from
4 femurs were prepared and analyzed the metaphase chromosomes for structural aberrations. They
5 reported that formalin, methanol, water and CP, respectively produced structural chromosomal
6 aberrations (%) such as aneuploidy (101, 65, 6, 3), breaks (3, 3, 1, 32), exchanges (40, 24,
7 10,17), aberrant chromosomes with gaps (45, 32, 2.1, 63) and without gaps (43, 27, 1.7, 55) in
8 bone marrow cells of male B6C3F1 mice. The cytogenetic effects seen in bone marrow suggest
9 that formalin or other chemicals given at a single high dose (gavage) were able to reach bone
10 marrow and induce genotoxicity, specifically, aneuploidy and sister chromatid exchanges. It is
11 recognized that the results reported were preliminary; however, they are significant in terms of
12 genotoxic potential of formaldehyde exposure via oral route of exposure.
13 In a different study Natarajan et al. (1983) failed to detect significant differences in MN
14 induction in bone marrow cells or CAs in spleen cells of male and female CBA mice given IP.
15 6.25, 12.5, and 25 mg/kg formaldehyde compared with saline-treated controls. However, the
16 same study showed a positive induction of MNs and CAs in vitro. The authors suggest that the
17 lack of genotoxicity in vivo may be due to the inability of formaldehyde to reach the target cells
18 in sufficient quantity to induce biological effects.
19 Kligerman et al. (1984) also found no difference in the incidence of SCEs or
20 chromosome breakage in the peripheral lymphocytes of male and female F344 rats exposed to
21 formaldehyde in air at 0.5, 6, or 15 ppm (0.61, 7.36, or 18.4 mg/m3) 6 hours/day for 5 days.
22 However, in a different study (Migliore et al., 1989), clastogenic effects, such as increased MNs
23 and CAs, were reported in GI epithelial cells of male Sprague-Dawley rats after oral exposures to
24 200 mg/kg formaldehyde. In this study, micronucleated cells and nuclear anomalies were
25 increased in a time-dependent manner in the stomach, duodenum, ileum, and colon of rats, and
26 the mitotic index was unchanged for these cells compared with controls at 16, 24, and 30 hours.
27 These clastogenic effects were seen without regenerative cell proliferation, supporting
28 formaldehyde-induced mutations as primary effects of formaldehyde rather than secondary to
29 regenerative cell proliferation.
30 Kitaeva et al. (1990) observed cytopathological and cytogenetic effects of formaldehyde
31 chronic inhalation in 0.5 and 1.5 mg/m3 doses in the female rat's germ cells and bone marrow
32 cells, where formaldehyde-induced harmful effects were seen in germ cells at <1.5 mg/m3 doses,
33 while the reliable clastogenic and cytogenetic effects on the marrow cells were induced even at
34 the 0.5 mg/m3 dose, suggesting differences among effects of small doses of formaldehyde on
35 different cell systems.
This document is a draft for review purposes only and does not constitute Agency policy.
4-43 8 DRAFT—DO NOT CITE OR QUOTE
-------
1 Dallas et al. (1992) observed a slight increase (7.6 and 9.2%) in CAs in pulmonary lavage
2 cells from male Sprague-Dawley rats exposed to 15 ppm (18.4 mg/m3) formaldehyde in air
3 6 hours/day, 5 days/week for 1 or 8 weeks by inhalation compared with corresponding controls
4 (3.5 and 4.8%), respectively. However, the small study, limited as it was to five animals/group,
5 showed statistically significant increase at the highest dose tested (15 ppm) but not at lower
6 doses (0.5 and 3 ppm). In the same study, no clastogenic effects were seen in bone marrow,
7 which is consistent with formaldehyde acting primarily at the site of first contact.
8 Speit et al (2009) investigated the genotoxicity of formaldehyde in peripheral blood
9 samples of Fischer-344 rats exposed to 0 to 15 ppm formaldehyde by whole-body inhalation for
10 4 weeks (6 h/day, 5 days/week). In this study, the authors found no significant increase in the
11 genotoxic assays such as comet assay with or without gamma-irradiation of blood samples (DNA
12 migration as determined by tail movement or intensity), sister chromatid exchange (SCE) assay
13 and micronucleus test (MNT) compared to controls. However, rats given 50 mg/kg
14 methylmethane sulfonate (MMS) by gavage for 4 hrs (positive control for Comet and SCE
15 assays) or 10 mg/kg cyclophosphamide (CP) given twice orally (positive control for MNT)
16 induced significant increase in genotoxicity in this study. The lack of genotoxicity in this study
17 is not surprising since earlier studies by Casanova-Schmitz et al (1984a) have shown that
18 formaldehyde does not cause toxicity to bone marrow possibly due to the inability of this
19 chemical to reach the bone marrow. Although MMS and CP used in this study were positive in
20 the genotoxicity assays, the data from positive controls can not be used for validation since the
21 exposure routes of formaldehyde (inhalation) and the positive controls (oral) were different.
22 No animal studies have examined clastogenic effects of formaldehyde in nasal or
23 respiratory epithelial cells. Therefore, it is unknown whether similar changes would occur in
24 response to exposure to formaldehyde via inhalation. However, the negative finding in bone
25 marrow cannot be considered definitive evidence on the question of the mutagenic potential of
26 formaldehyde for cells present at the POE. With weak positive results in pulmonary lavage cells
27 and clear clastogenicity in GI epithelial cells below exposures that trigger regenerative cell
28 proliferation, the existing evidence, however incomplete, supports the concept of genotoxic
29 action of formaldehyde at the POE.
30
31 4.3.4.2. Genotoxicity in Humans
32 The majority of the studies on the effects of formaldehyde in exposed humans have
33 measured various cytogenetic endpoints, such as MNs, SCEs, or CAs in nasal and oral mucosal
34 cells (considered to be in direct contact with formaldehyde) as well as peripheral lymphocytes.
35 Since genotoxicity at the proximal sites (oral, nasal) can be readily linked to the reactive nature
This document is a draft for review purposes only and does not constitute Agency policy.
4-439 DRAFT—DO NOT CITE OR QUOTE
-------
1 of formaldehyde, these studies are discussed first, noting where researchers also collected blood
2 lymphocyte samples. A subsequent discussion is focused on results in blood lymphocytes.
3 Finally, the few studies that measured DPXs in exposed humans are discussed. Table 4-89
4 provides a summary of human cytogenetic studies of formaldehyde.
5
6 4.3.4.2.1. Nasal, buccal, and oral mucosal cells.
1 Epithelial cells of the URT and oral cavity are potential targets of formaldehyde's DNA
8 reactivity and genotoxicity. Several studies indicate that formaldehyde exposure results in
9 measurable increases in SCEs, MN formation, and DPXs in nasal, buccal, and oral mucosal cells;
10 however, these genotoxic effects vary with the type of exposure. Study quality, sample size,
11 availability of exposure measurements, and assay methodology may in part contribute to
12 variability in study findings. The studies fall into three general categories: workers (industrial or
13 professional), students and staff attending anatomy and mortuary science courses, and subjects in
14 a controlled clinical trial.
15 Ballarin et al. (1992) observed significantly higher frequency of micronucleated cells in a
16 formaldehyde exposed group in a plywood factory compared with controls (0.9 ± 0.47 versus
17 0.25 ± 0.22,/> < 0.01). In this study, the frequency of MNs and cytology of respiratory nasal
18 mucosal cells was examined in 15 nonsmokers exposed to levels of formaldehyde that ranged
19 between 0.1 and 0.39 mg/m3 (-0.32 ppm) for an average of 6.8 years. Exposed subjects were
20 compared with age- and sex-matched controls.
21 Ye et al. (2005) reported significant increases in MNs per thousand cells in nasal mucosal
22 cells for 18 nonsmoking workers (2.70 + 1.50) in a formaldehyde manufacturing plant in the
23 Hugei province of China as compared with controls (1.25 + 0.41). In addition, higher
24 frequencies of SCEs in peripheral lymphocytes of workers were also reported (8.24 + 0.89 versus
25 6.38 + 0.41). In this study, the average age of workers was 29 + 6.8 years, the average duration
26 at work was 8.5 years (range 1-15 years), and the reported 8-hour TWA was 0.985 mg/m3
27 (0.8 ppm). The control group consisted of 23 undergraduate students with an average age of
28 19 + 2.3 years. The 8-hour TWA in the student dormitories was 0.011 mg/m3 (9 ppb). A group
29 of 16 waiters with an average exposure duration of only 12 weeks and an 8-hour TWA of
30 0.107 mg/m3 (90 ppb) was also included in the study. The incidence of MNs and SCEs in the
31 waiters was the same as that in controls. Overall, results from this study suggest that the
32 genotoxic potential of high-level formaldehyde exposure may have occupational risks in long-
33 term exposure.
34 However, in a different study, Speit et al. (2007b) showed that formaldehyde did not
35 induce MNs in exfoliated buccal mucosa cells of humans exposed up to a maximum of 1 ppm
This document is a draft for review purposes only and does not constitute Agency policy.
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
and a cumulative exposure of 13.5 ppm-hours over 2 weeks. In this study, volunteers exposed to
formaldehyde in closely controlled conditions (4 hours/day for 10 days) with a complex
exposure schedule, amounting to a cumulative total of 13.5 ppm-hours (16.6 mg/m3-hours), were
used. Samples of the buccal mucosa were taken from subjects 1 week before the start of the
experiment, at the start of the experiment, at the conclusion of the series of exposures, and at 7,
14, and 21 days after the completion of exposure. Thus, the subjects served as their own
controls. Two thousand cells per data point were assessed for the frequency of MNs on slides
that were coded by an independent quality assurance organization. As shown in Table 4-84, the
frequency of MN formation was statistically unchanged from that in controls. The apparent
slight increase in subjects evaluated at the conclusion of exposure was caused by frequencies of
MNs in two subjects (5.0 and 4.5 MNs per 1,000 cells). The data as reported show a high
variability, where the SD approaches or exceeds the mean for each sample point, suggestive of
data with an asymmetrical distribution.
Table 4-84. MN frequencies in buccal mucosa cells of volunteers exposed to
formaldehyde
Sampling point
Group
MN/1000 cells (± SD)
Control data
1 week before exposure
Immediately before exposure series
1
2
0.95 ±0.67
0.86 ±0.84
Test data
Immediately after exposure series
7 days after exposure
14 days after exposure
21 days after exposure
3
4
5
6
1.33 ±1.45
0.94 ±0.73
0.85 ±0.86
0.44±0.38a
"Statistically significantly different from control values (p < 0.05), as calculated by the authors.
Source: Speitetal. (2007b).
The best evidence of formaldehyde-induced clastogenic changes in peripheral
lymphocytes is found in studies of anatomy class and mortuary class students. Since genetic
damage accumulates with age, the studies in younger adults, where cells are analyzed before and
after exposure, may have greater sensitivity and fewer confounding factors.
Suruda et al. (1993) showed a 12-fold increase in the MN frequency of epithelial cells
from the buccal area of the mouth in mortuary science students exposed to embalming fluids
This document is a draft for review purposes only and does not constitute Agency policy.
4.441 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
containing formaldehyde following an 85-day exposure period (see Table 4-85). Overall,
students were exposed to 0.33 ppm (0.4 mg/m3) formaldehyde as an 8-hour TWA on days when
embalming was performed (an average of 6.9 embalmings). Blood, oral, and nasal samples were
collected pre- and postexposure. As shown in Table 4-85, nasal epithelial MNs increased by
22% (frequency of micronucleated lymphocytes increased by 28%). By contrast, SCE frequency
decreased by 7.5% after formaldehyde exposure.
Table 4-85. MN and SCE formation in mortuary science students exposed to
formaldehyde for 85 days
Sampling point
Before course
After course
Buccal mucosa
(MN/1,000)
0.046 ±0.17
0.60±1.27a
Nasal epithelium
(MN/1,000)
0.41 ±0.52
0.50 ±0.67
Blood
(MN/1,000)
4.95 ± 1.72
6.36±2.03a
Blood
(SCEs/cell)
7.72 ± 1.26
7.14 ±0.89
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 (p < 0.05), as calculated by the authors.
Source: Surudaetal. (1993).
Another group (Titenko-Holland et al., 1996) also reported a significant increase in MN
frequency of buccal, but not nasal, epithelial cells from mortuary students exposed to embalming
fluid. In this study, 28 out of 35 students were sampled before and after a 90-day embalming
class. The mean formaldehyde exposure for the subjects providing data on buccal cell MNs was
14.8 ± 7.2 ppm-hours (18.2 ± 8.8 mg/m3-hours) for the entire 90-day period and 16.5 ± 5.8 ppm-
hours (20.3 ±7.1 mg/m3-hours) for students providing data on nasal cell MNs. Cells were
recorded as having either whole chromosomes with centromeres (MN+) or acentric fragments
and no centromeres (MN ). Cells with multiple nuclei were present only in samples taken after
exposure to embalming fluid. There was a ninefold increase in the MN frequency in buccal cells
(p < 0.5) and only a twofold increase (p > 0.05) in nasal cells. In addition, there was a twofold
increase in the MN+ frequency in buccal cells (see Table 4-86). The authors suggested that
chromosomal breakage appears to be the primary mechanism of MN formation.
Ying et al. (1997), however, observed higher frequencies of MNs in the nasal exfoliative
cells (3.85 ± 1.48 versus 1.20 ± 0.676, paired t-test,/> < 0.001) and oral exfoliative cells (0.857 ±
0.558 versus 0.568 ± 0.317, p < 0.001) after formaldehyde exposure, although there was no
significant increase in the frequency of lymphocyte MNs (p > 0.05) in students exposed to
formaldehyde in anatomy classes (three classes per week for 3 hours over an 8-week duration).
In this study, blood samples and nasal swabs were collected before and after the study. The
TWA concentration of formaldehyde in anatomy laboratories and student dormitories was 0.508
This document is a draft for review purposes only and does not constitute Agency policy.
4-442 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
± 0.299 mg/m3 and 0.012 ± 0.0025 mg/m3, respectively, suggesting that nasal mucosa cells
exposed through respiration are the primary target of formaldehyde-induced genotoxicity.
Table 4-86. Incidence of MN formation in mortuary students exposed to
formaldehyde for 90 days
Sampling point
Pre-exposure
Postexposure
/rvalue
Buccal cells (n = 19)
Total MN
0.6 ±0.5
2.0±2.0a
0.007
MN+
0.4 ±0.4
1.1 ±1.3
0.08
MN"
0.1 ±0.2
0.9±l.la
0.005
Nasal epithelial cells (n = 13)
Total MN
2.0 ±1.3
2.5 ±1.3
0.20
MN+
1.2 ±1.3
1.0 ±0.8
0.31
MN"
0.5 ±0.5
1.0±0.6a
0.03
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
""Statistically significant at the level shown, as calculated by the authors.
Source: Titenko-Holland et al. (1996).
Ying et al. (1997), however, observed higher frequencies of MNs in the nasal exfoliative
cells (3.85 ± 1.48 versus 1.20 ± 0.676, paired t-test,/> < 0.001) and oral exfoliative cells (0.857 ±
0.558 versus 0.568 ± 0.317, p < 0.001) after formaldehyde exposure, although there was no
significant increase in the frequency of lymphocyte MNs (p > 0.05) in students exposed to
formaldehyde in anatomy classes (three classes per week for 3 hours over an 8-week duration).
In this study, blood samples and nasal swabs were collected before and after the study. The
TWA concentration of formaldehyde in anatomy laboratories and student dormitories was 0.508
± 0.299 mg/m3 and 0.012 ± 0.0025 mg/m3, respectively, suggesting that nasal mucosa cells
exposed through respiration are the primary target of formaldehyde-induced genotoxicity.
In a different study (Ying et al., 1999), however, the same group showed that exposure to
formaldehyde affected the composition of lymphocyte subsets (B cells, total T cells, T helper-
inducer cells, T cytotoxic-suppressor cells), but no significant difference was reported between
lymphocyte proliferation rate and SCEs at the given levels and durations of formaldehyde
exposure. This study involved 23 nonsmoking students exposed to 0.508 ± 0.299 mg/m3
formaldehyde for a period of 8 weeks (3 hours , 3 times per week).
Burgaz et al. (2002) reported significantly (p < 0.05) higher mean MN frequencies in
buccal mucosal cells from shoe workers as well as anatomy and laboratory workers (0.62 ±
0.45% and 0.71 ± 0.56%, respectively) compared with unexposed controls (0.33 ± 0.30%). In
this study, the measured air concentrations of formaldehyde in the breathing zone of the anatomy
and pathology laboratory workers were between 2 and 4 ppm (2.5 and 5 mg/m3). MN count per
This document is a draft for review purposes only and does not constitute Agency policy.
4.443 DRAFT—DO NOT CITE OR QUOTE
-------
1 3,000 cells was measured in buccal smears from shoe workers and from anatomy and pathology
2 staff, and eighteen male university staff were used as controls.
3 In a critical review, Speit and Schmid (2006) examined data from studies that have
4 reported the formation of MNs in nasal or buccal cells of persons either environmentally or
5 occupationally exposed to formaldehyde. The authors identified a number of issues relating to
6 study design, exposure regimen, and confounding factors, including MN levels in nasal and
7 buccal cells well below established background levels, reports limited by the number of cells
8 observed, variation in standard techniques, and nonconcordance between buccal and nasal
9 findings. However, the authors concluded that, despite these limitations, the weight of evidence
10 supports the finding that formaldehyde may be genotoxic in human cells in direct contact with
11 formaldehyde.
12
13 4.3.4.2.2. Peripheral blood lymphocytes.
14 Mature lymphocytes are present at the POE as intraepithelial lymphocytes and within
15 germinal centers in the mucosa. Because more lymphocytes may be available in the nasal
16 mucosa than the oral mucosa, mouth versus nose breathing may contribute to variability in
17 findings. Since some of the lymphocytes traffic around the body, it is reasonable to find
18 clastogenic effects in these relatively long-lived cells reflected in peripheral blood lymphocytes.
19 Thus, lymphocytes proliferating in response to antigen would be more vulnerable to DNA
20 reactivity of formaldehyde and to the clastogenic effects in general.
21 A cytogenetic evaluation by Fleig et al. (1982) of 15 employees exposed for an average
22 of 28 years in a formaldehyde manufacturing plant revealed no statistically significant increase
23 in the frequency of CAs in peripheral blood lymphocytes compared with a matched control
24 group. Although some of the workers are smokers in this study, smokers did not show an
25 increased incidence of CAs. Likewise, in a different study (Thomson et al., 1984), no exposure-
26 related differences were evident in the frequency of CAs and MNs in lymphocytes from six
27 pathology workers and five unexposed controls. This study did not account for differences in
28 smoking.
29 Bauchinger and Schmid (1985) observed an increased incidence of CAs (dicentric and
30 ring chromosomes) in the peripheral lymphocytes of 20 male paper mill workers and supervisors
31 exposed to formaldehyde (average exposure of 14.5 ± 7.2 years) compared with unexposed
32 workers. When workers and supervisors were analyzed separately, significant increases were
33 only seen for supervisors. The average length of exposure for supervisors (n = 11) and workers
34 (n = 9) was 18.9 years and 7.2 years, respectively. Information regarding formaldehyde
35 concentrations for the two groups was not provided. However, the incidence of SCEs among
This document is a draft for review purposes only and does not constitute Agency policy.
4.444 DRAFT—DO NOT CITE OR QUOTE
-------
1 workers was actually slightly lower than among the 20 controls. In contrast, Ye et al., (2005)
2 observed that the frequency of SCEs in peripheral lymphocytes of 18 nonsmoking formaldehyde
3 factory workers (mean exposure duration of 8.6 years) was significantly increased over
4 nonsmoking controls (8.24 + 0.89 versus 6.38 + 0.41) (described in Section 4.3.4.2.1).
5 Vargova et al. (1992) observed that the percentage of aberrant cells and number of breaks
6 per cell in the peripheral blood lymphocytes of formaldehyde-exposed workers was 3.08 and
7 0.045 versus 3.6 and 0.030 in controls in a pressed board factory, respectively, suggesting both
8 groups to be at an increased risk. However, normal unexposed population had only 1-2%
9 aberrant cells. The authors also noted that the mitotic index was significantly decreased in
10 exposed workers compared with controls. This study did not specify about the smoking
11 background of the two groups, however, the authors indicate that both exposed and unexposed
12 workers have similar habits and social status.
13 Kitaeva et al. (1996) evaluated the genotoxic effects of formaldehyde among 15
14 industrially exposed workers and 8 academic laboratory instructors and observed an increase in
15 the frequencies of CAs and MNs in the lymphocytes of exposed subjects compared with
16 6 unexposed controls. This study did not provide information about smoking history.
17 Shaham et al. (1996, 1997) found significantly higher levels of DPXs and SCEs in
18 peripheral blood lymphocytes of workers occupationally exposed to formaldehyde (physicians
19 and technicians) compared with unexposed control workers. The authors also observed a linear
20 relationship between years of exposure to formaldehyde and levels of DPXs and SCEs.
21 Although both studies have smokers and nonsmokers, smoking did not influence DPX formation
22 in either of the studies (Shaham et al 1996; 1997).
23 Formaldehyde-induced genotoxicity has also been reported in peripheral blood
24 lymphocytes of anatomy class students and mortuary workers. Vasudeva and Anand (1996) did
25 not observe significant differences in the incidences of CAs between the formaldehyde exposed
26 students and the matched, unexposed controls. In this study, peripheral blood lymphocytes from
27 30 medical students exposed to formaldehyde in a gross anatomy laboratory for 15 months with
28 average exposures of less than 1 ppm (1.23 mg/m3) formaldehyde were used. No smoking
29 history is available for this study.
30 He et al. (1998) used the cytokinesis-blocked MN (CBMN) assay to detect the frequency
31 of micronucleated peripheral lymphocytes in 13 students exposed to formaldehyde during a
32 12-week (10 hours/week) anatomy class. Sampling of breathing zone air showed a mean
33 concentration of 2.37 ppm (3.17 mg/m3). Ten students from the same school, without exposure
34 to formaldehyde, were used as controls. All study subjects were nonsmokers. CAs and SCEs
35 were observed in both groups, and there were significant increases (p < 0.01) in the frequencies
This document is a draft for review purposes only and does not constitute Agency policy.
4.445 DRAFT—DO NOT CITE OR QUOTE
-------
1 of micronucleated cells and CAs in the formaldehyde-exposed group compared with the control
2 group.
3 In a study involving 97 plasticware workers (34 males and 63 females) exposed to 0.5 to
4 0.9 mg/m3 formaldehyde, 4.4 to 6.2 mg/m3 styrene and 0.5 to 0.75 mg/m3 phenol for 2 months to
5 25 years, Lazutka et al (1999) observed significantly higher CAs than controls (nonexposed
6 donors matched by age and similar smoking habits as the exposed workers). Although workers
7 with short and long exposures showed significant increases in the frequency of CAs, the
8 cytogenetic damage did not increase with exposure duration.
9 Sari-Minodier et al. (2001), using the CBMN assay in anatomy/pathology laboratory
10 workers who were occupationally exposed to formaldehyde for a mean exposure period of 9 yrs
11 (range 1-16 yrs), reported higher frequency of micronucleated peripheral blood lymphocytes
12 (18.8% ± 13.1) than in matched controls (8.8% ± 4.4).
13 Shaham et al. (2002) observed a mean ± S.E. number of 0.27 ± 0.003 SCEs per
14 chromosome in the peripheral lymphocytes of a hospital pathology department workers with a
15 mean exposure period of 15.4 yrs (range 1-39 yrs) compared with 0.19 ± 0.002 SCEs in control
16 subjects from the administrative staff of the same hospital (p < 0.01) after adjusting for age, sex,
17 smoking habits, etc. This study involved 90 individuals employed in 14 hospital pathology
18 laboratories and 52 unexposed controls.
19 Yu et al. (2005) reported dose-dependent increase in MNs and comet assay parameters
20 (olive tail moment and comet tail length) in peripheral lymphocytes in 151 workers from two
21 plywood factories compared with 112 unexposed controls. Both exposed and control subjects
22 are nonsmokers in this study. The TWA exposure level in the working environment was
23 0.1-7.88 mg/m3 (0.08-6.42 ppm) formaldehyde compared with a background level of <0.01
24 mg/m3 (<0.008 ppm) formaldehyde applicable to controls. In the comet assay, the authors
25 observed olive tail moments averaging 0.93 (0.78-1.1), 3.03 (2.49-3.67), and 3.95 (3.53-4.43)
26 for control, low-, and high-exposure individuals, respectively. For the same subjects, comet tail
27 lengths were 6.78 (6.05-7.6), 11.25 (10.12-12.5), and 12.59 (11.8-13.43), respectively. In the
28 CBMN assay, MNs/100 cells were 0.27 ± 0.13, 0.41 ± 0.25, and 0.65 ± 0.36, respectively, for
29 control, low-, and high-exposure individuals.
30 In a population of 18 workers exposed to formaldehyde at a plant in China, with a mean
31 employment of 8.5 years (range 1 to 15 years), Ye et al (2005) examined nasal epithelial cells
32 and lymphocytes for cytogenetic effects. This study also included a second group of 16 waiters
33 who worked in a newly fitted ball room for 12 weeks with a low level exposure to formaldehyde
34 from building material, tobacco smoke and furniture and a group of 23 college students as a
35 control group. The background indoor air concentration of 0.009 ppm formaldehyde was
This document is a draft for review purposes only and does not constitute Agency policy.
4-446 DRAFT—DO NOT CITE OR QUOTE
-------
1 reported in students' dorms. Significantly increased frequencies of MNs in the nasal mucosal
2 cells and SCEs in peripheral blood lymphocytes were reported for the workers, but not the
3 waiters in this study.
4 Orsiere et al. (2006) reported no apparent effect on the DNA damage in peripheral blood
5 lymphocytes as assessed by a chemiluminescence microplate assay in pathology and anatomy
6 laboratory workers (n = 59) before and after a 1-day exposure to formaldehyde. This study had
7 59 exposed workers and 37 controls with similar smoking history. However, with the CBMN
8 assay, the authors reported statistically significant differences in the frequency of binucleated
9 micrenucleated cells (1.69 ± 0.93 versus 1.11 ± 0.6%) in exposed versus control subjects.
10 Discrimination between clastogenic and aneugenic events by using fluorescence in situ
11 hybridization (FISH) technique with a pan-centromeric DNA probe resulted in a higher rate of
12 binucleated micrenucleated cells (1.91 ± 1.01 versus 1.19 ± 0.56% in controls) and showed that
13 the frequency of centromeric nuclei was higher in the exposed group than in controls, though not
14 significantly. Among the centromeric MNs, the frequency of MNs with only one centromere
15 (Cl+MN) was significantly greater in pathologists/anatomists than in controls (1.1 ± 0.62 versus
16 0.31 ± 0.24%, p < 0.001). The authors interpreted their data on monocentromeric nuclei in
17 anatomists/pathologists as an indication that formaldehyde exposure might be associated with
18 aneugenic (rather than clastogenic) events.
19 Based on pooled analysis of two reports (larmarcovai et al., 2006a, b) (see Table 4-87),
20 MN frequency ratios in the peripheral lymphocytes of cancer patients, welders, and
21 anatomists/pathologists were significantly increased compared with the corresponding controls.
22 The data were taken from three biomonitoring studies by using CBMN/FISH. The incidence of
23 MNs was scored and then evaluated further for the presence of centromere-negative MNs
24 (C-MNs), centromere-positive MNs (C+MNs), and, for the latter case, those containing a single
25 centromere (Cl+MNs) and those containing two or more centromeres (Cx+MNs). Applying
26 their findings to considerations of the aneugenic mechanism of action of formaldehyde, the
27 authors hypothesized that the use of centromeric signals enables the identification of endpoints
28 representing impaired chromosomal migration (with Cl+MN formation) or centrosome
29 amplification (with Cx+MN formation).
30 Recently, Costa et al. (2008) observed a significant increase in the genotoxicity of
31 formaldehyde-exposed pathological anatomy laboratory workers (n = 30) compared with
32 controls (n = 30) in cytogenetic assays. In this study, the authors evaluated the level of exposure
33 to formaldehyde near the breathing zone of workers, and TWA of exposure was calculated for
34 each subject, giving a mean level of exposure to be 0.44 ± 0.08 ppm (range: 0.04-1.58 ppm). As
35 compared with control subjects, peripheral blood lymphocyte cultures of formaldehyde exposed
This document is a draft for review purposes only and does not constitute Agency policy.
4.447 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-87. Multivariate repression models linking genomic
instability/occupational exposures to selected endpoints from the MN assay
Study populations
Cancer patients versus
controls
Welders versus controls
Pathologists/anatomists
versus controls
Number
10/10
27/30
18/18
MNa
1.85
(1.18-2.87)
1.37
(1.09-1.72)
1.28
(0.86-1.90)
C-MN
2.05
(1.07-3.94)
1.39
(0.99-1.95)
0.79
(0.46-1.36)
C+MN
1.81
(1.02-3.21)
1.37
(1.03-1.83)
1.65
(1.05-2.59)
Cl+MN
1.68
(0.80-3.53)
1.10
(0.80-1.53)
3.29
(2.04-5.30)
Cx+MN
1.28
(0.63-2.59)
1.31
(0.99-1.74)
0.68
(0.38-1.20)
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
aBolded values indicate statistical significance (p < 0.05).
Source: larmarcovai et al. (2006b).
workers showed significant increases in MN frequency (5.47 ± 0.76 versus 3.27 ± 0.69;
p = 0.003), SCEs (6.13 ± 0.29 versus 4.49 ± 0.16; p < 0.05), and comet assay as determined by
tail length (TL) (60.00 ± 2.31 versus 41.85 ± 1.97;;? < 0.05). Smoking did not affect MN, TL
and SCE. In addition, Costa et al. (2008) observed a positive correlation between formaldehyde
exposure levels and MN frequency (r = 0.384; p = 0.001) and TL (r = 0.333; p = 0.005) (see
Table 4-88). However, polymorphic genes of xenobiotic metabolizing and DNA repair enzymes
did not show any significant effect on the genotoxic endpoints. This is the lowest level of
exposure to formaldehyde in the studies observed so far, wherein a clear indication of genotoxic
effects of formaldehyde was demonstrated.
Table 4-88. Genotoxicity measures in pathological anatomy laboratory
workers and controls
Controls
(n = 30)
Exposed
(n = 30)
^value
MN assay
Mean MN ± SEM
(range)
3.27 ±0.69
(0-17)
5.47 ±0.76
(1-17)
0.003
SCEs
Mean SCE ± SEM
(range)
4.49 ±0.16
(3.10-3.06)
6.13 ±0.29
(3.64-8.80)
<0.05
Comet assay
Mean TL (|iM) ± SEM
(range)
41.85 ±1.97
(28.85-66.52)
60.00 ±2.31
(33.76-99.09)
O.05
23
24
25
26
Source: Costa etal. (2008).
This document is a draft for review purposes only and does not constitute Agency policy.
4-448 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.3.5. Summary of Genotoxicity
2 Formaldehyde's genotoxicity has been demonstrated in a variety of in vitro and in vivo
3 test systems including bacteria (S. typhimurium and E. coli), molds (N. crassa), yeasts (S.
4 cerevesiae), fruit flies (Drosophila), plants, several mammalian primary cells and cell lines,
5 tumor cell lines, human lymphoblasts and lymphocytes, in vivo studies in rodents (different
6 strains of mice, rats and hamsters), nonhuman primates (monkeys) and peripheral blood
7 lymphocytes of occupationally exposed workers (industrial as well as professional) workers.
8 Among these test systems several genotoxicity endpoints, such as DPXs (or DPCs), DNA
9 adducts, DDXs, SSB, cytogenetic endpoints (MNs, SCEs, CAs), mutations (point, dominant
10 lethal, deletion, heritable), p53 mutations and mutantp53 expression in rat nasal SCCs, other
11 genetic endpoints such as UDS, DNA repair inhibition, etc), and clastogenic effects in human
12 buccal, nasal and peripheral blood lymphocytes and chromosomal changes (monosomy and
13 trisomy) in lymphocytes.
14 Formaldehyde forms predominantly DPXs that are detected in cell-free systems and
15 single cells in vitro. DPXs are also formed in nasal epithelia but not in extranasal passages of
16 rodents, which are completely removed within a day after formation. In vivo data in human and
17 mammalian cells demonstrate that formaldehyde is genotoxic at the site of first contact,
18 including cells of the mouth or the nose. DPXs are also detected in nasal and extranasal tissues
19 of monkeys, suggestive of direct effects of formaldehyde in tissues that correspond to observed
20 tumor sites (nasal and nasopharynx) in humans. In addition, this is used as a basis for cross-
21 species comparison with humans. Formaldehyde-DNA adducts are labile and constitute a minor
22 fraction of the DNA-reaction products and are less likely to play an important role in the
23 genotoxicity of formaldehyde. There is limited literature on the formation of DDXs after
24 formaldehyde exposure, but its role in genotoxicity is not clear.
25 Formaldehyde clastogenicity has been demonstrated by the induction of SCEs, SSBs,
26 MNs, and CAs in cultured mammalian cells. Formaldehyde induces mutations in salmonella and
27 escherichia bacterial strains that contain an AT base pair at the primary reversion site that is used
28 to detect oxidative compounds and cross-linking agents without metabolic activation by
29 exogenous enzyme-activating systems. Formaldehyde induces mutations in cultured mammalian
30 cells at levels that do not cause significant toxicity.
31 Formaldehyde exposure causes differential induction of MNs in human nasal epithelial
32 and buccal epithelial cells, which is significant in industrial exposure workers and students
33 working in anatomy or mortuary science, respectively. However, recent data and data from
34 larger studies support a finding of increased MNs in blood lymphocytes, although the issue
35 remains controversial because of issues relating to study design, exposure regimen, and
This document is a draft for review purposes only and does not constitute Agency policy.
4.449 DRAFT—DO NOT CITE OR QUOTE
-------
1 confounding factors, including MN levels in nasal and buccal cells well below established
2 background levels, reports limited by the number of cells observed, variation in standard
3 techniques, and nonconcordance between buccal and nasal findings (Speit and Schmid, 2006).
4 Several clastogenic effects, such as induction of MNs, SCEs, and CAs, were seen in human
5 peripheral blood lymphocytes; also a recent study showed increased monosomy 7 and trisomy 8
6 in the lymphocytes of formaldehyde exposed workers. Formaldehyde exposure also causedp53
1 mutations in rat nasal carcinomas with the expression of mutant p53 protein.
8 Overall, induction of DPXs as a predominant lesion in vitro and in vivo, clastogenicity,
9 and mutagenicity with locus-specific mutations in nonhuman and human cells supports the
10 concept of genotoxic action of formaldehyde at the POE.
11 A summary of the genotoxicity of formaldehyde in humans is presented in Table 4-89.
12
13 4.4. SYNTHESIS AND MAJOR EVALUATION OF NONCARCINOGENIC EFFECTS
14 The adverse health effects due to formaldehyde exposure have been extensively studied
15 in humans and in animal models. Studies of human exposure include occupational exposures,
16 environmental exposures, and clinical studies of intentionally exposed subjects (see Section 4.1).
17 Occupational exposures are primarily due to inhalation and dermal contact. Animal studies are
18 available for a variety of routes of exposure, including inhalation, oral, dermal, and intravenous
19 and IP. injections (see Section 4.2). Additionally, as discussed in Chapter 3, in vitro studies
20 address biological activity and the metabolic fate of formaldehyde.
21 Taken together, the human and animal studies support numerous health effects, not only
22 at the POE as expected for a reactive gas but also on pulmonary function, neurobehavioral
23 function, reproduction, development, immunomodulation, and sensitization (atopy, asthma). The
24 discussion below provides a description of the adverse effects seen in each area, summarizing the
25 data for both human and animal studies. MOA data are discussed where information regarding
26 formaldehyde's biological activity may be linked to the observed adverse health effects.
27
28 4.4.1. Sensory Irritation
29 Sensory irritation of the eyes, nose, and throat is reported in humans upon direct contact
30 with formaldehyde gas during inhalation exposures (Holmstrom and Wilhelmsson, 1988;
31 Ritchie and Lehman, 1987) and includes irritation resulting from acute exposures (Lang et al.,
32 2008; Yang et al., 2001; Krakowiak et al., 1998; Kulle, 1993; Green et al., 1989, 1987; Kulle et
33 al., 1987; Sauder et al., 1987, 1986; Schachter et al., 1987, 1986; Witek et al., 1987; Day et al.,
34 1984; Bender et al., 1983; Weber-Tschopp et al., 1977). Controlled exposures in inhalation
35 chambers confirm the specificity of these responses to formaldehyde exposure and allow for
36 assessment of these symptoms through both subjective and objective measures (Kulle, 1993;
This document is a draft for review purposes only and does not constitute Agency policy.
4-450 DRAFT—DO NOT CITE OR QUOTE
-------
-------
its
Table 4-89. Summary of human cytogenetic studies
Study population
TV
Exposure time (years)
Range
Mean
Formaldehyde
concentration (ppm)
Range
Mean
(TWA)
Cytogenetic observations
CAs
SCEs
MNs
Reference
Analyses of nasal and/or buccal cells
Plywood workers
Age and sex matched controls
Male mortuary science students
Female mortuary science students
Mortuary science students3
Female anatomy faculty
Male anatomy faculty
Controls (Females)
Anatomy students
Anatomy /pathology staff
Controls (University staff)
Workers at a formaldehyde plant
Controls
Volunteers
15
15
22
7
28
8
5
7
25
28
18
18
23
21
2-19
6.8
Buccal and nasal swabs taken
before and after first 9 weeks of
embalming course
Buccal and nasal swabs taken
before and after first 9 weeks of
embalming course
NA
23.6
25.6
Buccal and nasal swabs taken
before and after 8-week
anatomy course
1-13
1-15
4.70
8.5
10 days
0.32-0.83
0.1-4.3
0.1-4.3
NA
0.06-1.06
2-4
(1)
1.4
1.4
NA
(0.508)
NA
0.8
13.5 ppm-hours
+ nasal
+ buccal
-nasal
- buccal
-nasal
+buccalb
-nasal
+ buccal
- buccal
+ buccal
+ nasal
+ buccal
+ nasal
- buccal
Ballarinetal. (1992)
Surudaetal. (1993)
Titenko -Holland et
al. (1996)
Kitaevaetal. (1996)
Yingetal. (1997)
Burgaz et al. (2002)
Ye et al. (2005)
Speit et al. (2007b)
Analyses of peripheral lymphocytes
Manufacturing workers
Age and sex matched controls
15
15
23-35
28
<5 1971
<1 later
"
Fleigetal. (1982)
& s
I- 1
2J I
5'
3 ,
J^.
rv
L/i
to
fe
H
U
O
-------
ts
Table 4-89. Summary of human cytogenetic studies (continued)
^ 8
1!
f^ S
2 5'
3 ,
Study population
Pathology workers
Controls
Anatomy students0
Papermakers
Controls
Wood workers
Controls
Male embalming students
Female embalming students
Manufacturing workers
Anatomy faculty
Controls
Medial students
Controls
Anatomy students
Controls (students)
Physicians
Technicians
Controls (age matched/unexposed)
Anatomy students
TV
6
5
8
20
20
25
19
22
7
15
8
6
30
30
13
10
6
7
20
23-25
Exposure time (years)
Range
4-11
Mean
10-week class
2-30
<5 to <16
14.4
Blood sampled before and after
first 9 weeks of embalming
course
10
17
Sampled near end of 15 -month
term
12-week class
2-24
2-25
10
15
Blood samples taken before and
after 8 -week anatomy course
Formaldehyde
concentration (ppm)
Range
1.08-1.99d
0.08-0.6e
<3
0.45-8.6
0.15-4.3
Up to 4
NA
<1
3.1-2.8
0.06-1.06
Mean
(TWA)
0.9-5.8
1.2
0.3
NA
NA
1.4
NA
NA
2.37g
1.6
(0.508)
Cytogenetic observations
CAs
+f
+
ND
—
+
-
SCEs
+
-
+
+
+
—
MNs
"
+
+
+
—
Reference
Thomson et al.
(1984)
Yager etal. (1986)
Bauchinger and
Schmid (1985)
Vargova et al.
(1992)
Suruda etal. (1993)
Kitaeva etal. (1996)
Vasudeva and
Anand (1996)
He at al. (1998)
Shaham etal. (1997)
Ying etal. (1999,
1997)
fe
H
o
o
-------
its
& s
I- 1
2J I
5'
3 ,
Table 4-89. Summary of human cytogenetic studies (continued)
Study population
Female anatomy/pathology lab
workers
Controls (Women)
Hospital pathology workers'1
Controls
Workers at a formaldehyde plant
Controls
Workers at two plywood factories
Controls
Pathology or anatomy workers
Controls
Pathologists
Controls
Pathological anatomy lab workers
Controls (21 females and 9 males)
Plasticware workers
Controls (nonexposed donors)
Wood workers
Controls
School children (1984)
School children (1985)
School children (1986)
Controls (1984)
TV
10
27
90
52
18
23
151
112
59
37
18
18
30
30
97
90
40
22
20
16
18
17
Exposure time (years)
Range
1-16
1-39
1-15
ND
ND
ND
0.5-27
2 mo to 25 yrs
NR
Mean
8.9
15.4
8.5
11
Formaldehyde
concentration (ppm)
Range
1.2-15.1
0.04-0.71
0.72-5.6
<0.1-20.4k
0.4-7.0k
0.04-1.58
0.5-0.9 mg/m3
NR
Mean
(TWA)
NA
0.4
2.24
0.8
0.08-6.42
2k
2.3k
0.44
0.26
0.11
0.03
0
Cytogenetic observations
CAs
+
+
+
+
SCEs
+J
+
+
+
MNs
+
-
+
+
+
+
Reference
Sari-Minodier et al.
(2001)
Shaham et al. (2002)
Ye et al. (2005)
Yu et al. (2005)
Orsiere et al. (2006)
larmarcovai et al.
(2006a, b)
Costa et al. (2008)
Lazutkaetal.,1999
Chebotarev et al.
(1986)
Neri et al. (2006)
fe
H
o
o
-------
Table 4-89. Summary of human cytogenetic studies (continued)
Study population
Preschool controls (1984)
Preschool children (1984)
Phenolformaldehyde resin workers
Controls
TV
24
13
31
74
Exposure time (years)
Range
0.33-30 yr
Mean
Formaldehyde
concentration (ppm)
Range
0.17-0.3
Mean
(TWA)
0
0.41
0
Cytogenetic observations
CAs
+
SCEs
MNs
Reference
Suskov and
Sazonova(1982)
& s «
111
f^ XJ '.\1
k ^ S
°S 2 5'
s ,
fe
H
aSame population in Suruda et al. (1993) but different slides used. Nineteen complete slide sets forbuccal analysis and 13 complete slide sets for nasal epithelial
cell analysis.
bNot dose related; both low- and high-exposure groups had same SCE increase.
°Each student sampled before and after 10-week anatomy class.
dBreathing zone samples.
eRoom air samples.
Increase only in 11 supervisors. See text for details.
8Average breathing zone during dissection procedure.
hExposed and controls from 14 hospitals.
'Low- and high-exposure groups established but numbers not provided.
JNot dose related; both low and high groups had same SCE increase.
"Described as "mean concentrations for sampling times of 15 minutes."
CAs = chromosomal aberrations; SCEs = sister chromatid exchanges; MNs = micronuclei; TWA = time-weighted average; ND = not determined; NA = not
applicable.
O
O
-------
1 Holness and Nethercott, 1989; Green et al., 1987; Kulle et al., 1987; Sauder et al., 1986; Weber-
2 Tschopp et al., 1977). Eye irritation may be reported as itching, burning, and general discomfort.
3 Tearing, redness of the eyes, and increased blink frequency are observed and may be quantified
4 in exposure under controlled conditions (Lang et al., 2008; Yang et al., 2001; Andersen and
5 Molhave, 1983; Weber-Tschopp et al., 1977; Schuck et al., 1966). Eye irritation appears to be
6 the most sensitive endpoint in most individuals and may be observed after short exposures
7 (195 minutes at 0.5 ppm: Lang et al. [2008]; 30 seconds at 1.65 ppm: Yang et al. [2001]).
8 Itching, burning, and discomfort of the nose, which may be accompanied by increased
9 mucous production (runny nose), are reported by individuals exposed via inhalation (Krakowiak
10 et al., 1998; Kulle, 1993; Green et al., 1987; Kulle et al., 1987; Weber-Tschopp et al., 1977).
11 Throat irritation may also be described subjectively as itching and burning and is often
12 accompanied by a cough (Krakowiak et al., 1998). Symptoms of eye and mucous membrane
13 irritation are also reported in numerous rodent studies and support the health effects reported in
14 humans (see Section 4.1.1.1). Although dermal contact may result in dermatitis and an apparent
15 hypersensitivity reaction, symptoms do not present upon contact as sensory irritation. There are
16 no human or animal data that assess sensory irritation from oral exposures.
17 The time to onset of sensory irritation symptoms and severity of the sensory irritation are
18 a function of both the air concentration and duration of exposure. Additionally, nose and throat
19 irritation becomes more prominent at higher exposures and longer duration of exposure (Kulle,
20 1993; Kulle et al., 1987). Controlled human laboratory exposures (Yang et al., 2001; Kulle,
21 1993; Kulle et al., 1987; Cain et al., 1986; Andersen and Molhave, 1983) provide more direct
22 exposure-response evidence for sensory irritation. These studies are limited to healthy
23 nonsmoking individuals. Two studies (Cain et al., 1986; Andersen and Molhave, 1983)
24 document discomfort and irritation of the eye in response to acute exposures as low as 0.25 ppm.
25 Dose-response relationships are reported in a number of different ways: as an incidence of the
26 reported symptom among subjects, as a score for severity of the symptom, or in some cases as a
27 subjective measure, such as blink frequency for eye irritation.
28 Symptoms of sensory irritation, including eye irritation (burning watering, increased
29 blinking), nasal irritation (rhinitis, itching/burning), throat/respiratory tract irritation (wheezing,
30 coughing, phlegm production), have been reported in numerous worker cohorts. Occupational
31 exposure environments include hospital and medical settings, students, and industrial workers
32 (Takahashi et al., 2007; Takigawa et al., 2005; Krakowiak et al., 1998; Akbar-Khanzadeh et al.,
33 1994; Uba et al., 1989; Horvath et al., 1988; Schachter et al., 1987). Formaldehyde levels often
34 vary in a work environment and peak as well as average exposures may be used to report
35 occupational exposures. Although sensitive individuals often remove themselves from an
This document is a draft for review purposes only and does not constitute Agency policy.
4-456 DRAFT—DO NOT CITE OR QUOTE
-------
1 irritating workplace (the healthy worker effect), eye, nose, and throat symptoms are still reported
2 in this environment. Among workers in a plant where formaldehyde resins were used, those
3 exposed to an average of 210 ppb formaldehyde (range of 40-400 ppb) reported increased
4 symptoms, including nasal and eye irritation, above those in the control population (Holmstrom
5 and Wilhelmsson, 1988).
6 These effects have been noted in students, particularly medical students, who are exposed
7 to formaldehyde in cadaver labs. In a study of 24 formaldehyde-exposed anatomy students
8 (personal breathing zone samples 0.73 ppm, range 0.49-0.93) (Kriebel et al., 1993), eye, nose,
9 and throat irritation was present when comparing rates of irritation from the end or middle of
10 class to before the start of class. Takahashi et al. (2007) showed that 143 medical students
11 reported various symptoms (including eye and throat irritation) and that the percentage of
12 students reporting symptoms increased between the beginning (measured after the first day of
13 class) and the end of the course (2 months later). After the first day of class, approximately 35%
14 of students reported eye soreness and about 15% reported throat irritation.
15 Sensory irritation has also been reported in occupational settings, including particleboard
16 manufacturing, woodworking, and embalming, with average formaldehyde concentrations of
17 300-400 ppb. Symptoms of irritation were observed at a frequency of 40-65% among exposed
18 workers compared to 6-25% among unexposed comparison groups (Horvath et al., 1988;
19 Alexandersson and Hedenstierna, 1988; Holmstrom and Wilhelmsson, 1988; Holness and
20 Nethercott, 1989).
21 Eye, nose and throat irritation also were observed in association with residential exposure
22 to formaldehyde. Studies with formaldehyde measurements and analyses that adjusted for
23 potential confounders (e.g., age, gender and smoking status) observed a clear exposure-response
24 relationship for average formaldehyde concentrations or cumulative exposure (Ritchie and
25 Lehnen, 1987; Hanrahan et al., 1984; Liu et al., 1991).
26 Ritchie and Lehnen (1987) surveyed residents in 2,000 homes classified as having
27 formaldehyde concentration <0.1 ppm, 0.1-0.3 ppm, and >0.3 ppm. Increases in the prevalence
28 of eye irritation of 1-2%, 12-32%, and 86-93%, respectively were observed with increasing
29 categories of formaldehyde exposure. Liu et al. (1991) report irritant effects associated with
30 formaldehyde exposure in mobile homes, where formaldehyde concentrations ranged from the
31 0.01 ppm detection limit to 0.46 ppm. Eye irritation (60%), nose/throat irritation (10-20%), or
32 headache (<10%) were reported in residents.
33
This document is a draft for review purposes only and does not constitute Agency policy.
4-457 DRAFT—DO NOT CITE OR QUOTE
-------
1 MOA
2 The mucosae of the URT, oral cavity pharynx, and upper airways are complex tissues,
3 where epithelial and goblet cells predominate. In addition, the nasal mucosa is highly enervated.
4 The main nerves include the trigeminal nerve and olfactory sensory cells (olfactory epithelium,
5 the vomeronasal organ, and the organ of Masera) (Feron et al., 2001). A possible MOA for
6 sensory irritation includes formaldehyde-induced stimulation of the trigeminal nerve (though
7 whether formaldehyde acts as a direct agonist is unknown). Trigeminal nerve stimulation in the
8 nasal passages transmits signals to the CNS, which then sends efferent signals back to the nasal
9 tissues, causing sensory irritation, and possibly systemically via vagal nerve stimulation,
10 resulting in more systemic effects.
11 Animal studies are potentially useful models for understanding mechanisms of toxicity,
12 especially where sufficient human data do not exist. While experimental animal studies provide
13 a model of secondary effects, rodents also demonstrate KB, an effect not seen in humans. Thus,
14 species that exhibit bradypnea (like mice and rats) may not be appropriate for assessing
15 respiratory endpoints. The mechanism underlying KB includes formaldehyde binding to the
16 sensory nerve endings of the trigeminal nerve, where signals travel to the CNS. The vagus nerve
17 transmits the efferent signal to produce smooth muscle contraction. The animals become
18 inactive, their core temperatures decrease by several degrees C, and their respiratory rates and
19 minute volumes decrease. However, this is not to say that trigeminal nerve stimulation is not an
20 appropriate potential mechanism of action in other species or in humans. Since trigeminal nerve
21 stimulation has been independently confirmed in species without KB, this mechanism may be a
22 viable explanation for the observed effects.
23
24 4.4.2. Pulmonary Function
25 Several adequately conducted and reported studies evaluated chronic effects of
26 occupational (Malaka and Kodama, 1990; Herbert et al., 1994; Horvath et al., 1988; Holness and
27 Nethercott, 1989; Alexandersson et al., 1982; Alexandersson and Hedenstierna, 1989) or
28 residential exposure to formaldehyde (Krzyzanowski et al., 1990; Franklin et al., 2000). Small
29 decreases in preshift lung function relative to unexposed groups were reported in several
30 different occupational settings including plywood, oriented strand board and carpentry (Malaka
31 and Kodama, 1990; Herbert et al., 1994; Alexandersson et al., 1982; Alexandersson and
32 Hedenstierna, 1989). The studies included appropriate comparison groups, formaldehyde
33 measurements, detailed reporting of methods, and adjustment for potential confounding
34 variables, including age and smoking status (in some studies weight and ethnicity). Lung
35 function values were presented as percent of expected based on age, height, and sex or adjusted
This document is a draft for review purposes only and does not constitute Agency policy.
4-458 DRAFT—DO NOT CITE OR QUOTE
-------
1 for these variables in the analysis. Studies of long-term exposure also reported increased
2 respiratory symptoms such as cough, increased phlegm, asthma, chest tightness and chest colds
3 in exposed workers (Malaka and Kodama, 1990; Herbert et al., 1994; Pourmahabadian et al.,
4 2006, Alexandersson et al., 1982; Alexandersson and Hedentierna 1989).
5 Other studies evaluated pulmonary function among occupational groups with exposure to
6 formaldehyde but the results of these studies are less informative because spirometric values
7 were not adjusted for age, height or gender, statistical analyses were not reported, or the studies
8 suffered from small sample size, methodological, analytical or reporting deficiencies
9 (Khamgaonkar and Fulare, 1991; Pourmahadadian et al. (2006); Kilburn et al., 1985; Main and
10 Hogan, 1983; Ostojic et al., 2006; Holmstrom and Wilhelmsson, 1988; Nunn et al., 1990). Also,
11 occupationally exposed groups compared to nonoccupational referent groups may have exhibited
12 a healthy worker effect masking any formaldehyde related lung function decrements (Holness
13 and Nethercott, 1989).
14 A longitudinal study by Alexandersson and Hedenstierna (1989) documented an
15 association of lung function decrements between 1980 and 1984 with time-weighted average
16 occupational formaldehyde concentrations of 0.42-0.5 mg/m3 (340-400 ppb). Statistically
17 significant annual decreases in FEVi/FVC and FEF25-75o/0 were noted over 5 years in nonsmokers
18 after correction for normal aging and reference lung function spirometry values. The decrease in
19 FEF25-75o/0 was 0.212 ± 0.066 L/second (mean ± SD) for each year of exposure and was
20 significant (p < 0.01). The decrease in FEVI/FVC was 0.4 ± 0.2 (p < 0.01). Across-sectional
21 study with a participation rate of 93% among carpentry workers observed statistically significant
22 decrements in FEVi, FEVi/FVC, and FEF25-75o/0 associated with an 8-hour time-weighted
23 average formaldehyde exposure of 1.13 ppm (Malaka and Kodama, 1990). In multiple
24 regression models adjusting for age, height, weight, cigarettes/day, and dust, formaldehyde as a
25 continuous variable was a significant predictor for FEVi, FEVi/FVC, and FEF25-75o/0. Each unit
26 increase in formaldehyde (ppm-years) was associated with a decrease of 0.015 liters, 0.347%,
27 and 0.043 1/s in FEVi, FEVi/FVC, and FEF25-75%, respectively. The strongest response was for
28 FEF25-75o/0; which showed a 12% drop in observed function compared with expected function in
29 the unexposed.
30 Kryzanowski et al. (1990) described a well-designed and executed cross-sectional study
31 of residential formaldehyde exposure in a large, representative sample that provided clear
32 evidence of a linear relationship between increased formaldehyde exposure and decreased peak
33 expiratory flow rate (PEFR) among children. A statistically significant linear relationship
34 between increased household mean formaldehyde exposure and decreased PEFR was reported in
35 children (B = -1.28 ± 0.46 L/minute per ppb formaldehyde). The average formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-459 DRAFT—DO NOT CITE OR QUOTE
-------
1 concentration was 26 ppb, with a maximum sample value of 140 ppb. Decrements in PEFR
2 associated with increasing formaldehyde concentrations also were observed among adults
3 beginning at an average concentration of 40 ppb. While Franklin et al. (2000) did not observe an
4 association between a one-time measurement of FVC or FEVi among children aged 6-13 years
5 and indoor concentrations of formaldehyde, levels of exhaled nitric oxide (NO) were higher in
6 children exposed to average formaldehyde levels >0.05 ppm compared to < 0.05 ppm. These
7 findings indicate that formaldehyde may increase lower airway inflammation at concentrations
8 associated with effects on pulmonary function.
9 The pulmonary function measures associated with formaldehyde exposure are consistent
10 with bronchial constriction, inflammation, or chronic obstructive lung disease. Decreases in
11 spirometric values, including vital capacity (VC), forced expiratory volume (FEV), forced vital
12 capacity (FVC) and FEV/FVC have been documented. Decreases in lung volume (FEVi, FVC)
13 indicate possible pulmonary obstruction (narrowing of the airways during exhalation) (Pellegrino
14 et al., 2005). Early changes in small airways are observed as reductions in expiratory flow in the
15 terminal portion of the spirogram (PEF, FEF75o/0, MEF25%.75%). These changes may be observed
16 even if FEVi is not affected.
17 Lung function deficits have been reported in pre- versus postshift measurements among
18 occupational groups (Herbert et al., 1994; Horvath et al., 1988). Students have also shown
19 decrements in lung function that are associated with exposure to formaldehyde in laboratories
20 Kriebel et al., 1993; Kriebel et al., 2001. Kriebel and colleagues (1993) observed a 2% cross-lab
21 decrement in PEF among anatomy students during the first two weeks of a 3-hour laboratory that
22 they attended once per week . The cross-lab pulmonary response was attenuated over the 10-
23 week duration of the course (formaldehyde geometric average concentration of 0.73 ppm).
24 While the acute effects of formaldehyde exposure appeared to diminish after several weeks of
25 exposure, the effect on prelaboratory PEF across 10 weeks was a 2.7 ± 1.1 L/minute per week
26 drop that was statistically significant (p < 0.01) in a model adjusting for random person effects,
27 asthma, interaction between time and asthma, and eye as well as nose symptoms of irritation.
28 Prevalence of eye and nose symptoms was associated with decreased PEF (p < 0.02).
29 Overall, acute formaldehyde exposures (0.5-3 ppm) have not induced significant
30 pulmonary deficits in healthy, nonexercising volunteers in controlled human exposure studies
31 (Kulle et al., 1987; Schachter et al., 1986; Schachter et al., 1987; Witek et al., 1986; Day et al.,
32 1984; Andersen and Molhave, 1983). However, it is unclear whether the data analysis in these
33 reports had the statistical power to substantiate the small deficits reported in occupational and
34 student studies. The studies exposed small numbers of diverse individuals, often including males
35 and females of varying age, and some included current smokers. Some studies report the
This document is a draft for review purposes only and does not constitute Agency policy.
4-460 DRAFT—DO NOT CITE OR QUOTE
-------
1 absolute values of the lung function parameters without adjustment to individual expected
2 function or the unexposed baseline for each individual (Kulle et al., 1987; Andersen and
3 Molhave, 1983; Day et al., 1984). In other studies that reported lung function values as a percent
4 of baseline, the variation of the mean change in lung function parameters is quite large, nearly
5 equaling the reported value and exceeding it in several cases (Witek et al., 1986; Schachter et al.,
6 1986; 1987). The absence of normalized raw data, combined with large individual variation,
7 limit the interpretation of these studies.
8 Small but statistically significant deficits in pulmonary function (e.g., decreased FEVi,
9 FVCi, FEVs, specific airways conductance) due to acute formaldehyde exposure (2 or 3 ppm)
10 have been reported in healthy volunteers in controlled human exposure studies using exercise
11 (Green et al., 1987, 1989; Sauder et al., 1986;). Although changes in lung function parameters
12 averaged over experimental groups were generally small, some individuals exhibited clinically
13 significant deficits, even after only 2 hours of exposure (Green et al., 1987). This differential
14 response suggests susceptibility in certain subjects (Green et al., 1987). Other studies that
15 included an exercise component did not report a difference in response among healthy volunteers
16 (Schachter et al., 1986; Kulle et al., 1987). Acute controlled studies that evaluated responses
17 among asthmatics reported no changes in pulmonary function associated with formaldehyde
18 exposure (Sheppard et al., 1984, Ezratty et al., 2007; Harving et al., 1990; Green et al., 1987,
19 1989; Sauder et al., 1987; Witek et al., 1987, 1986; Krakowiak et al., 1998). These findings
20 suggest that a brief exposure to formaldehyde may not trigger a response in the airways of
21 asthmatic individuals in the absence of allergen. However, the large variation in pulmonary
22 response among the individuals (healthy and asthmatic) that participated in the experimental
23 exposure studies suggests that some individuals may be more sensitive to formaldehyde.
24 Several animal studies document increased airway resistance and bronchial constriction
25 following inhalation exposure to formaldehyde (Nielson et al., 1999; Swiecichowski et al., 1993;
26 Biagini et al., 1989; Amdur et al., 1960). A study using cynomolgus monkeys (Biagini et al.,
27 1989) demonstrated that methacholine-induced bronchial constriction can be similarly induced
28 by acute formaldehyde exposure (10 minutes at 2.5 ppm). Thus, formaldehyde exposure
29 simulated bronchial constriction observed after methacholine challenge, but these effects may
30 not occur by a similar MO A. Similar results were reported in guinea pigs (Swiecichowski et al.,
31 1993; Amdur et al., 1960), rats (Ohtsuka et al., 1997), and mice (Nielson et al., 1999).
32 Taken as a whole, studies of occupational exposure to formaldehyde, as well as
33 residential exposure to low indoor formaldehyde concentrations support an association with
34 deficits in pulmonary function among adults and children. Respiratory symptoms also were
35 reported at the same exposure levels. A longitudinal study that documented a progressive
This document is a draft for review purposes only and does not constitute Agency policy.
4-461 DRAFT—DO NOT CITE OR QUOTE
-------
1 decrease in pulmonary function among nonsmokers over 5 years (Alexandersson and
2 Hedenstierna, 1989) supports the conclusion that formaldehyde exerts a progressive chronic
3 effect on pulmonary function. Formaldehyde exposures used in controlled human chamber
4 studies have not caused functional deficits among healthy, nonexercising volunteers or among
5 asthmatics, although wide variability in responses suggests that some individuals may be
6 sensitive responders. A few studies that incorporated exercise reported small decreases in
7 pulmonary function related to formaldehyde exposure. These effects have been corroborated in
8 animal studies exposed to formaldehyde.
9
10 MO A
11 Formaldehyde-induced inflammation of the airways may contribute to observed
12 decreases in measures of pulmonary function. Even short-term inflammatory reactions could
13 reduce the effective diameter of the conductive airways, resulting in lower respiratory volumes in
14 a number of functional tests. Formaldehyde-induced trigeminal nerve stimulation contributes to
15 airway inflammation, which in turn would reduce airway function. Chronic exposures may
16 result in increased sensitization or chronic inflammatory responses, which could contribute to the
17 effects seen in the worker and residential populations.
18 Formaldehyde-induced pulmonary function deficits may also be in part a result of smooth
19 muscle contraction in repose to trigeminal nerve stimulation. Trigeminal nerve stimulation
20 transmits signals to the CNS. The resulting efferent signal from the vagal nerve produces
21 smooth muscle contraction and may result in decreased pulmonary function. Efferent signaling
22 has also resulted in release of substance P and other neuromodulatory compounds, which may
23 contribute to BC and sensitization of pulmonary responses (asthma, atopy).
24
25 4.4.3. Hyper sensitivity and Atopic Reactions
26 A large number of studies have investigated the potential association between
27 formaldehyde exposure and a continuum of adverse health effects ranging from decrements in
28 pulmonary function to asthma. In general, epidemiologic studies of adults have reported varied
29 results between null findings and positive findings but have not consistently distinguished
30 between studies in which formaldehyde may be causing an increase in the incidence of asthma
31 (e.g., phenotypic switching), increasing the prevalence of asthma, initiating an asthma attack or
32 worsening the severity of an attack. Formaldehyde may itself be an allergen or it may potentiate
33 the ability of other allergens to cause phenotypic switching or increase the sensitivity of atopic
34 individuals. Thus formaldehyde exposure among nonatopic individuals could theoretically cause
35 phenotypic switching in the presence or absence of allergens possibly resulting in a diagnosis of
This document is a draft for review purposes only and does not constitute Agency policy.
4-462 DRAFT—DO NOT CITE OR QUOTE
-------
1 asthma. Formaldehyde could also cause an asthma attack or potentiate the influence of other
2 stimuli on the risk of asthma attacks.
3 Asthma is a specific manifestation of IgE-mediated hypersensitivity, characterized by
4 BHR and airway inflammation, resulting in lower airway obstruction (Fireman, 2003; Kuby,
5 1991). A variety of hypersensitivity reactions have been reported following exposure to
6 formaldehyde. Rashes and skin reactions have been reported in some individuals after dermal
7 exposures to formaldehyde. Increased expression of Th-2 cytokines in the lymph nodes of mice
8 given dermal applications of formaldehyde does indicate the involvement of an immune
9 component to the observed sensitization (Dearman et al., 2005; Hilton et al., 1998; Arts et al.,
10 1997). However, the response does not appear to be IgE mediated (Arts et al., 1997; Lee et al.,
11 1984). Gorski et al. (1992) observed an increase in formaldehyde-mediated neutrophil burst in
12 dermatitis patients exposed in a controlled chamber study and suggests a putative role of
13 oxidative stress and reactive oxygen species (ROS). Some case reports of bronchial asthma in
14 occupational settings suggest direct respiratory tract sensitization to formaldehyde gas (Lemiere
15 et al., 1995; Burge et al., 1985; Hendrick et al., 1982; Hendrick and Lane, 1977, 1975; Stenton
16 and Hendrick, 1994, Nordman et al., 1985). However, the development of an allergic asthmatic
17 response to formaldehyde appears to be a rare occurrence.
18 The few studies that reported an association of formaldehyde exposure with a new
19 diagnosis of asthma are consistent with the notion that formaldehyde may be an allergen or that
20 formaldehyde exposure may result in sensitization to common allergens among nonatopic
21 individuals, referred to as phenotypic switching (Rumchev et al., 2001; Smedje and Norback,
22 2001). Rumchev et al. (2002) reported that residential formaldehyde exposure was associated
23 with an increased risk of incident asthma in a population-based case-control study of 192
24 children aged 6 months to 3 years. The asthma diagnosis was based on a recent discharge from
25 the emergency department of a children's hospital in Perth, Australia, with a primary diagnosis
26 of asthma. The assessment of formaldehyde exposure was based on in-home measurements (8-
27 hour passive sampler) in two seasons and associations with asthma were adjusted for several
28 potential confounding factors including, other indoor air pollutants, house dust mite
29 concentrations, humidity and temperature, family history of asthma and allergy, smoking, and
30 other risk factors and demographic variables. Multiple regression models indicated a 3%
31 increased risk with each 10 |J,g/m3 increase in formaldehyde concentration in the home (OR
32 reported per unit formaldehyde increase 1.003 [95% CI: 1.002-1.003]). When formaldehyde was
33 categorized into four exposure groups, odds ratios were increased at 50-50 ng/m3 and above 60
34 Hg/m3 compared to 10-29 and 30-49 ng/m3. A 39% increased risk was observed at the highest
35 exposure category (p < 0.05). Asthma incidence also was evaluated in a four year follow-up of
This document is a draft for review purposes only and does not constitute Agency policy.
4-463 DRAFT—DO NOT CITE OR QUOTE
-------
1 1732 students, aged 7-13 years, from 39 schools in Sweden who completed a mailed
2 questionnaire in 1993. Among the 1347 students who responded in 1997 (78%), 56 (4.5) were
3 diagnosed with asthma by a physician. While asthma incidence was not associated with
4 formaldehyde exposure in the entire cohort (OR 1.2 [95% CI: 0.8-1.7]), among 22 students with
5 no history of atopy at baseline, a higher concentration of formaldehyde was associated with an
6 odds ratio for incident asthma of 1.7 (95% CI: 1.1-2.6). Mold concentration also was associated
7 with incident asthma. The authors did not discuss the correlation of formaldehyde and mold
8 concentrations and do not appear to have adjusted for mold concentrations in the formaldehyde
9 analyses. However, when the units for both mold and formaldehyde are standardized to the
10 geometric mean standard deviation, the magnitude of the formaldehyde effect is larger.
11 Therefore, it is unlikely that the association with formaldehyde was the result of uncontrolled
12 confounding by mold concentrations.
13 Other studies of asthma prevalence have observed an association with indoor
14 formaldehyde concentrations among children exposed at home (Krzyzanowski et al., 1990;
15 Garrett et al., 1999) and at school (Zhao et al., 2008). A large, representative study of 202
16 households (mean formaldehyde level of 26 ppb) found that among children aged 6-15 years old
17 and exposed to environmental tobacco smoke, the prevalence of physician-diagnosed asthma was
18 45.5% for those with measured levels of formaldehyde in the kitchen >60 ppb (N= 11). The
19 prevalence of asthma dropped to 0% for levels 41-60 ppb (N= 12) and 15.1% for levels <40 ppb
20 (N= 106) (chi-squared trend test/? < 0.05). No trend in asthma prevalence was seen for children
21 who were not exposed to environmental tobacco smoke (Krzyzanowski et al., 1990).
22 Krzyzanowski et al. also reported a statistically significant linear relationship between increased
23 household mean formaldehyde exposure and decreased peak expiratory flow rates among
24 children. PEFR is a diagnostic tool used to identify new asthma cases. Two studies of self-
25 reported asthma reported null results (Gee et al., 2005; Tavernier et al., 2006; Palczynski et al.,
26 1999). However, Tavernier et al. did not report the range of formaldehyde concentrations
27 evaluated so the variation of the concentration of formaldehyde is not known. Among the 187
28 children aged 15 years or less studied by Palczynski et al., only 9 were defined as having asthma.
29 Ambient formaldehyde concentrations outdoors on school grounds also were associated
30 with cumulative asthma in a cross-sectional study of students at 10 schools in China (Zhao et al.,
31 2008). Ambient formaldehyde concentrations of 5.8 ng/m3 were associated with an odds ratio
32 for cumulative asthma of 4.6 (95% CI: 1.1-19.5). The odds ratio was adjusted for age, gender,
33 parental asthma or allergy and home factors including environmental tobacco smoke. Another
34 study that evaluated ambient formaldehyde exposure reported an increase in the severity of
35 asthma symptoms in a panel of 22 Hispanic children, 10-16 years old, with a minimum one year
This document is a draft for review purposes only and does not constitute Agency policy.
4-464 DRAFT—DO NOT CITE OR QUOTE
-------
1 history of doctor diagnosed asthma (Delfino et al., 2003). A 24-hour average formaldehyde
2 concentration at a central site monitor of 7.21 ppb (range 4.27-14.02 ppb) was associated with an
3 increase in asthma symptom scores on the next day (OR 1.37 (95% CI: 1.04-1.89).
4 A recent meta-analysis of formaldehyde exposure supports an association with asthma in
5 children (McGwin et al., 2010). Of the seven studies that were included in the meta-analysis, six
6 reported increased risks of asthma associated with exposure to formaldehyde. The results of the
7 random-effect model results showed an overall effect estimate of OR =1.17 (95% CI: 1.01-
8 1.036). The three studies with the highest statistical weights based on the inverse of the variance
9 of the study ORs were for the studies by Rumchev et al. (2002), Garrett et al. (1999) and
10 Krzyzanowski et al. (1990).
11 A cross-sectional study of residential exposure to formaldehyde among 88 adults from
12 Sweden reported an increase in reports of nocturnal breathlessness in association with mean
13 bedroom formaldehyde concentrations of 29 |J,g/m3 compared to 17 |J,g/m3 (Norback et al., 1995).
14 The adjusted odds ratio was 12.5 (95% CL2.0-77.9) per a 10-fold increase in the indoor
15 concentration, higher in magnitude than the odds ratios for toluene, terpenes and volatile organic
16 compounds. This finding is supported by the observation by Krzyzanowski et al. (1990) of
17 decreased morning peak expiratory flow rate among adult smokers exposed to >40 ppb
18 formaldehyde in their homes. Another study of pregnant women found no association with
19 asthma prevalence for personal exposure to > 47 ppb compared to < 18 ppb (Matsunaga et al.,
20 2008). However, the number of women with a diagnosis of asthma was small (N= 21) and the
21 confidence limits were wide (OR 2.65 [95% CI: 0.63-11.11]).
22 The epidemiologic literature also indicates that formaldehyde may increase the
23 prevalence of allergic sensitization to common allergens, as well as the severity of responses
24 (Garrett et al., 1999; Delfino et al., 2003; Cassett et al., 2006).
25 Garrett et al. (1999) reported that the prevalence and severity of allergic sensitization to
26 12 common allergens was increased in relation to formaldehyde levels in the homes of 148
27 children, aged 7-14 years, living in Victoria, Australia based on the highest of four seasonal
28 4-day formaldehyde measurements in the home (p < 0.001). The proportion of atopic children
29 between groups of the highest recorded formaldehyde level was 0.33, 0.64, and 0.75 for <20
30 ug/m3, 20-50 ug/m3, and >50 ug/m3, respectively (test for trend,/? < 0.001). The differences in
31 prevalence were statistically significant (p = 0.001). In logistic regressions, the crude association
32 for atopy with an increase in bedroom formaldehyde concentration per 10 ug/m3 was OR = 1.34
33 which increased when adjusted for parental asthma and gender to an odds ratio of 1.40 per 10
34 ug/m3 (95% CI: 0.98-2.00). Thus, parental asthma was not a confounder of the association
35 between formaldehyde and prevalence of atopy. The adjusted odds ratio for atopy with an
This document is a draft for review purposes only and does not constitute Agency policy.
4-465 DRAFT—DO NOT CITE OR QUOTE
-------
1 increase in highest recorded formaldehyde per 20 ug/m3 was 1.42 (95% CI: 0.99-2.04). Passive
2 smoking, the presence of pets, indoor nitrogen dioxide concentrations, airborne fungal spores
3 and house-dust-mite allergens also did not influence the effect estimates and were unlikely to be
4 confounders. The authors reported that mean respiratory symptom scores adjusted for parental
5 asthma and parental allergy increased with increasing formaldehyde exposure categories (<20
6 ug/m3, 20-50 ug/m3, >50 ug/m3,/? < 0.05). Delfino et al. (2003) also reported an association of
7 the severity of respiratory symptoms with ambient formaldehyde concentration on the previous
8 day among asthmatic children. The association between formaldehyde concentrations and
9 severity of allergic sensitization was analyzed using two measures, number of positive skin prick
10 tests and the ratio of wheal diameters after skin pricks with allergens compared to the histamine
11 wheal size. Average levels of both measures of severity were higher in the two higher
12 formaldehyde groups compared to the lowest group (p < 0.05). Further, both measures were
13 linearly related to increasing formaldehyde categories in regression models controlling for
14 parental asthma and allergy and sex. The role of formaldehyde in exacerbating allergic
15 responses to common allergens among atopic individuals was demonstrated in a controlled
16 human exposure study by Cassett et al. (2006) using formaldehyde concentrations of 100 ug/m3
17 and a 30 minute cross-over exposure protocol. Another study using different allergens,
18 dosimeters, and study protocol did not report an effect of formaldehyde on allergic responses
19 (Ezratty et al., 2007).
20 Exacerbation of response after formaldehyde exposure has been demonstrated in animal
21 studies as well. Sadakane et al. (2002) demonstrated that formaldehyde exposure exacerbated
22 sensitization and challenge with Der f and suggested that formaldehyde exposure may aggravate
23 eosinophilic infiltration and goblet cell proliferation that accompanies allergic responses.
24 Several animal studies report increased airway resistance and BC due to inhalation exposures to
25 formaldehyde (Nielsen et al., 1999; Swiecichowski et al., 1993; Biagini et al., 1989; Amdur,
26 1960). Changes in pulmonary resistance were observed as early as 10 minutes after exposure
27 (Biagini et al., 1989), and reported effect levels ranged from 0.3 to 13 ppm. BHR is commonly
28 associated with allergic Type I hypersensitivity reactions but is not sufficient to demonstrate that
29 an agent induces Type 1 hypersensitivity.
30 In conclusion, the epidemiologic studies of formaldehyde exposure among children
31 support the finding that low indoor and outdoor concentrations result in increased asthma
32 incidence and prevalence. Formaldehyde exposure is associated with increased prevalence of
33 allergic sensitization to common allergens and with increased severity of asthmatic symptoms
34 and the allergic response to allergen challenge. The studies of formaldehyde exposure among
This document is a draft for review purposes only and does not constitute Agency policy.
4-466 DRAFT—DO NOT CITE OR QUOTE
-------
1 adults is sparse but suggest that exposure increases the prevalence of respiratory symptoms.
2 These effects are supported by findings of experimental animal studies.
O
4 MOA
5 The MOA underlying this response has not been elucidated. Formaldehyde-induced IgE
6 production has been reported in some studies (Vandenplas et al., 2004; Wantke et al., 1996a).
7 Other studies suggest that this effect does not appear to be immunogenic in nature (Fujimaki et
8 al., 2004; Lee et al., 1984). Although formaldehyde exposure has been reported to alter cytokine
9 levels and immunoglobulins in some experimental systems (Fujimaki et al., 2004a; Ohtsuka et
10 al., 2003), these immunomodulatory effects do not support immunogenically mediated type 1
11 hypersensitivity.
12 These decrements may be mediated via neurogenic potentiation (Sadakane et al., 2002;
13 Riedel et al., 1996; Tarkowski and Gorski, 1995). Tarkowski and Gorski (1995) suggest that
14 formaldehyde may increase permeability of respiratory epithelium and destruction of
15 immunologic barriers. Tachykinin NK1 receptor and various neuropeptides (NGF and substance
16 P) have been implicated in formaldehyde-induced sensitization and lend weight of evidence to a
17 neurogenic MOA (Van Schoor et al., 2000; Ito et al. 1996).
18
19 4.4.4. Upper Respiratory Tract Histopathology
20 Several studies in occupational workers have reported increased squamous cell
21 metaplasia and reduced mucociliary clearance in nasal and buccal swabs from humans
22 occupationally exposed to formaldehyde (Holmstrom et al., 1989; Holmstrom and Wilhelmsson,
23 1988). Evidence of genotoxic effects include increased MNs and CAs in nasal and buccal
24 epithelial cells from both workers and students exposed to formaldehyde (Ying et al., 1997;
25 Titenko-Holland et al., 1996; Suruda et al., 1993) and suggest a potential association between
26 genotoxicity and altered histopathology.
27 Numerous animal experimental studies in multiple strains of rats, mice, hamsters, rabbits,
28 and monkeys describe formaldehyde-induced URT pathology (F16-Neyret et al., 2001; Roemer
29 et al., 1993; Reuzel et al., 1990; Monticello et al., 1989; Zwart et al., 1988; Wilmer et al., 1987;
30 Morgan et al., 1986b, 1983; Swenberg et al., 1986; Buckley et al., 1984). Effects are first
31 observed in the anterior respiratory mucosa and progress through the nasal passages with
32 increasing exposure concentration and time. The first observed effect includes damage to the
33 mucociliary apparatus of the nasal passages in response to formaldehyde. Studies conducted
34 both in vivo and in vitro demonstrate that formaldehyde disrupts mucus flow and ciliary beat that
35 are dependent on concentration and duration of exposure. Mucociliary apparatus deficits have
This document is a draft for review purposes only and does not constitute Agency policy.
4-467 DRAFT—DO NOT CITE OR QUOTE
-------
1 been recorded even after 18 hours of recovery following formaldehyde exposure. The
2 breakdown of the mucociliary apparatus may allow for increased infection and allow the
3 underlying epithelium to come into contact with exogenous chemicals.
4 Formaldehyde is highly reactive and may impact all cells in the nasal mucosa, including
5 epithelial cells (ciliated, columnar, and cuboidal), goblet cells, sensory neurons, and
6 intraepithelial lymphocytes. The histologic changes of these processes have been described in all
7 laboratory animals examined and progress from the anterior nares to the posterior regions of the
8 nasal passages, including the ETs and olfactory epithelium if the concentration and duration of
9 exposure are great enough.
10 Humans and nonhuman primates have significantly less complex nasal passages than
11 rodents. Formaldehyde has lower peak flux in human nasal tissues compared with rodents,
12 which are obligate nose breathers, but will penetrate more deeply into the human respiratory tract
13 than in rodents, since humans lack the autonomic RA response. Additionally, humans may
14 switch to mouth breathing in the presence of an irritant gas, thus bypassing the sensitive nasal
15 passages and increasing the tissue dose in the mouth and throat. These differences have been
16 demonstrated by using nonhuman primates where, at comparable concentrations, tissue
17 pathology and increased cell proliferation progressed further into the respiratory tract than in
18 rodents (Monticello et al., 1989). Nonhuman primates share common structural respiratory
19 components and patterns of breathing and do not have a reflex autonomic apnea response.
20 Despite the anatomical and physiological differences in breathing patterns and different
21 exposure parameters between humans and rodents, similar toxic effects are reported in tissues at
22 the POE in humans and laboratory animals. Several occupational studies have reported
23 increased squamous cell metaplasia in nasal and buccal samples in response to formaldehyde
24 exposure (Ballarin et al., 1992; Boysen et al., 1990; Holmstrom et al., 1989), paralleling the
25 histologic effects seen in experimental animal studies. A few human epidemiology studies
26 suggest increased NPCs (see Section 4.5) as well as oral/buccal tumors in response to
27 formaldehyde exposure (Shangina et al., 2006; Laforest et al., 2000).
28 The observed formaldehyde-induced URT toxicity is related to its high reactivity and
29 solubility. Moreover, additional interspecies differences in the surface area and configuration of
30 the nasal passages and upper airways will influence the areas of high formaldehyde flux in POE
31 tissues.
32
33 MO A
34 Formaldehyde-induced damage to the mucociliary apparatus of the nasal passages may
35 occur because formaldehyde may disrupt mucus flow and ciliary beat that is dependent on
This document is a draft for review purposes only and does not constitute Agency policy.
4-468 DRAFT—DO NOT CITE OR QUOTE
-------
1 concentration and duration of exposure. Formaldehyde reacts with the mucosal glycoproteins
2 and thus may contribute directly to the breakdown of the mucus layer. As formaldehyde reaches
3 the cells of the pseudostratified epithelium in the nasal passages, it exerts a range of effects from
4 direct damage to cell membrane, intracellular proteins, and DNA to alterations in GSH pools and
5 increased ROS. Adaptive effects include increased mucus flow and goblet cell proliferation as
6 well as the transition of respiratory epithelium to more insensitive cuboidal cells. With
7 continued exposure at sufficient concentration, squamous metaplasia develops, creating a
8 protective layer of keratinized cells. Gradually, this damage exceeds the cell's ability to
9 compensate for and repair damage; chronic nasal lesions develop, and the cells die both through
10 general necrosis as well as programmed cell death, depending on the severity of the cellular
11 damage (Monticello et al., 1989; Swenberg et al., 1983).
12 Genotoxic effects have been reported in nasal and buccal lesions taken from both workers
13 and students exposed to formaldehyde (Ying et al., 1997; Titenko-Holland et al., 1996; Suruda et
14 al., 1993). MN formation occurs in the more sensitive pseudostratified epithelium of the nasal
15 passages, nasopharynx, and upper airways, since there is only one layer of epithelial cells that are
16 constantly regenerating. However, the genotoxicity observed in buccal cells is more difficult to
17 explain, since buccal basal cells are usually covered by protective keratinized cell layers. Cuts,
18 sores, or other buccal lesions would increase basal epithelial cells' vulnerability to direct
19 exposure to formaldehyde.
20
21 4.4.5. Toxicogenomic and Molecular Data that may Inform MO As
22 Over the past several years, studies have begun to examine the effects of formaldehyde
23 exposure on gene and protein expression. These include studies on in vivo and in vitro changes
24 in the global expression of mRNA (transcriptomics) and proteins (proteomics) in the tissues and
25 cells of humans and rodents exposed to formaldehyde. Currently, nine "-omics" studies from
26 five research groups are available. These studies are summarized in Section 5.2 and are
27 evaluated and discussed in the context of their relevance to informing MO As and the dose-
28 response characterization briefly here.
29 In 2002, EPA released the Interim Policy on Genomics (U.S. EPA, 2002c), which
30 addresses how to use genomic data in regulatory decision making. Although the policy
31 encourages research in genomics, it places limits on its use, stating that genomic data alone are
32 not sufficient as a basis for decision making. These data thus cannot currently be utilized as the
33 critical effect in a chemical risk assessment but can be utilized in a weight-of-evidence approach
34 on a case-by-case basis. The Science Policy Council developed a white paper entitled Potential
3 5 Implications of Genomics for Regulatory and Risk Assessment Applications at EPA (U. S. EPA,
This document is a draft for review purposes only and does not constitute Agency policy.
4-469 DRAFT—DO NOT CITE OR QUOTE
-------
1 2004). This report described three areas where genomic data might be applied in risk assessment
2 at EPA: MOA analysis, susceptible population, and mixtures assessments. The genomic data on
3 formaldehyde thus may be applied to a discussion of MOA.
4 Toxicogenomics studies have investigated the gene and protein expression changes
5 resulting from formaldehyde exposure in a variety of respiratory tissues, including nasal tissue
6 (Andersen et al., 2008; Thomas et al., 2007; Hester et al., 2005, 2003), and, in lung tissue (Lee et
7 al., 2008; Li et al., 2007; Sul et al., 2007, 2006; Im et al., 2006) used human tracheal cell lines to
8 study genomic changes after exposure to formaldehyde in vitro. Unfortunately, these studies are
9 not directly comparable because different gene chip technology platforms were used in different
10 tissues, in both in vivo and in vitro study designs. In general, the gene and protein expression
11 changes reflect changes in apoptotic pathway genes, oxidative stress, and tissue remodeling.
12 Andersen et al. (2008) concluded that there was a threshold level where exposure to
13 formaldehyde (6 ppm) does not elicit changes in nasal epithelium of F344 rats. Overall,
14 Andersen et al. (2008) concluded that genomic changes were no more sensitive than tissue
15 responses and that formaldehyde, being an endogenous chemical, is well handled until some
16 threshold is achieved when toxicity rapidly ensues with genomic and histologic changes. At
17 about 6 ppm, this largely involves tissue remodeling (and protection), but regenerative
18 hyperplasia occurs at higher doses. Andersen et al. (2008) conclude that there is a threshold
19 where exposure to formaldehyde does not elicit changes in F344 nasal epithelial tissue over the
20 duration examined in this study (i.e., 15 days). Andersen et al. (2008) argue that this is
21 consistent with bioassays that indicate no tumor formation in rodents below 6 ppm
22 formaldehyde.
23 The primary conclusion in the Andersen et al. (2008) paper is that genomic changes,
24 including those suggestive of mutagenic effects, did not temporally precede or occur at lower
25 doses than phenotypic changes in the tissue. The authors stated as follows:
26
27 "The genomic signatures related to these transitions are for cell membrane and
28 extracellular components, then inflammation and cellular stress, and eventually
29 apoptosis. Importantly, these hierarchical models indicate that the tissue
30 responses at low dose concentrations are qualitatively different from those at high
31 concentrations and linear extrapolations or extrapolations that specify similar
32 modes of action at high and low doses would be inappropriate."
33
34 4.4.6. Noncancer Modes of Actions
35 Noncancer health effects of interest span numerous organ systems and include
36 reproductive and developmental effects, neurological/neurobehavioral effects, and a complex
37 interaction between inflammation and immune and adverse pulmonary function. To date, no
This document is a draft for review purposes only and does not constitute Agency policy.
4-470 DRAFT—DO NOT CITE OR QUOTE
-------
1 -omics studies have examined changes in reproductive tissue or altered gene expression in
2 developing animals. In regard to neurological/behavioral effects, one study (Lu et al. [2008],
3 described in Section 4.1.1.6) has reported elevations in the mRNA for NMDA receptor subunits
4 in brain homogenates following exposure to 2.4 ppm. Hester et al. (2003) reported a significant
5 increase in NMDA receptor subunit transcripts, along with other neuropeptide genes, in nasal
6 tissue of rats instilled into nostril with 400 mM formaldehyde (Described in Section 4.2.1.2.2.1).
7 Together, these changes may relate to formaldehyde-induced sensory irritation and, perhaps,
8 changes throughout the brain.
9 In regard to inflammatory, immune, and pulmonary effects, transcript and protein
10 changes in rodent tracheal tissue and lung tissue indicate that exposure to 3 to 38 ppm
11 formaldehyde results in genes involved in oxidative stress and cell proliferation and may
12 additionally increase airway ADH3 levels (Lee et al., 2008; Sul et al., 2007; Yi et al., 2007; Im et
13 al., 2006; Yang et al., 2005). Together, these data provide evidence for adverse pulmonary
14 effects that may exacerbate or facilitate asthma.
15 In lung tissue, Yang et al. (2005) identified three proteins up regulated and one protein
16 down regulated following 15 days of exposure to about 28 ppm formaldehyde. None of the
17 proteins corresponded with transcript changes reported by Sul et al. (2007). Interestingly, Sul et
18 al. (2007) reported that only two transcripts were significantly up regulated in the lung in
19 response to 5-10 ppm formaldehyde, while 19 were down regulated. In this regard, it is worth
20 considering that changes in proteins may not relate to their regulation but rather to their overall
21 percent composition in a cell (relative to other protein changes) before and after exposure. In
22 addition to transcript changes, Sul and colleagues (2007) reported DNA damage and lipid
23 peroxidation and noted that the observed down regulation of GR would facilitate oxidative stress,
24 while the down regulation of phospholipase A2 (PLA2) might represent a mechanism for
25 mitigating lipid peroxidation. It is worth noting that an increase in either of these genes could
26 also be argued to support similar conclusions (i.e., that GR is up regulated to increase GSH
27 levels and that PLA2 up regulation explains lipid peroxidation); this highlights the problem with
28 interpreting these data. Nevertheless, these studies indicate adverse effects (e.g., oxidative stress,
29 lipid peroxidation, cell proliferation etc.) in the rodent lung in response to 5-30 ppm
30 formaldehyde.
31 In a hypothesis-driven study by Yi et al. (2007), formaldehyde exposure was shown to
32 increase lung ADH3 levels. Several studies indicate that allergic responses and hyperreactivity
33 are uncoupled and may relate to ADH3 expression and activity. Que et al. (2005) has shown
34 that, in a rodent asthma model, ADH3 knockout mice exhibit similar signs of inflammation but
35 are protected from bronchoconstriction. Lino dos Santos Franco et al. (2006) provided evidence,
This document is a draft for review purposes only and does not constitute Agency policy.
4-471 DRAFT—DO NOT CITE OR QUOTE
-------
1 in rodents, that formaldehyde may induce inflammatory responses (e.g., leukocyte infiltration in
2 the lung) through neurogenic mechanisms but that bronchial tone is mediated by NO. The latter
3 effect is likely to be mediated by S-nitrosoglutathione GSNO and thus influenced by ADH3.
4 Interestingly, the single nucleotide polymorphisms (SNPs) that Wu et al. (2007) reported as
5 linked to other polymorphisms in the promoter region was the one demonstrating protection
6 against asthma. Hedberg et al. (2001) reported that at least one SNP in the promoter region
7 reduced ADH3 expression. Together, these data suggest that reduced ADH3 expression might
8 lower GSNO turnover and bronchial tone, thereby reducing signs of asthma. In this regard, it is
9 conceivable that wheezing and bronchoconstriction are the symptoms that lead to medical
10 intervention and not the inflammation per se. Thus, while ADH3 polymorphisms may not cause
11 asthma, ADH3 polymorphisms may influence hyperresponsivity and the likelihood of asthma
12 diagnosis. This is discussed further in Section 4.6 on susceptible populations. Formaldehyde
13 has been shown to accelerate GSNO breakdown (Staab et al., 2008a; Yi et al., 2007); thus,
14 pulmonary responses to formaldehyde may represent a balance between mechanisms that induce
15 NO (i.e., inflammation) and those that terminate GSNO (i.e., ADH3).
16 In regard to -omics changes in blood samples, the apparent limited distribution of
17 formaldehyde may suggest that these changes are secondary to effects at sites of contact but
18 could also indicate systemic distribution. As noted elsewhere in this report, bradypnea can
19 induce changes in dosimetry as well as changes in core body temperature and blood gases
20 (hypoxia itself induces hypothermia in rodents, and thus the reduction in minute volume and gas
21 changes may both contribute to hypothermia). These physiological responses (hypothermia and
22 hypoxia) surely induce changes in gene expression. Observed gene and protein changes in the
23 blood following formaldehyde exposure could also relate to irritation and inflammation at sites
24 of contact. In this regard, Im et al. (2006) reported changes in cytokines indicative of Th-2
25 responses. Altogether, the authors identified 32 proteins altered in the plasma of rats exposed to
26 formaldehyde. Although no coherent mechanisms are apparent from these changes, the authors
27 posited that they could serve as biomarkers for formaldehyde exposure. The concordance of
28 such changes across species remains to be demonstrated.
29 Li et al. (2007) identified dose-response relationships for six genes in human blood
30 samples that were putatively associated with formaldehyde exposure. Three of these genes are
31 reported to inhibit apoptosis and were posited as supporting in vitro data by Tyihak et al. (2001);
32 however, Li and colleagues (2007) did not report any increase in blood cell count or in Hs 680.Tr
33 cell counts in vitro (i.e., these changes were not phenotypically linked to changes in cell kinetics
34 or hematology). However, these findings are not inconsistent with those of Hester et al. (2003)
35 that indicated no significant increase (or decrease) in nine genes involved in apoptotic pathways.
This document is a draft for review purposes only and does not constitute Agency policy.
4-472 DRAFT—DO NOT CITE OR QUOTE
-------
1 Finally, serum and glucocorticoid-induced protein kinase 1 (SGK1) was elevated in blood
2 samples and was posited to relate to possible inflammatory and immune responses.
O
4 4.4.7. Immunotoxicity
5 Results from studies that evaluated the immunotoxicity of formaldehyde are mixed. For
6 example, most human studies that investigated systemic immune effects by measuring increases
7 in formaldehyde-specific IgE are negative (Doi et al., 2003; Kim et al., 2001; Palcynski et al.,
8 1999; Krakowiak et al., 1998; Wantke et al., 1996; Grammer et al., 1990). Vandenplas et al.
9 (2004) reported a transiently positive increased formaldehyde-specific IgE titer in occupationally
10 exposed workers. In contrast, Thrasher et al. (1990, 1988, 1987) reported positive
11 formaldehyde-specific IgE liters in small (six to eight person) case studies of exposed workers,
12 and Carraro et al. (1997) reported elevated IgE liters in smokers. Grammer et al. (1990) did not
13 report any differences in albumin IgE in formaldehyde-exposed workers compared with controls.
14 In a residential study, Pross et al. (1986) found that formaldehyde insulation in homes had no
15 effect on tested human immunologic parameters.
16 One study suggests that immune system parameters are perturbed by formaldehyde
17 exposure. Lyapina et al. (2004) reported decreased immune resistance in all 29 workers exposed
18 to formaldehyde. This effect was associated with decreased neutrophil respiratory burst activity.
19 A LOAEL of 700 ppb was established.
20 Results from animal studies are mixed as to whether formaldehyde causes
21 immunotoxicity. Leach et al. (1983) reported systemic immunomodulation in F344 rats that was
22 attributed to formaldehyde exposure, but the formaldehyde effects on measures of humoral and
23 cell-mediated immunity were not confirmed in B6C3F1 mice (Dean et al., 1984). Jakab et al.
24 (1992) detected no differences in phagocytic ability of alveolar MPs from mice after
25 formaldehyde exposure. Formaldehyde-exposed rats that were injected with pneumococcus
26 antigen or tetanus toxoid produced antibodies in amounts similar to nonexposed, infected control
27 animals (Holmstrom et al., 1989b).
28 However, specific immune parameters appear to be affected by formaldehyde exposure.
29 For example, increased host resistance and hydrogen peroxide release from peritoneal MPs were
30 reported and confirmed (Adams et al., 1987; Dean et al., 1984) and suggest a putative role for
31 ROS. Increased host resistance may be mediated by formaldehyde-induced chronic
32 inflammation and respiratory mucosal damage that causes an up regulation in MPs and therefore
33 increases host immunity. Jakab et al. (2002) reported reduced pulmonary bacterial resistance in
34 mice after exposure to formaldehyde, as determined by increased bacterial loading. This result
This document is a draft for review purposes only and does not constitute Agency policy.
4-473 DRAFT—DO NOT CITE OR QUOTE
-------
1 contrasts with Dean et al. (1984) and is attributed to differential exposure regimens and
2 experimental design.
O
4 Mode of Action
5 Circulating immune cells present in the mouth and upper airways, such as intraepithelial
6 lymphocytes, direct a local inflammatory response but may also contribute to systemic responses
7 through secreted cytokines and soluble factors released into the bloodstream (Togias, 1999).
8 Altered host resistance and hydrogen peroxide release from peritoneal MPs were reported
9 and confirmed (Adams et al., 1987; Dean et al., 1984) and suggest a putative role for ROS.
10 Indeed, increased neutrophilic ROS have been associated with formaldehyde-induced dermatitis
11 (Gorski et al., 1992), and, conversely, decreased neutrophil respiratory burst activity has been
12 shown in workers with history of formaldehyde-induced respiratory tract inflammation (Lyapina
13 et al., 2004). Oxidative stress may occur directly as a result of formaldehyde exposure or as a
14 secondary consequence to inflammatory responses.
15
16 4.4.8. Effects on the Nervous System
17 There is considerable evidence that formaldehyde exposure causes adverse effects on the
18 nervous system following inhalation at relatively low exposure levels (see Sections 4.1.1.6 and
19 4.2.6) but little or no information regarding a possible mechanism of action for these effects.
20 Data regarding adverse effects on the nervous system following oral exposure are very limited,
21 reflecting a data gap in this area. Relevant data in animals and humans for several types of
22 neurological endpoints, following inhalation exposure, are summarized below.
23
24 4.4.8.1. Irritant Threshold Detection
25 Humans are exquisitely sensitive to the irritant properties of formaldehyde, as has been
26 discussed previously (see Section 4.1.1.1). Animal data confirm the irritant properties of this
27 compound at very low concentrations (Wood and Coleman, 1995).
28
29 4.4.8.2. Behavioral Effects
30 Limited data in humans, as well as more robust data in animals, provide evidence of
31 behavioral changes following exposure to formaldehyde at levels as low as 0.1 ppm. Studies in
32 animals have found effects that persist for days to weeks after termination of exposure. In spite
33 of significant limitations, the available human data are consistent with the animal findings.
34 Several types of behavior have been evaluated in animals following formaldehyde
35 exposures. The most consistent findings, demonstrated by multiple laboratories and in multiple
This document is a draft for review purposes only and does not constitute Agency policy.
4.474 DRAFT—DO NOT CITE OR QUOTE
-------
1 species, have been changes in motor activity, habituation, and learning/memory task
2 performance. Motor activity and habituation have been evaluated under a variety of exposure
3 conditions, using both rats and mice. Consistent decreases in activity have been seen in adult
4 animals (Malek et al., 2004, 2003 a, b; Usanmaz et al., 2002). Senichenkova (1991) and
5 Sheveleva (1971) also found changes in motor activity in offspring following in utero exposure,
6 including decreased habituation in juvenile rats exposed in utero. Decreased performance in
7 learning and/or memory paradigms have been seen in multiple laboratories, also in both rats and
8 mice (Lu et al., 2008; Malek et al., 2003c; Pitten et al., 2000).
9 Data from controlled human exposures are very limited, but studies have shown
10 decreased performance in addition tasks and reaction time tasks following acute exposures to
11 formaldehyde (Lang et al., 2008; Bach et al., 1990). In contrast, Andersen and Molhave (1983)
12 indicated they found no change in performance on several types of tasks (including addition,
13 multiplication, and card punching) following acute exposure to volunteers, but supporting data
14 were not provided.
15 Data for humans are also available from epidemiology studies of individuals who were
16 occupationally exposed to formaldehyde. A variety of neurobehavioral deficits, including lack
17 of concentration and loss of memory, disturbed sleep, impaired balance and dexterity, and
18 changes in mood, were identified (Kilburn et al., 1987, 1985). However, most of the individuals
19 evaluated in these studies were also occupationally exposed to other solvents, raising questions
20 regarding possible confounding of the results due to multiple exposures. In addition, the
21 formaldehyde exposure information provided in the studies is not sufficient to permit a reliable
22 dose-response assessment for the effects identified. The types of effects seen in humans in the
23 available epidemiology studies are, however, consistent with those seen in available animal
24 studies.
25 In general (and noting the differences in exposure paradigms and types of tasks),
26 behavioral effects in animals and humans appear to occur at similar exposure levels. Animal
27 studies demonstrated LOAELs as low as 100 ppb following acute or repeated exposures (Malek
28 et al., 2003b, c); human controlled exposure studies have found effects in that same range, with
29 LOAELs of approximately 300 ppb following acute exposures (Lang et al., 2008; Bach et al.,
30 1990).
31
32 4.4.8.3. Neurochemistry, Neuropathology, and Mechanistic Studies
33 Limited data are available regarding neurochemical and neuropathological sequelae of
34 formaldehyde exposure. Studies from one laboratory (Sorg et al., 2004, 2001) have suggested
35 that behavioral sensitization to formaldehyde is linked to alterations in HPA control of
This document is a draft for review purposes only and does not constitute Agency policy.
4-475 DRAFT—DO NOT CITE OR QUOTE
-------
1 corticosterone and changes in mesolimbic dopamine pathways. Neurochemical changes in
2 response to formaldehyde exposure have also been documented in other laboratories (Fujimaki et
3 al., 2004b; Hayashi et al., 2004). Some of these data appear to be conflicting, and there are no
4 definitive data supporting a specific mechanism for formaldehyde effects on the nervous system
5 at this time. Neuropathological data are also limited, although data from one laboratory
6 (Sarsilmaz et al., 2007; Asian et al., 2006) suggest a concern for changes in brain structure in
7 neonatal rats following exposure at 6 or 12 ppm. No human data are available that address these
8 endpoints. However, a prospective cohort study of nearly one million people followed for 15
9 years reported strongly significant dose-response associations between death from ALS and
10 exposure to formaldehyde of a known duration, with longer exposures associated with greater
11 risk (Weisskopf et al., 2009). This large, well-designed prospective cohort study strongly
12 supports the causal association of neuropathological effects in humans following long-term
13 formaldehyde exposure.
14
15 4.4.8.4. Summary
16 Overall, there is strong evidence that formaldehyde exposure via inhalation may cause
17 adverse effects on nervous system function in experimental animals at relatively low levels of
18 exposure (LOAELs as low as 100 ppb). Although human data regarding neurotoxicity following
19 formaldehyde inhalation are limited, available data provide support that the types of effects seen
20 in humans are similar to those found in animal studies. Evidence from available human
21 controlled inhalation exposure studies indicates that humans may be affected at doses similar to
22 those used in animal studies; however, the human data are extremely limited.
23 There are insufficient data to evaluate the potential for neurotoxicity following oral
24 exposure to formaldehyde. Limited evaluations of brain weight or histopathology in available
25 chronic or subchronic oral studies found no evidence of formaldehyde-induced changes (Til et
26 al., 1989, 1988; Tobe et al., 1989; Johannsen et al., 1986). However, reliable studies examining
27 nervous system function or focused studies of neuropathology following oral exposure to
28 formaldehyde are not available.
29
30 4.4.8.5. Data Gaps
31 Major data gaps were found regarding the evaluation of changes in nervous system
32 structure or function following formaldehyde exposure by both the inhalation or oral routes.
33 With respect to inhalation exposure, none of the available human studies resulted in data
34 sufficient to conduct a reliable dose-response assessment for changes in nervous system function.
35 Most of the available animal inhalation studies used short exposure durations (acute or short-
This document is a draft for review purposes only and does not constitute Agency policy.
4-476 DRAFT—DO NOT CITE OR QUOTE
-------
1 term), precluding a reliable evaluation of neurotoxicity following chronic exposure. Available
2 data for neurodevelopmental exposures are also quite limited, consisting of evaluation of
3 neuropathology in only one brain region and functional evaluations focused only on changes in
4 motor activity.
5 Major data gaps also exist regarding neurotoxicity following oral exposure, with no
6 relevant human data and extremely limited animal data. Available oral exposure studies were
7 insufficient to permit a reliable evaluation of the potential for neurotoxicity following oral
8 exposure to formaldehyde.
9
10 4.4.9. Reproductive and Developmental Toxicity
11 A number of studies have been identified that indicate an effect of formaldehyde
12 exposure on reproductive and developmental outcomes. Human data are described in
13 Section 4.1.1.7, and animal studies are addressed in Section 4.2.7 of this document.
14
15 4.4.9.1. Spontaneous Abortion and Fetal Death
16 Several epidemiologic studies reported increases in risk of spontaneous abortion
17 following maternal occupational formaldehyde exposure (Taskinen et al., 1999, 1994; John et al.,
18 1994; Seitz and Baron, 1990; Axelsson et al., 1984). While these finding have been questioned
19 (Collins et al., 2001b), upon careful examination, none of the principal biases in epidemiologic
20 studies, including information bias, selection bias, and confounding, appear to be more likely
21 causes of these reported findings than the conclusion that they may reflect an underlying causal
22 process. While each of these occupational studies focused on women who were coexposed to
23 formaldehyde and other chemicals, the occupational groups were quite different and had
24 different sets of coexposures. The woodworkers in the Taskinen et al. (1999) study were
25 potentially coexposed to organic solvents related to painting and lacquering, dusts, and phenols,
26 none of which was shown to be an independent predictor of adverse risk. The cosmetologists
27 studied by John et al. (1994) were coexposed to hair dyes, bleach, alcohol-based disinfectants,
28 and chemicals specific to services, such as fingernail sculpturing, but, in analyses that were
29 specifically adjusted for other work exposures and their potentially confounding effects, the
30 investigators reported an increased risk for the use of formaldehyde-based disinfectants. The
31 laboratory workers studied by Axelsson et al. (1984) were potentially coexposed to a wide range
32 of solvents, but the miscarriage rate was highest among those exposed to formaldehyde. For a
33 potential confounder to entirely explain an observed effect of another exposure, it must be more
34 strongly associated with the adverse outcome. It does not appear that the collective results of
This document is a draft for review purposes only and does not constitute Agency policy.
4-477 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde exposures associated with increased risk of spontaneous abortion—often in spite
2 of exposures being crudely measured—can be explained by information bias or confounding.
3 Taken together, these findings are consistent with an adverse effect of formaldehyde
4 exposure on the risk of pregnancy loss. The single study with the strongest quantitative
5 assessment of that risk is Taskinen et al. (1999), and the results presented are of sufficient quality
6 to support quantitative risk assessment by using the LOAEL/NOAEL approach.
7 This study was a well-designed population-based case-control study that appears to have
8 been well executed and appropriately analyzed. The study population of Finnish women was
9 well defined and adequately selected to allow for meaningful comparisons of health effects
10 between individuals with different levels of exposure to formaldehyde. The participation rate
11 was 64%, which is low enough to raise a concern about the potential for selection bias.
12 However, the authors noted that selection bias has not influenced the results of other
13 reproductive epidemiology studies reporting results on smoking, irregular menstruation, and
14 earlier miscarriages, which are known to lengthen the time to pregnancy (Bolumar et al., 1996;
15 Sallmen et al., 1995; Baird and Wilcox, 1985). Furthermore, there is no evidence to support
16 conjecture that an individual's decision to participate in this study would be differential with
17 respect to their workplace formaldehyde exposures while being nondifferential with respect to
18 the other exposures of interest, including organic solvents, wood dust, and phenols. Since the
19 women who chose to participate in this study were not likely to be aware of the specific
20 hypotheses under investigation nor could they have known the formaldehyde exposures that were
21 independently estimated by an industrial hygienist, selection bias is not a likely explanation for
22 the findings of adversity.
23 Data on pregnancy history, including spontaneous abortions, were collected by
24 questionnaire. Spontaneous abortion is the most common adverse outcome of pregnancy (Klein
25 et al., 1989), and retrospective self-report of spontaneous abortion has been found to match well
26 with prospectively collected reproductive histories (Wilcox and Horney, 1984). Many
27 spontaneous abortions, however, are missed with self-reporting, with the magnitude likely
28 exceeding 25%, but only rarely do women self-report false positive events (Wilcox and Horney,
29 1984). The effect of such an undercount is to cause a bias towards the null when the likelihood
30 of undercounting is unrelated to formaldehyde exposure. The implication is that the observed
31 association of increased risk of spontaneous abortion associated with occupational exposure to
32 formaldehyde may be an underestimation of the true risk.
33 The findings by Taskinen et al. (1999) of reduced fertility and increased risk of
34 spontaneous abortion are internally consistent and coherent with other reports of increased risk
35 of pregnancy loss associated with exposure to formaldehyde (John et al., 1994; Taskinen et al.,
This document is a draft for review purposes only and does not constitute Agency policy.
4-478 DRAFT—DO NOT CITE OR QUOTE
-------
1 1994; Seitz and Baron, 1990; Axelsson et al., 1984). Absent evidence of alternative explanation
2 for these findings, it is concluded that exposure to formaldehyde is associated with pregnancy
3 loss and diminished fertility.
4 In animal studies, Sheveleva (1971) noted an increase in preimplantation loss in rats
5 exposed to 0.04 and 0.4 ppm formaldehyde by inhalation on GDs 1-19, and Kitaev et al. (1984)
6 observed evidence of degeneration in harvested embryos on GD 2, following 4 months of
7 maternal inhalation exposure to 0.41 ppm formaldehyde in rats. In a second series of tests
8 reported in Kitaev et al. (1984), female rats were exposed to 0.41 and 1.22 ppm formaldehyde for
9 4 months to test the hypothesis that the embryo degeneration could have been the result of
10 disrupted reproductive hormone levels in the dams. Ovarian weight and blood levels of LH were
11 increased at 0.41 ppm (but not at 1.22 ppm), and significantly increased levels of FSH were
12 observed at 1.22 ppm. Kitaev et al. (1984) proposed that effects at the 0.41 ppm might be related
13 to disruption of the hypothalamic-pituitary axis and that at the higher exposure level (1.22 ppm)
14 frank toxic effects to the embryo were observed. The increased FSH levels at 1.22 ppm may also
15 be indicative of hormonal perturbations induced by formaldehyde exposure that could affect
16 pregnancy maintenance in humans. The finding of treatment-related increased preimplantation
17 loss in rats appears to support the evidence of spontaneous abortion in the epidemiologic data. In
18 addition, a dominant lethal study in rats by Odeigah (1997) identified significant
19 postimplantation loss following premating IP. formaldehyde exposures to males, suggesting a
20 potential MOA involving germ cell toxicity. Nevertheless, a number of developmental toxicity
21 studies in rats did not report treatment-related embryolethality following gestation exposures to
22 formaldehyde. These included inhalation studies by Martin (1990), Saillenfait et al. (1989), and
23 Kilburn and Moro (1985), a series of studies by Gofmekler and Bonashevskaya (1969),
24 Gofmekler (1968), and Pushkina et al. (1968), and studies by Senichenkova and Chebotar (1996)
25 and Senichenkova (1991). It is noted, however, that, to the extent that these studies evaluated
26 embryonic or fetal death, the observations were conducted late in gestation and the studies may
27 not have been designed to detect the preimplantation losses as observed in Kitaev et al. (1984)
28 and Sheveleva (1971). Additionally, a number of the reports for these studies did not include
29 sufficient details to engender a high degree of confidence in the reported results. Fetal death was
30 also not observed in oral studies with formaldehyde in beagle dogs (Hurni and Ohder, 1973) and
31 rats (Seidenberg and Becker, 1987).
32
This document is a draft for review purposes only and does not constitute Agency policy.
4-479 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.4.9.2. Congenital Malformations
2 The effect of occupational exposures to formaldehyde on the incidence of congenital
3 malformations was examined by Dulskiene and Grazuleviciene (2005), Taskinen et al. (1994),
4 and Hemminki et al. (1985). Results were mixed.
5 In animal studies, the most frequently observed structural anomaly noted following
6 inhalation exposures to formaldehyde during gestation was a delay in fetal (i.e., 1st stage) testes
7 descent (Senichenkova and Chebotar, 1996; Senichenkova, 1991; Kilburn and Moro, 1985),
8 although similar findings were not reported by Saillenfait et al. (1989) or Martin (1990) in what
9 appear to be well-conducted prenatal developmental toxicity studies. No study in the available
10 database specifically examined the 2nd stage of postnatal testes descent in pups. Thus, there is no
11 evidence to determine if the observed effect represented a developmental delay or if it was
12 related to disruptions in male reproductive tract ontogeny, which is dependent on normal levels
13 of fetal testicular testosterone and on the expression of insulin-like hormone-3 (ins!3) in fetal
14 Leydig cells (Klonisch et al., 2004). Senichenkova (1991) observed an increased incidence of
15 other organ anomalies following formaldehyde exposure during gestation; however, the
16 anomalies are not characterized in the report. Alterations on fetal organ weights and/or size were
17 noted in several studies (Kilburn and Moro, 1985; Gofmekler, 1968), but it is difficult to
18 ascertain if these findings represented agenesis, hypoplasia, or evidence of systemic organ
19 toxicity. Histopathologic evaluation of pup tissues following maternal gestational exposures to
20 0.01 and 0.81 ppm formaldehyde was conducted by Gofmekler and Bonashevskaya (1969),
21 revealing reduced glycogen content in the myocardium, the presence of iron in hepatic Kupffer
22 cells, and a positive Schiff reaction in the basement membrane (indicating functional alterations
23 in the renal tubule) at both exposure levels.
24
25 4.4.9.3. Low Birth Weight and Growth Retardation
26 A population-based study (Grazuleviciene et al., 1998) reported an association between
27 atmospheric formaldehyde exposure and low birth weight, with an adjusted OR of 1.37 (95% CI:
28 0.90-2.09).
29 A number of inhalation studies in rats identified reduced fetal weight as an adverse
30 outcome of in utero formaldehyde exposure and are supportive of the association noted in
31 humans. In a study that exposed pregnant rats to formaldehyde during GDs 6-20, Saillenfait et
32 al. (1989) observed significantly decreased male and female fetal rat weights (78 and 81% of
33 control values, respectively) at 40 ppm formaldehyde. In a study that exposed the dams from
34 GDs 6-15, Martin (1990) found decreased fetal weights at exposure levels of 5 and 10 ppm. In
35 both studies, observations of reduced or delayed skeletal ossification (i.e., the thoracic vertebrae
This document is a draft for review purposes only and does not constitute Agency policy.
4-480 DRAFT—DO NOT CITE OR QUOTE
-------
1 in Saillenfait et al. [1989] and the pubes and ischial bones in Martin [1990]) were consistent with
2 the fetal weight deficits. Kilburn and Moro (1985) also reported fetal body weight decreases in
3 rats at an inhalation exposure level of 30 ppm. Conversely, increased fetal body weight as
4 compared with controls (generally considered to be nonadverse) was noted by Gofmekler (1968)
5 and Pushkina et al. (1968) at maternal formaldehyde exposure levels of 0.1 and 0.81 ppm
6 administered for approximately 2-3 weeks prior to mating and then throughout gestation.
7 Increased fetal weight was also noted in rats by Senichenkova (1991) and Senichenkova and
8 Chebotar (1996) following maternal exposures to 0.41 ppm formaldehyde throughout gestation.
9 Studies that assessed the effects of oral administration of formaldehyde on development
10 are quite limited. The only oral study identified that found a treatment-related effect on offspring
11 growth was a study using beagle dogs (Hurni and Ohder, 1973). In this study, formaldehyde was
12 administered at doses of 3.1 or 9.3 mg/kg-day in the feed during gestation, and pup weight
13 decrements at postnatal week 8 were 6.3 and 12% in the low- and high-dose groups, respectively.
14
15 4.4.9.4. Functional Developmental Outcomes (Developmental Neurotoxicity)
16 Indications of effects on the developing nervous system were observed in several rodent
17 studies, although no similar epidemiologic findings in children were identified. These studies
18 (Sarsilmaz et al., 2007; Asian et al., 2006; Weiler and Apfelbach, 1992; Senichenkova, 1991;
19 Sheveleva, 1971) are described in detail in Section 4.2.6. In the studies by Asian et al. (2006)
20 and Sarsilmaz et al. (2007), neonatal rats were exposed to formaldehyde for 30 days at 6,000 or
21 12,000 ppb. Decreases in the volume of discrete areas of the brain were observed at both
22 exposure levels in both studies, and, additionally, decreased cell numbers were noted in a region
23 of the hippocampus in the Sarsilmaz et al. (2007) study. Weiler and Apfelbach (1992) exposed
24 juvenile rats to 0.25 ppm formaldehyde for 130 days or adult rats to 0.5 ppm formaldehyde for
25 90 days. Olfactory thresholds measured in this study were significantly higher in the rats that
26 had been exposed as juveniles than in those that had been exposed only as adults. Sheveleva
27 (1971) observed alterations in spontaneous mobility in 1- and 2-month-old pups from dams that
28 had been exposed to formaldehyde at 0.04 or 0.4 ppm throughout gestation. In the Senichenkova
29 (1991) study, maternal rats were exposed to 400 ppb formaldehyde during GDs 1-19, and
30 functional observational testing was conducted on the juvenile offspring. Changes in open-field
31 motor activity, exploratory activity, and habituation were observed in the offspring.
32
33 4.4.9.5. Male Reproductive Toxicity
34 A number of laboratory animal studies have reported effects of formaldehyde exposure
35 on male reproductive system endpoints. These effects include decreased testes weight, changes
This document is a draft for review purposes only and does not constitute Agency policy.
4-481 DRAFT—DO NOT CITE OR QUOTE
-------
1 in Leydig cell quantity and quality, degeneration of seminiferous tubules, decreased testosterone
2 levels, alterations in biomarkers of toxicity in the testes, and alterations in sperm measures
3 (Galilapour et al., 2007; Xing et al., 2007; Zhou et al., 2006; Ozen et al., 2005, 2002; Sarsilmaz
4 et al., 1999; Odeigah, 1997; Majumder and Kumar, 1995; Chowdhury et al., 1992; Til et al.,
5 1989, 1988; Tobe et al., 1989; Johanssen et al., 1986; Maronpot et al., 1986; Cassidy et al., 1983;
6 Appelman et al., 1982; Guseva, 1972). Following concurrent exposures to formaldehyde in air
7 and drinking water for 6 months, Guseva (1972) found decreases in testicular nucleic acid levels.
8 In a study conducted by Chowdhury et al. (1992), rats were administered IP. injections of
9 formaldehyde for 30 days, and evidence of decreased testes weight, serum testosterone levels,
10 and Leydig cell quality was observed. Sarsilmaz et al. (1999) followed up on these findings
11 (exposing male rats to formaldehyde via inhalation at 10 and 20 ppm for 4 weeks) and found
12 decreases in Leydig cell quantity and the percentage of cells with normal nuclei. Hypothesizing
13 that the reported decreases in Leydig cell quality may have been the result of oxidative stress and
14 damage, Ozen et al. (2002) evaluated biomarkers of such changes and found that testicular zinc
15 and copper levels were decreased and iron levels were increased following exposures of adult
16 male rats to 10 and 20 ppm formaldehyde for 4 or 13 weeks. Additionally, relative testes weight
17 was decreased in a dose- and duration-dependent manner, although this effect had not been
18 observed by Sarsilmaz et al. (1999). Ozen et al. (2005) noted decreased serum testosterone
19 levels, decreased seminiferous tubule diameters, and increased levels of heat shock protein in
20 spermatogonia, spermatocytes, and spermatids of rats following 91 days of exposure to 10 ppm
21 formaldehyde. A study by Golalipour et al. (2007) observed decreased numbers of testicular
22 germ cells, altered spermatogenesis, and reduced seminiferous tubular diameter and epithelial
23 height in rats following 18 weeks of formaldehyde inhalation exposure; the severity of the
24 seminiferous tubule pathology was positively correlated to the number of hours/week of
25 exposure. Zhou et al. (2006) found decreased testis weight, atrophy of seminiferous tubules,
26 edematous interstitial tissue, and alteration of epididymal sperm count, morphology, and motility
27 in rats after 2 weeks of formaldehyde exposure at 8 ppm. Abnormal sperm were also observed in
28 mice by Xing et al. (2007) after 13-weeks of inhalation exposure at 16.9 ppm, and Cassidy et al.
29 (1983) reported sperm abnormalities in rats following a single oral dose of 200 mg/kg.
30 Additionally, Majumder and Kumar (1995) observed significantly reduced sperm count, motility,
31 and viability following 30 days of IP. injection of 10 mg/kg-day formaldehyde to male rats.
32 Also in this study, the ability of formaldehyde to affect sperm parameters was confirmed with in
33 vitro testing. A study conducted by Odeigah (1997) demonstrated epididymal sperm count and
34 morphology abnormalities following five IP. doses of >0.125 mg/kg formaldehyde and
This document is a draft for review purposes only and does not constitute Agency policy.
4-482 DRAFT—DO NOT CITE OR QUOTE
-------
1 additionally identified dominant lethal effects (decreased live embryos and increased dead
2 implants) following mating of treated male rats with untreated females.
3 Although Til et al. (1989) reported low incidences of Ley dig cell tumors in
4 formaldehyde-treated rats in a chronic drinking water study, no alterations in testes weight or
5 histopathologic lesions of the testes were observed in subchronic inhalation studies conducted by
6 Appleman et al. (1982) or Maronpot et al. (1986) or in subchronic or chronic oral studies by
7 Johanssen et al. (1986), Til et al. (1988), or Tobe et al. (1989).
8 No epidemiologic studies have identified an association between formaldehyde exposure
9 and alterations in the male reproductive system (e.g., see Ward et al. [1984]).
10
11 4.4.9.6. Female Reproductive Toxicity
12 The available database for the assessment of the effects of formaldehyde exposure on the
13 female reproductive system was limited. In addition to the findings of spontaneous abortions, as
14 described above, Taskinen et al. (1999) examined fecundability in a cohort of healthy female
15 wood-processing industry workers and found that conception was significantly delayed in
16 women who were occupationally exposed to formaldehyde. The FDR, a ratio of average
17 incidence densities of pregnancies for exposed female employees compared with unexposed
18 female employees, was lower in women exposed to mean formaldehyde levels of approximately
19 0.33 ppm (range: 0.15-1.00 ppm) compared with controls (adjusted FDR = 0.64 [95% CI:
20 0.28-0.92]). An FDR <1.0 is indicative of delayed conception, which is an indicator of reduced
21 fertility. The subfertility observed in this study is supportive of the association observed
22 between formaldehyde exposure and spontaneous abortion, since subclinical pregnancy losses
23 are increased in women with compromised fertility (Gray and Wu, 2000; Hakim et al., 1995),
24 and both spontaneous abortion and subfertility can be related to exposure to environmental
25 toxicants (Correa et al., 1996).
26 As described above, formaldehyde exposures to female rats resulted in decreased ovarian
27 weight and altered LH and FSH levels (Kitaev et al., 1984). Maronpot et al. (1986) reported
28 endometrial hypoplasia and lack of ovarian luteal tissue in female mice exposed to 40 ppm
29 formaldehyde for 13 weeks. Additionally, it is noted that, in developmental toxicity studies that
30 included repeated exposures of dams before mating and/or during gestation, reports of adverse
31 pregnancy outcomes were few. Gofmekler (1968) reported an increase in pregnancy duration
32 and decrease in litter size; however, this finding was not observed in other studies.
33 With the exception of spontaneous abortion and increased time to pregnancy, associations
34 of formaldehyde exposure with adverse female reproductive system outcomes were not observed
35 in the available epidemiologic data.
This document is a draft for review purposes only and does not constitute Agency policy.
4-483 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.4.9.7. Mode of Action
2 A strong case cannot be made for any one MO A that explains one or more of the
3 reproductive and developmental outcomes observed in formaldehyde epidemiologic or
4 toxicology studies. This is due to a number of issues, including the following:
5 (1) inconsistencies in study findings for the toxicology studies, which may be explained by study
6 quality issues (see detailed descriptions of studies in Sections 4.1 and 4.2); (2) few studies that
7 allow for comparisons because no study was performed with the same study design (e.g., stage of
8 exposure, dose, species, and strain); (3) few mechanistic studies available to test hypothesized
9 MO As; and (4) a bias that is pervasive in the formaldehyde literature that outcomes observed
10 beyond the POE (the nose) are not expected from inhalation exposure, which is the route of
11 exposure for most of the developmental and reproductive studies. This discussion presents
12 putative MO As that have been hypothesized by study authors and the studies that support the
13 hypothesized MO As. The four hypothesized MO As are not mutually exclusive. They could be
14 acting alone for certain endpoints (in which case the others are not operable) or in concert for
15 certain endpoints.
16 The focus of this discussion is on analyzing possible MO As for the developmental and
17 reproductive outcomes that were noted most consistently, across toxicology studies and, in some
18 cases, across human and animal studies. These outcomes include developmental delays, fetal
19 loss, and sperm quality and quantity effects.
20 An endocrine-disrupting MOA is supported by some of the reproductive and
21 developmental epidemiology and toxicology studies. For example, the decreases in fetal body
22 weight (Martin, 1990; Saillenfait et al., 1989), delayed ossifications (Senichenkova and
23 Chebotar, 1996; Senichenkova, 1991; Martin, 1990; Saillenfait et al., 1989), and delayed
24 eruption of incisors (Senichenkova, 1991) noted in rats after gestational exposure to
25 formaldehyde are consistent with developmental delays. In turn, developmental delays can result
26 from effects on the hypothalamic-gonadal-pituitary axis in the dam that cause hormonal level
27 changes in the pup; however, hormone levels in pups were not measured. Kilburn and Moro
28 (1985) also observed organ size changes and undescended testes after developmental
29 formaldehyde exposure. These outcomes can also be explained by an endocrine MOA. There
30 are three studies that directly tested for changes in hormones after formaldehyde exposure.
31 Kitaev et al. (1984) observed ovarian weight and serum LH and FSH increases after inhaled
32 formaldehyde in adult female rats. Chowdhury et al. (1992) assessed serum testosterone levels
33 in adult rats after formaldehyde IP exposure and found significant decreased testosterone and
34 testes weights and a decrease in 3-p,A-5-hydroxy steroid dehydrogenase in Leydig cells,
35 suggesting that formaldehyde affects steroidogenesis. Ozen et al. (2002) also reported
This document is a draft for review purposes only and does not constitute Agency policy.
4-484 DRAFT—DO NOT CITE OR QUOTE
-------
1 significant serum testosterone level decreases as well as decreased mean seminiferous tubule
2 diameters. Furthermore, the steroidogenesis MOA leading to reduced testosterone is also
3 consistent with the sperm quality and quantity decrements observed in the studies by Ozen et al.
4 (2002), Sarsilmaz et al. (1999), and Odeigah (1997) studies.
5 In human studies, delayed time to pregnancy and increased incidence of spontaneous
6 abortion (Taskinen et al., 1999), consistent with some study findings from the toxicology
7 literature, could also be explained by an endocrine MOA. Alterations in hormone levels could
8 lead to pregnancy maintenance problems. Extrapolating the Chowdhury et al. (1992) results of
9 the steroidogenesis MOA to females, formaldehyde exposure could affect progesterone levels
10 required for pregnancy. However, progesterone levels were unchanged in the female rat in the
11 one study that assessed progesterone (Kitaev et al., 1984). Consistent with an endocrine
12 mediated MOA, Maronpot et al. (1986) observed endometrial hypoplasia and lack of ovarian
13 luteal tissue in females exposed to formaldehyde.
14 A second hypothesized MOA for some of the developmental and reproductive outcomes
15 is genotoxicity of the gametes. Such an MOA could explain pregnancy loss in humans
16 (Taskinen, et al., 1999) and preimplantation loss in animal studies (Xing et al., 2007; Kitaev et
17 al., 1984; Sheveleva, 1971) and fetal viability (e.g., litter size decreases) after formaldehyde
18 exposure. Consistent with male gamete genotoxicity, Odeigah (1997) and Xing et al. (2007)
19 observed reduced fertile matings and live embryos, and increased dead implants in a dominant
20 lethal study.
21 Oxidative stress/damage is another MOA that is consistent with testicular toxicity, sperm
22 effects, and reduced embryo viability. Ozen et al. (2002) investigated the mechanism of
23 oxidative stress being responsible for the testes quality effects by assessing testicular iron,
24 copper, and zinc levels. Zinc and copper levels were reduced in the rat testes, consistent with an
25 oxidative stress MOA. Ozen et al. (2002) also reported increased iron levels and decreased zinc
26 levels in the lung, consistent with oxidative stress. Another study (Zhou et al., 2006) that
27 investigated the oxidative stress MOA in the testes observed significant changes in oxidative
28 stress biochemical markers (decreases in SOD, GPX, GSH, and an increase in MDA levels).
29 The authors also assessed the protective effect of coadministration with vitamin E, an
30 antioxidant, on decreased testes weight, biochemical alterations, histopathologic effects, and on
31 sperm count, motility, and morphology. The study of Pushkina et al. (1968) found reduced
32 levels of Vitamin C, another antioxidant, in the fetus and maternal liver after formaldehyde
33 exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
4-485 DRAFT—DO NOT CITE OR QUOTE
-------
1 The MO As proposed are not mutually exclusive and in fact could interact with one
2 another. For example, an endocrine MOA could lead to oxidative stress, and that oxidative stress
3 could lead to genotoxicity.
4
5 4.4.9.8. Data Gaps
6 The inhalation developmental toxicity studies conducted on formaldehyde and described
7 in Section 4.2.7 comprise an adequate assessment of prenatal developmental toxicity for
8 application to inhalation reference concentration determination. The assessments of postnatal
9 developmental toxicity and of reproductive function following inhalation of formaldehyde are
10 less complete. It is notable that, although the database contains some studies that assess various
11 aspects of reproductive organ system toxicity, particularly in males, there is no assessment of
12 multigenerational reproductive function, such as would be evaluated in a two-generation
13 reproductive toxicity study, nor is there an assessment of potential reproductive effects of
14 formaldehyde exposure in human males.
15 Adequate assessments of developmental and reproductive toxicity following oral
16 exposures to formaldehyde have not been conducted.
17
18 4.5. SYNTHESIS AND EVALUATION OF CARCINOGENICITY
19 4.5.1. Cancers of the Respiratory Tract
20 Epidemiologic studies of formaldehyde-exposed workers provide sufficient evidence of a
21 causal association between formaldehyde exposure and nasopharyngeal cancer (see
22 Section 4.1.2.1.1) as well as nasal and paranasal cancers (see Section 4.1.2.1.2). The
23 epidemiologic evidence of association between formaldehyde exposure and other upper
24 respiratory tract cancers (see Section 4.1.2.1.3) is consistent with, and supportive of, a causal
25 association but insufficient on its own to reach a causal conclusion. However, taken together
26 with the causal evidence of an association between formaldehyde and nasopharyngeal cancer and
27 sinonasal cancer in neighboring tissues of the upper respiratory tract and sites of first contact
28 with inhaled formaldehyde, along with the strongly supportive evidence of association in
29 animals, the evidence is sufficient to conclude that formaldehyde is causally related to cancers of
30 the upper respiratory tract as a group.
31 The observational evidence from epidemiologic studies reporting associations between
32 formaldehyde exposure and increased risk of nasopharyngeal cancer supports a conclusion of a
33 causal association for this specific cancer. However, epidemiologic studies of rare outcomes
34 such as nasopharyngeal cancer, which has an expected incidence of 1 per 100,000 people per
35 year in the United States, do not typically have great statistical power to rule out the null
This document is a draft for review purposes only and does not constitute Agency policy.
4-486 DRAFT—DO NOT CITE OR QUOTE
-------
1 hypothesis (i.e., no association). However, the weight of evidence of the several studies
2 reviewed in Section 4.1.2.1.1 provide an accumulation of consistent observational evidence
3 sufficient to exclude chance as an explanation for the association. Additionally, there is
4 insufficient evidence of consistent confounding or other bias across the studies that were
5 considered; thus, confounding and bias were also ruled out as explanations for the observed
6 association. The lack of a convincing and consistent alternative hypothesis of causation - in
7 spite of repeated examinations - further supports the conclusion that the association between
8 formaldehyde exposure and increased risk of nasopharyngeal cancer is causal.
9 The single strongest cohort study, Hauptmann et al. (2004), shows a statistically
10 significant exposure-response relationship between formaldehyde exposure and upper respiratory
11 tract cancers. Hauptmann et al. (2004) demonstrated significant excess risk of nasopharyngeal
12 cancer in exposed workers based on U.S. population death rates (standardized mortality ratio
13 [SMR] = 2.1; 95% confidence index [CI] 1.05-4.21) in a large cohort of 25,619 industrial
14 workers. In addition to the SMR based on an external comparison population, relative risks
15 (RRs) were presented based on internal comparisons of workers in order to minimize potential
16 selection bias due to the well known healthy worker effect. Statistically significant exposure-
17 response relationships between increased formaldehyde exposure and increased risk of
18 nasopharyngeal cancer were reported for two different metrics of exposure (peak and cumulative
19 exposure). Relative risks for nasopharyngeal cancer were also elevated for increased duration of
20 exposure to formaldehyde and for the average intensity of exposure. These analyses controlled
21 for potential confounders including calendar year, age, sex, race, and pay category. While
22 exposure measurement error is likely to be present in any epidemiologic study, there was no
23 evidence of any differential measurement error that could have produced the observation of a
24 spurious association. Any nondifferential measurement error would likely have attenuated the
25 effect of formaldehyde was smaller than that which would otherwise have been observed in the
26 absence of measurement error.
27 The case-control studies similarly also report associations between formaldehyde
28 exposure and cancer mortality for nasopharyngeal cancer. Although other risk factors for
29 nasopharyngeal cancer (e.g., Epstein-Barr Virus) and the predominant nasopharyngeal cancer
30 histological subtype (SCC versus undifferentiated) vary significantly across the world, case-
31 control studies consistently provide evidence of an association between occupational exposure to
32 formaldehyde and nasopharyngeal cancer (Vaughn et al., 1986a; Vaughn et al., 2000; Roush et
33 al., 1987; Hildesheim et al., 2001; West et al., 1993). In their more recent study, Vaughn et al.
34 (2000) used worker histories to estimate each individual worker's formaldehyde exposure.
35 Workers with more than 1.10 ppm-years of cumulative exposure were found to be at
This document is a draft for review purposes only and does not constitute Agency policy.
4-487 DRAFT—DO NOT CITE OR QUOTE
-------
1 significantly higher risk of nasopharyngeal cancer, with an odds ratio (OR) of 3.0 (95% CI: 1.3-
2 6.6) (Vaughn et al., 2000). Two different exposure metrics, duration of exposure and cumulative
3 exposure, were positively associated with increased risk of nasopharyngeal cancer, with a
4 significant test for trend (p = 0.014 and 0.033, respectively). The OR also increased in
5 magnitude as the probability of "Ever" having occupational exposure increased, from OR =1.6
6 among those whose exposure was judged to be "Possible, probable or definite" to OR =13.3
7 among those with "Definite" exposure (p-trend < 0.001).
8 Nasopharyngeal cancer histological subtype analysis indicates that these associations
9 held for both SCC and epithelial nasopharyngeal cancer, but not for the undifferentiating and
10 nonkeritinizing nasopharyngeal cancer (Vaughn et al., 2000). However, formaldehyde exposure
11 is also associated with risk of nasopharyngeal cancer in Taipei, Taiwan, where greater than 90%
12 of the cases had nonkeritinizing and undifferentiated carcinomas and less than 10% of the cases
13 were diagnosed as having SCCs (Hildesheim et al., 2001). These reported associations were
14 strengthened by considering higher probability of exposure (RR = 2.6; 95% CI: 1.1-6.3), greater
15 intensity of exposure (RR = 2.1; 95% CI: 1-4.2) and EBV seropositive cases (RR = 2.7; 95% CI:
16 1.2-5.9) (Hildesheim et al., 2001). Case-control studies have also linked residential exposure to
17 formaldehyde, specifically for years of residence in mobile homes (Vaughn et al., 1986b) and the
18 use of mosquito coils in the Philippines (West et al., 1993). Independent testing of 6 brands of
19 East Asian mosquito coils evaluated the emission rates of carbonyl compounds in the mosquito
20 smoke and reported that formaldehyde and acetaldehyde had the highest emission rates. Among
21 the three experiments on each of the six brands, the range of formaldehyde concentrations was
22 from 0.87 |ig/m3 (1 ppb) to 25 |ig/m3 (31 ppb).
23 As a group, other URT sites of direct contact with formaldehyde upon inhalation (i.e.,
24 salivary gland, mouth, nasal cavity and larynx) also showed evidence of a trend in increasing
25 relative risks with increasing average intensity and peak exposure in the Hauptmann et al. (2004)
26 cohort study, although these trends did not reach the level of statistical significance. The results
27 from other cohort studies and case-control studies are mixed (between positive associations and
28 null findings) for associations between formaldehyde exposure and specific cancers of the URT
29 (IARC, 2006). For rare cancers, extremely large cohorts would be needed to have the statistical
30 power to detect an association for tumors defined by individual sites (e.g., mouth, salivary gland,
31 hypopharynx). Results vary in the smaller cohort studies, where a single case may result in an
32 elevated risk but taken together the evidence is considered suggestive (see Section 4.1.2). Case-
33 control studies have been useful to better understand potential associations between
34 formaldehyde exposure and rare cancers of the URT. Luce et al. (2002) evaluated pooled data
35 from 12 case-control studies and demonstrated a statistically significant increased risk between
This document is a draft for review purposes only and does not constitute Agency policy.
4-488 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde exposure and sinonasal cancer. A case-control study by Gustavsson et al. (1998)
2 suggested an association between formaldehyde exposure and oral squamous cell carcinoma
3 (SCC), esophageal, and laryngeal cancers, with odds ratios (ORs) of 1.28, 1.90, and 1.45,
4 respectively. However, the individual ORs were not statistically significant. Hypopharyngeal
5 cancer was linked with formaldehyde exposure with an OR of 3.78 (95% CI: 1.50-9.49) in
6 another case-control study (Laforest et al., 2000). While the data on site-specific cancers of the
7 URT is somewhat sparse, they are consistent with a carcinogenic hypothesis and in their large
8 cohort study, Hauptmann and colleagues (2004) concluded that in spite of the small numbers of
9 deaths from cancers of the URT, the positive associations with average intensity and peak
10 exposure were consistent with the carcinogenicity of formaldehyde at these sites of first contact.
11
12 4.5.1.1. Supporting Animal Evidence
13 Animal studies, primarily rodent bioassays, strongly support the causal relationship
14 between of formaldehyde exposure and URT carcinogenicity. Formaldehyde-induced cancers
15 are primarily seen in the nasal passages (Kerns et al., 1983; Monticello et al., 1996; Tobe at al.,
16 1985; Kamata et al., 1997; Sellakumar et al., 1985), but it should be noted that rodents, unlike
17 humans, are obligate nose breathers and have convoluted nasal turbinates. Chronic animal
18 studies report tumor incidence in a variety of rodent models. Study descriptions are provided
19 above in detail (see Section 4.2.2, Table 4-42).
20 In rodent studies of the respiratory tract, only nasal tumors are considered to be induced
21 by formaldehyde. Repeated exposures to 10-15 ppm formaldehyde result in gross nasal lesions
22 and high incidence of nasal tumors (see Table 4-42, Section 4.2.1). Although increased cell
23 proliferation, squamous metaplasia, dysplasia, and focal necrotic lesions have been noted in the
24 larynx and trachea in some studies, no tumors in these locations have been reported in the rodent
25 studies. The majority of studies were conducted using rats (F344, Wistar, or Sprague-Dawley),
26 and all studies of 18 months or greater in mice and rats show evidence of formaldehyde-induced
27 nasal carcinogenicity. The nasal tumors are primarily SCCs, although papillomas, polypoid
28 adenomas, adenocarcinomas, fibrosarcomas, and esthesioneuroepitheliomas have been reported
29 (Kamata et al., 1997; Monticello et al., 1996; Morgan et al., 1986a, b; Takahashi et al., 1986;
30 Sellakumar et al., 1985; Kerns et al., 1983; Albert et al., 1982). Although hyperplasia, dysplasia,
31 and squamous metaplasia of the respiratory epithelium have been observed beyond the nasal
32 cavity, other respiratory tract tumors have not been reported to be significantly increased by
33 formaldehyde exposure alone.
34 Increased tumor incidence and decreased latency are correlated with increasing
35 formaldehyde exposure concentration. Reviewing data from the only lifelong inhalation study
This document is a draft for review purposes only and does not constitute Agency policy.
4-489 DRAFT—DO NOT CITE OR QUOTE
-------
1 (i.e., until "natural death") with multiple exposure groups, nasal SCCs occurred much earlier in
2 the high-exposure animals. For example, tumors are first noted at 8 and 9 months following
3 exposure for high-exposed (15 ppm) female and male F344 rats versus tumors not arising until
4 24 months in low-exposed rats (2 ppm) (Kerns et al., 1983). In a follow-up study by Monticello
5 et al. (1996), the incidence of SCC in rats exposed at 15 ppm was 47%, with the first tumor noted
6 at 12 months. The incidence of SCC in male rats exposed at 10 ppm was 22%, with the first
7 tumors observed at 18 months after exposure. Moreover, of the 90 rats exposed at 6 ppm for 20
8 months, only one SCC was noted. No SCCs were detected in rats exposed to 0.7 or 2 ppm
9 formaldehyde. These incidence rates are not mortality-adjusted (see Chapter 5, Section 5.3.4)
10 and include animals from each scheduled sacrifice (3, 6, 12, and 18 months). In a lifelong study
11 of male Sprague-Dawley rats exposed to 10 ppm formaldehyde, the cumulative nasal tumor
12 incidence was calculated as a function of time of exposure (Sellakumar et al., 1985). After 2
13 years of exposure, the adjusted probability of nasal carcinoma was greater than 60%.
14 There is some evidence that less-than-lifetime exposure to formaldehyde can induce nasal
15 tumors over an extended observation period. Two studies, both in male Wistar rats, report nasal
16 tumors in response to less-than-lifetime exposures (Woutersen et al., 1989; Feron et al., 1988).
17 A 13-week exposure at 20 ppm followed by an observation period of 30 months (inclusive of
18 exposure) in Wistar rats resulted in six nasal tumors including three nasal SCCs, one cystic SCC
19 of the nasolacrimal duct, one carcinoma in situ and an ameloblastoma, while no tumors were
20 noted in the corresponding air-exposed controls (Feron et al., 1988). A limited number of
21 formaldehyde-related tumors were noted from 4 or 8 weeks of exposure followed by 30 months
22 of observation. Although the tumor incidence of these less-than-lifetime exposures is low, this is
23 consistent with the 2-year bioassays in Wistar rats. Wistar rats are more resilient to
24 formaldehyde-induced nasal toxicity than F344 or SD rats (see Section 4.2.1), and only 1 of 26
25 (4%) Wistar rats exposed at 10 ppm for 28 months developed SCC (Woutersen et al., 1989)
26 versus 22% in F344 rats (Monticello et al., 1996).
27 The specificity of formaldehyde-induced tumors in the nasal passages of rodents is
28 believed, at least in part, to be a function of tissue dose. Computational fluid dynamics (CFD)
29 modeling used to predict formaldehyde tissue flux during inhalation exposures suggests that at
30 comparable concentrations, tissue flux in the nasal passages of rodents is more intense than for
31 nonhuman primates and humans. Modeling predicts a different pattern of formaldehyde flux into
32 URT tissues of rodents compared to humans, where formaldehyde penetrates more deeply into
33 the respiratory tract of primates than rodents even considering nose-only breathing for primates
34 (see Section 3.4). Humans will generally switch to mouth breathing when sensing an irritating
35 smell and during physical exertion, resulting in direct exposures to the mouth and greater tissue
This document is a draft for review purposes only and does not constitute Agency policy.
4-490 DRAFT—DO NOT CITE OR QUOTE
-------
1 flux in tissues beyond the bypassed nasal passages. Therefore, species differences in tissue dose
2 may contribute to formaldehyde-induced tumors in humans beyond the nasal passages, which are
3 not evident in rodent bioassays.
4
5 4.5.2. Lymphohematopoietic Malignancies
6 4.5.2.1. Background
1 Lymphohematopoietic (LHP) cancers include neoplasms of both lymphoid and myeloid
8 cell origins. Cancers of the immune system are described as leukemia if they primarily involve
9 cells from peripheral blood and bone marrow at diagnosis and lymphomas if they constitute a
10 solid tumor (Robbins, 2004). Some forms of leukemia which present as an immature immune
11 cell phenotype are believed to arise from lymphomyeloid stem cells or progenitor cells normally
12 found in the bone marrow (e.g., acute lymphoblastic leukemia (ALL) and acute myeloid
13 leukemia (AML)) (Greaves, 2004). However, multiple myeloma, lymphomas and some
14 leukemias may arise from mature functional lymphocytes present outside of the bone marrow
15 (Greaves, 2004; see Figure 4-32).
16
Developmental origins of blood cell cancers
Germ Line or
Early Embryonic
Haemopoietic
Stem cell
(lymphomyeloid)
Rare Acute Leukemia
Acute Lymphoblastic Leukemia (ALL)
Acute Myeloid Leukemia (AML)
Myelodysplastic syndrome
Chronic Myeloid Leukemia (CML)
17
18
19
20
21
22
Myeloid
Stem Cell
Lymphoid
Stem Cell
Mature
lymphocyte
Subsets
Source: M.F. Greaves, IARC Press, (2004)
Childhood ALL and AML
Cutaneous T Lymphoma
Adult T-cell Leukemia
B Chronic Lymphocytic Leukemia
non-Hodgkin's Lymphoma
B or T Prolymphocytic Leukemia
Hairy Cell Leukemia
Myeloma
Figure 4-32. Developmental origins for cancers of the lymphohematopoietic
system (Adapted from Greaves (2004).
This document is a draft for review purposes only and does not constitute Agency policy.
4-491 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.5.2.2. All LHP Malignancies
2 Epidemiologic studies involving formaldehyde-exposed workers provide sufficient
3 evidence of a causal association between formaldehyde exposure and all LHP malignancies (see
4 Section 4.1.2.2.1, summarized in Section 4.1.2.2.1.4). Positive associations between
5 formaldehyde exposure and LHP cancers have been reported for chemical workers (Wong et al.,
6 1983; Bertazzi et al., 1986), embalmers (Walrath and Fraumeni, 1983, Walrath and Fraumeni,
7 1984; Hayes et al., 1990), anatomists and pathologists (Harrington and Shannon 1975; Hall et al.,
8 1991; Levine at al 1984; Stroup et al., 1986; Matanoski et al., 1989) (see Table 4-90). However,
9 clear associations (in terms of overall standardized mortality ratios [SMRs] or proportional
10 mortality ratios [PMRs]) were not reported in analyses for garment workers, iron-foundry
11 workers, and a large U.S. industrial cohort (Pinkerton et al., 2004; Andjelkovich et al., 1995;
12 Beane-Freeman et al., 2009; Marsh et al., 1996), although associations were observed in some of
13 these studies when exposure-response relationships were considered. Several published meta-
14 analyses are available which more formally assess the strength of association between
15 formaldehyde exposure and mortality from all LHP cancers. Pooled SMRs indicate stronger
16 associations for professional workers (embalmers, anatomists, and pathologists) than industry
17 workers (see Table 4-91). Bosetti et al. (2008) found similar relationships, with a pooled SMR
18 of 1.31 (95% CI: 1.16-1.47) for 'professionals' (i.e., embalmers, anatomists and pathologists)
19 versus a pooled estimate of 0.85 (95% CI: 0.74-0.96) for industrial workers. A recent meta-
20 analysis by Zhang et al. (2009) reports a summary relative risk (RR) of 1.25 (95% CI: 1.09-1.43)
21 for both professional and industry workers for all LHP cancers (ICD 9 codes 200-209). These
22 researchers identified 19 cohort study analyses, including cohort study updates. Zhang et al.
23 (2009) used the reported RR from the highest exposure category to increase statistical power and
24 reduce uncertainty regarding confounding or other bias. The criteria for study inclusion and
25 exclusion applied by Zhang et al. (2009) appear to be appropriate and the methodology for using
26 myeloid-specific results where possible also appears to be appropriate. This meta-analysis is
27 supportive of a causal association between formaldehyde and LHP malignancies.
28 The apparent differences by industry/profession may reflect many influences, including
29 exposure potential and demographic characteristics. External analysis (use of the general
30 population for comparison) relies on the assumption that cancer incidence rates are expected to
31 be similar between the general population and the study population in the absence of exposure.
32 The 'healthy worker effect' is well known, and there may be differences in the magnitude of this
33 selection bias by industry or profession. For instance, LHP cancer incidence and mortality have
34 many risk factors including socioeconomic status. Therefore, the consistent positive findings in
35 professional workers versus mixed results in industrial workers could be influenced by the
36
This document is a draft for review purposes only and does not constitute Agency policy.
4-492 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
Table 4-90. Summary of cohort and case-control studies which evaluated the
incidence of all LHP cancers in formaldehyde-exposed populations (ICD-8
Codes: 200-209) and all leukemias (ICD-8 Codes: 204-207). (See Table 4-9 for
complete study details and findings)
Study population
Study details
All LHP cancers
Leukemia
Reference
SMR Analysis1
Pathologists and technicians
(n = 2,079)
Pathologists and technicians
(n = 2,720)
Pathologists
(« = 4,512)
Ontario Undertakers
(n = 1,477)
Male Anatomists
(n = 2,327)
Male pathologists
(n = 4,485)
Male pathologists
(w = 6,lll)
Chemical industry workers,
men
(«= 14,014)
Chemical workers
(n = 2,026)
Industrial workers
(« = 25,619)
Industrial workers
(n = 7,328)
Garment workers
(«= 11,098)
Years of study 1955-1973
1974-1980
Years of study 1974-1987
Mortality from 1950-1977
Mortality from 1925-1979
Mortality through 1977
Participants from 1912-
1950 membership rolls.
Mortality followed
through 1978.
Mortality from 1941-2000
Mortality followed
through 2004
Mortality followed
through 1998
2.0 (p< 0.01)
{pathologists}
0.5
{technicians}
NR
1.44
(0.69-2.63)
1.24
1.20
(0.7-2.0)
NR
1.25
(0.95-1.62)
NR
1.36
(0.5-2.95)
0.94
(0.84-1.06)
0.89
0.97
(0.74-1.26)
0.6
{pathologists}
0.5
{technicians}
0.91 (0.05-4.29)
men
9.26(0.47-43.9)
women
1.52 (0.41-3.89)
1.60
1.5 (0.7-2.7)
1.06
1.35 (0.92-1.92)
0.91(0.62-1.39)
1.02
(0.85-1.59)
1.09 (0.70-1.62)
Harrington and
Shannon, 1975
Harrington and
Oakes, 1984
Hall etal., 1991
Levineetal.,1984
Stroupetal., 1986
Logueetal., 1986
Matanoskietal.,
1989
Coggonetal.,
2003
Wong etal., 1983
Beane-Freemen et
al., 2009
Marsh etal., 1996
{Subset of NCI
cohort reported by
Hauptmann et al.,
2003}
Pinkerton et al.,
2004
This document is a draft for review purposes only and does not constitute Agency policy.
4-493 DRAFT—DO NOT CITE OR QUOTE
-------
Table 4-90. Summary of cohort and case-control studies which evaluated the
incidence of all LHP cancers in formaldehyde-exposed populations (ICD-8
Codes: 200-209) and all leukemias (ICD-8 Codes: 204-207). (See Table 4-9 for
complete study details and findings) (continued)
Study population
Study
details All LHP cancers
Leukemia
Reference
SMR Analysis"
Resin plant workers
(n = 1,330)
Plastic manufacturing
(n = 5,932)
Embalmers, New York
(n= 1,132)
Embalmers, CA
(n = 1,007)
Embalmers, U.S.
(n = 4,046)
Employed between
1959-1980
Mortality through 1986
Licensed between
1925-1980
Licensed between
1925-1980
2.01
1.69
(salaried workers)
0.93
(hourly workers)
1.15
1.22
1.39
(1.15-1.63)
White 1.31
(1.06-1.59)
Nonwhite 2.41
(1.35-3.97)
NR
1.98
(salaried workers)
0.98
(hourly workers)
1.32
1.75 (p< 0.05)
1.24
(<20 years)
2.21 (p< 0.05)
(>20 years)
1.52b
(0.98-2.35)
White 1.44
(p<0.05)
Nonwhite 2.72
(p<0.05)
Bertazzietal.,
1986
Dell and Teta,
1995
Walrath and
Fraumeni, 1983
Walrath and
Fraumeni, 1984
Hayes etal., 1990
Case-Control Studies"
American cancer Society
Cancer Prevention Study II:
(n = 362,828 men)
White men diagnosed with
leukemia
(Iowa and Minnesota)
(n = 622)
Results for men reporting
formaldehyde exposure,
and occupations related to
formaldehyde exposure
Recruited in 1980-1983
1.22 (0.84-1.77)
(formaldehyde
exposed)
3.44(1.11-10.68)
{formaldehyde
exposure and
occupation}
NR
0.96 (0.54-1.71)
(formaldehyde
exposed)
5.79 (1.44-23.25)
{formaldehyde
exposure and
occupation}
1.0 (0.7-1.4) Low
0.7 (0.2-2.6) High
Stellmanetal.,
1998
Blair etal., 1993
4
5
aRelative risk estimate (SMR or OR) presented with 95% confidence intervals, where available.
bSMR for leukemia for the total group calculated from the published data for lymphatic leukemia (204, myeloid leukemia_(205X
and other/unspecified_(206,207).
This document is a draft for review purposes only and does not constitute Agency policy.
4.494 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
Table 4-91. Secondary analysis of published mortality statistics to explore
alternative disease groupings within the broad category of all
lymphohematopoietic malignancies
All lymphohematopoietic
malignancies
ICD-8 Codes
200-209
U.S. embalmers
(Whole cohort)
(Hayes etal., 1990)
SMR
(95% CI)
1.39a
(1.15-1.67)
U.S. industry
(peak exposure metric: >4 ppm
vs. >0 to <2 ppm)
(Beane-Freeman et al., 2009)
Relative risk
(95% CI)
1.37
(1.03-1.81)c
Alternative Disease Groupings
Exclude myeloid leukemia
Solid tumors of lymphoid origin
(Lymphosarcoma and
reticulosarcoma, Hodgkin's
lymphoma, non-Hodgkin's
lymphoma and multiple
myeloma)
200-204,
206-209
200-203
1.35a
(1. 13-1.72) b
1.24a
(0.94-1.61) b
1.31d
(0.97-1.75) 4e
1.33d
(0.93-1.90) d-e
5
6
10
11
12
13
14
15
16
17
18
19
20
SMR =
Obs •
Exp
b Fischer's exact confidence intervals.
c See Table 2 of Beane-Freeman et al. (2009).
^ = DeathsComparisonGroup/Person-TimeComparisonGrollp
Deathsreflrea Group /Person - Timereferent Group
108-19
103-14
FT,
referent Group
PT
Comparison Group
where
PT,
zferent Group
PTr,
Comparison Group
Var(logRR) =
108C
103" x 1.37"
1
= 0.765
1
L>eatnsComparison Group
Deaths
referent Group
appropriateness of the comparison to the general population - that is, a differential extent of
selection bias. Interestingly, salaried workers, but not the hourly workers, in an Italian plastic
manufacturing plant had elevated SMRs for LHP cancers (1.69 (95% CI: 1.07-2.53) and
0.93(95% CI: 0.62-1.35), respectively) (Dell and Teta, 1995). Without knowledge of which
worker group is most similar to the comparison population with respect to LHP cancers
mortality, one cannot discern if this potential effect of demographic variability accentuates
effects in professional/salaried workers or obscures the effects in industrial/hourly-wage
workers.
This document is a draft for review purposes only and does not constitute Agency policy.
4-495 DRAFT—DO NOT CITE OR QUOTE
-------
1 The only study which has data to inform the effects of either exposure level or the
2 appropriateness of an external comparison group on the association between formaldehyde
3 exposure and all LHP cancer mortality is the National Cancer Institute (NCI) cohort study of
4 industrial workers (Blair et al., 1986; Beane-Freeman et al., 2009), which presents relative rates
5 based on internal comparisons for 3 different exposure metrics. Although SMR analysis with an
6 external comparison group did not indicate increased mortality from all LHP cancers (0.94, CI:
7 0.84-1.06, for the exposed workers), internal analysis using the low-exposed workers as the
8 comparison group demonstrates positive exposure-response relationships for increased mortality
9 from all LHP malignancies cancers and peak exposure across the study periods (1965-2004) (see
10 Figure 4-7 and 4-8) (Beane-Freeman et al., 2009), with a statistically significant trend (p < 0.05)
11 for every year since 1977. These results, indicating a positive exposure-response relationship
12 among plant workers, who most likely have similar demographic characteristics, are noteworthy
13 given the apparent lack of association when SMRs for the same cohort are calculated against
14 mortality rates for the general population. The lack of an apparent association with SMRs may
15 be attributable to the healthy worker effect and/or some other difference between the exposed
16 workers and the general population.
17 Although the association between formaldehyde exposure and all LHP cancer mortality
18 in industrial and professional cohorts is mixed, the strength of the internal analysis of the NCI
19 cohort, in the absence of positive SMRs compared to the general population, suggests that SMR
20 analyses may not be the most appropriate methodology for assessing LHP cancer mortality.
21 Given the potential for demographic differences between an industrial workforce and the general
22 population, the results of the internal analysis of the NCI industrial cohort provide a higher
23 quality analysis—and therefore should be given significantly more weight than SMR analyses of
24 industrial workers that could not distinguish their findings from the null. Given the consistency
25 and strength of the positive associations for all LHP cancers cancer mortality in professional
26 cohorts (embalmers, anatomists and pathologists) taken together with the strong positive results
27 of the NCI cohort, human epidemiologic evidence are sufficient to conclude that there is a causal
28 association between formaldehyde exposure and mortality from all LHP malignancies (as a
29 group).
30
31 4.5.2.3. All Leukemia
32 Epidemiologic studies involving formaldehyde-exposed workers provide sufficient
33 evidence of a causal association between formaldehyde exposure and all leukemia as a group
34 (see Section 4.1.2.2.1, summarized in Section 4.1.2.2.1.4). Like the analysis of all LHP cancers,
35 an analysis of all leukemia combines diseases which differ significantly in cell of origin and
This document is a draft for review purposes only and does not constitute Agency policy.
4-496 DRAFT—DO NOT CITE OR QUOTE
-------
1 etiology, including acute and chronic forms of both myeloid and lymphatic leukemia. This class
2 also includes other leukemia (e.g., erythraemia) and a general category of 'other and unspecified
3 leukemia' (ICD-8 207). Regardless, there is some utility in evaluating the all leukemia mortality
4 data because many studies provided results for this grouping. Also, the diagnosis of leukemia
5 versus solid LHP tumors is fairly distinct thereby limiting misclassification of the endpoint.
6 Although results are mixed across the studies (see Table 4-90), an association between
7 formaldehyde exposure and leukemia mortality is supported by cohort analyses of embalmers,
8 pathologists and anatomists (Hayes et al., 1990; Walrath and Fraumeni, 1983; Walrath and
9 Fraumeni 1984; Hall et al., 1991; Levine et al., 1984; Stroup et al., 1986; Matanoski et al., 1989).
10 Formaldehyde exposure and formaldehyde-related occupation are associated with leukemia
11 diagnosis in a case-control study (RR = 5.79 (95% CI: 1.44-23.25), but not formaldehyde
12 exposure alone (RR = 0.96; 95% CI: 0.54-1.71) (Stellman et al., 1998) (see Section 4.1.2.2.1 for
13 complete study summaries).
14 In contrast, SMR analyses of the industrial cohorts do not indicate a similar association
15 (Coggon et al., 2000; Beane-Freeman et al., 2009, Pinkerton et al., 2004). Although the SMR
16 analysis provided for the NCI cohort does not indicate a positive association for all leukemia
17 using an external reference group (Beane-Freeman et al., 2009), the SMR for exposed versus
18 unexposed workers within the cohort suggests all leukemia is elevated 2.1-fold with this internal
19 comparison (95% CI: 0.99-4.56)6. A positive exposure-response relationship further strengthens
20 the association of formaldehyde exposure to leukemia mortality (Beane-Freeman et al., 2009).
21 Where the referent group is defined as 'low exposed' individuals, leukemia is elevated in the
22 highest peak exposure category (RR = 1.42; 95% CI: 0.92-2.18) compared to both the referent
23 group and the unexposed category (RR = 0.59; 95% CI: 0.25-1.36), and there is a statistically
24 significant trend across all groups (p = 0.02). Categorical analysis for the average intensity and
25 cumulative exposure metrics suggests greater mortality in the high-exposure groups versus the
26 'low exposed' individuals (RR= 1.10 [95% CI: 0.68-1.78] and 1.11 [0.7-1.74]. respectively), but
27 analysis of individual results across the exposure-response range indicates cumulative exposure
28 is a better predictor (p = 0.08 for trend across all exposed and unexposed.)
29 Several meta-analyses have been conducted for formaldehyde exposure and leukemia
30 which indicate a positive association (see Section 4.1.2.2.1.3). Collins et al. (2004) report an
31 overall RR for 18 available studies of 1.1 (CI: 1.0-1.2), suggesting an association of leukemia
32 with formaldehyde exposure. This association was stronger for both pathologists/anatomists
1
6Var
C
In
V
( SMRExposed ^
^ bMKNon exp osed ^
\ i
j UDSExposed
n exp osed
This document is a draft for review purposes only and does not constitute Agency policy.
4-497 DRAFT—DO NOT CITE OR QUOTE
-------
1 (1.4; CI: 1.0-1.9) and embalmers (RR = 1.6; 1.2-2.0), but there was no association when
2 considering only industrial workers (RR = 0.9; 0.8-1.0). Study design also impacted the
3 apparent strength of association, with stronger associations seen in case-control studies
4 (RR = 2.4; 0.9-6.5) versus cohort studies (RR = 1.0; 0.9-1.2). Bosetti et al. (2008) reported an
5 association between formaldehyde exposure and leukemia mortality with a pooled RR of 1.39
6 (95% CI: 1.15-1.68) for 8 groups of professional workers. In the same analysis, the pooled RR
7 for the 4 industrial cohorts was 0.90 (0.75-1.07). Zhang et al. (2009) reported a pooled RR of
8 1.54 (95% CI: 1.18-2.00) for all cohorts identified in their meta-analysis, although this pooled
9 RR should be considered with some caution, as myeloid leukemia alone was included in the
10 analysis where available (Zhang et al., 2009).
11 While the epidemiologic evidence for a causal association between formaldehyde and all
12 leukemia as a group is not at strong as for all LHP as a group, the repeated identification of an
13 association in multiple meta-analyses taken together with the clear causal association between
14 myeloid leukemia demonstrated by Hauptmann et al. (2009) and the consistent evidence reported
15 by Beane-Freeman et al. (2009) are sufficient to conclude that there is a causal association
16 between formaldehyde exposure and mortality from all leukemia as a group (see
17 Section 4.1.2.2.1.4).
18
19 4.5.2.4. Subtype Analysis
20 Given the associations discussed above between formaldehyde exposure and both all
21 LHP cancers and all leukemia, further analysis is needed to examine if the observed increase in
22 all LHP cancers is primarily a reflection of increased leukemia, or if other types of LHP cancers
23 may be elevated as well. Although analysis of mortality data by subtype may provide a better
24 understanding of the specific disease associations, there are potential pitfalls as well. Chief
25 among these concerns are the potential for disease misclassification (especially in studies with
26 older mortality data) and lack of statistical power as the number of observed cases is reduced by
27 considering subtypes. Case control studies by design address specific diseases and are well-
28 suited for subtype analysis, but often provide little exposure information. The following analysis
29 will draw from the available data to examine which forms of LHP malignancies may be
30 associated with formaldehyde exposure.
31 There has been speculation that the association between formaldehyde exposure and
32 increases in all LHP cancers and all leukemia are driven by increased myeloid leukemia (Pyatt et
33 al., 2008; Heck and Casanova, 2004; Golden et al., 2006). If this were the case, then mortality
34 from LHP cancers other than myeloid should not be elevated, once the excess mortality from
35 myeloid leukemia is accounted for. Only 2 studies provide the data to evaluate this hypothesis—
This document is a draft for review purposes only and does not constitute Agency policy.
4-498 DRAFT—DO NOT CITE OR QUOTE
-------
1 both conducted by the NCI (Hayes et al., 1990; Beane-Freeman et al., 2009). From the
2 published data, crude mortality statistics can be calculated for alternative disease groupings (see
3 Table 4-91). In the NCI embalmer study (Hayes et al., 1990), only myeloid leukemia was
4 statistically elevated in the subtype analysis. For the NCI industrial cohort (Beane-Freeman et
5 al., 2009), elevations were also seen for Hodgkin's lymphoma relative to the referent group. In
6 both cases, the association between formaldehyde exposure and LHP malignancies remains when
7 myeloid leukemia is dropped from the analysis. Further, similar associations are found when all
8 leukemia and myeloproliferative diseases are dropped from the analysis and only solid tumors of
9 lymphoid origin are included (lymphosarcoma and reticulosarcoma, Hodgkin's lymphoma, non-
10 Hodgkin's lymphoma and multiple myeloma). These reanalyses illustrate the need for a more
11 careful subtype analysis to assess the potential for associations between formaldehyde exposure
12 and various forms of LHP cancers.
13
14 4.5.2.5. Myeloid Leukemia
15 The associations between myeloid leukemia and formaldehyde exposure are positive and
16 consistent (see Table 4-92, Section 4.1.2.2.4 for summary and evaluation of studies). Of the six
17 studies which formally assess myeloid leukemia mortality by SMR analysis, five are positive,
18 including cohorts of both professional and industrial workers although only two reach statistical
19 significance (Pinkerton at al., 2003; Stroup et al., 1986; Hayes et al., 1990; Walrath and
20 Fraumeni, 1984; Walrath and Fraumeni, 1983). Myeloid leukemia mortality was not increased
21 in the NCI industrial worker cohort by SMR analysis, but, there is evidence in the internal
22 analyses that myeloid leukemia is elevated by peak exposure (1.78 (0.97-3.64), and when
23 exposed are compared to unexposed within the cohort (1.38 (0.65-2.97). The strongest
24 association is seen in the recent update of US embalmers (11.2 (1.3-95.6 for ever exposed).
25 Hauptmann et al. (2009) provide several alternative exposure metrics which all show positive
26 associations (number of embalmings, average formaldehyde intensity, cumulative formaldehyde
27 exposure) (see Section 4.1.2.2.1 for study details).
28 Further subtype analysis to distinguish between acute and chronic myeloid leukemia is
29 problematic. Although Walrath and Fraumeni (1983, 1984) note that AML is prominent in their
30 analyses of New York and California licensed embalmers; they do not provide PMR (actually
31 SMR) analyses for CML. Walrath and Fraumeni (1983 and 1984) report leukemia cell types—
32 for both studies the majority of myeloid leukemia are acute (5/6 and 4/6, respectively, for New
33 York State and California embalmers). However, SMRs cannot be calculated for AML versus
34 CML in this paper, as comparison rates are not available from the 1920's through the 1960's—
35 the timeframe with the majority of deaths. The authors do contrast the observed rate of AML in
36 the cohort to the background rate for AML in white men in the 1970s—but given the potential
This document is a draft for review purposes only and does not constitute Agency policy.
4-499 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 4-92. Summary of studies which provide mortality statistics for
myeloid leukemia subtypes
Study population
Myeloid Leukemia
Acute
Myeloid
Leukemia
Chronic
Myeloid
Leukemia
Reference
SMR Analysis"
Garment workers
(«= 11,098)
Anatomists
(« = 2,317)
Industrial workers
(» = 25,619)
Embalmers, U.S.
(jV= 4,046)
Embalmers, U.S.
(N =4,046)
Embalmers (NY)
(n =
Embalmers (CA)
(» =
1.44 (0.08-2.37)
NR
0.90 (0.67-1.21)
SMR Ratio
1.38 (0.65-2.97)
(exposed/unexposed)
11.2
(1.3-95.6)
1.57
(1.01-2.34)
1.46
(0.54-3.19)
1.50
(0.55-3.26
1.34
(0.61-2.54)
NR
NR
NR
1.52
(0.85-2.52)
NR
NR
1.39
(0.38-3.56)
8.8b
NR
NR
1.84
(0.79-3.62)
NR
NR
Pinkerton at al., 2003
Stroupetal., 1986
Beane-Freeman et al.,
2009
Hauptmann et al., 2009
Hayes etal., 1990
Walrath and
Fraumeni,1983,
Walrath and Fraumeni,
1984
Case-Control Studies"
White men diagnosed with
leukemia
(Iowa and Minnesota) (n =
622)
NR
Low:
0.9(0.5-1.6)
High:
NR
Low:
1.3 (0.6-3.1)
High:
2.9 (0.3-24.5)
Blair etal., 1993
4
5
6
7
8
9
10
11
12
13
14
15
16
17
"Relative risk estimate (SMR, or OR) presented with 95% confidence intervals, where available.
Yeukemia SMR 1.5 (0.7-2.7) {5 of 10 deaths due to myeloid}; Chronic Myeloid Leukemia (CML) SMR of 8.8.
misclassification of late stage CML as AML, especially historically, this may not be an
appropriate comparison. Therefore, although these studies support an association between
formaldehyde exposure and myeloid leukemia in general (Walrath and Fraumeni, 1983; 1984),
the reported AML and CML subtype information does not allow a satisfactory subtype analysis
for myeloid leukemia.
Several studies do present a formal subtype analysis within myeloid leukemia. Similar
SMRs are reported for AML 1.34 (0.61-2.54) and CML (1.39(0.38-3.56)) in garment workers
(Pinkerton at al., 2003). The initial NCI study of US embalmers indicates a slightly stronger
association of formaldehyde exposure to CML (1.84 (0.79-3.62) compared to AML (1.52
This document is a draft for review purposes only and does not constitute Agency policy.
4-500 DRAPT—DO NOT CITE OR QUOTE
-------
1 (0.85-2.52) (Hayes et al., 1990.) Stroup et al. (1986) report that all of the myeloid
2 leukemiaobserved in anatomists was CML with an SMR of 8.8. Finally, a population-based
3 case-control study of white men in Iowa and Minnesota provides positive SMRs for CML based
4 on estimated formaldehyde exposure (Low: 1.3 (0.6-3.1); High: 2.9 (0.3-24.5)), but no
5 association for AML (Low exposed: 0.9 (0.5-1.6)) (Blair et al., 1993). Therefore, although few
6 cases exist for further subtype analysis, the available data indicate either no differences in SMRs
7 for acute myeloid leukemia (AML) versus chronic myeloid leukemia (CML) (Hayes et al., 1990;
8 Pinkerton et al., 2003) or suggest CML is more prominent (Blair et al., 2000; Stroup et al., 1986).
9 The meta-analysis by Zhang et al. (2009) evaluated the studies of formaldehyde exposure
10 and myeloid leukemia available at the time including Hauptmann et al. (2003), Pinkerton at al.,
11 2003; Hayes et al., 1990; Stroup et al., 1986; Walrath and Fraumeni, 1984, 1983. While the
12 findings of Hauptmann et al. (2003) on the NCI cohort have been recently updates by those of
13 Beane-Freeman et al. (2009) who updated the cohort, the Zhang et al. (2009) analysis provide the
14 only formal meta-analysis specific to myeloid leukemia. Zhang et al. (2009) reported a
15 statistically significant summary RR of 1.90 (95% CI: 1.31-2.76).
16 The study by Hauptmann et al. (2009) stands out among the studies of embalmers and
17 professionals in the funeral industry based on the strength of the quantitative exposure data and
18 the demonstration of exposure-response relationships which provide causal evidence of an
19 association between formaldehyde exposure and increased risk of myeloid leukemia. These
20 results were internally consistent and demonstrated statistically significant associations that were
21 unlikely the result of chance. As this nested case-control study was based on the cohorts of
22 Hayes et al. (1990) and those of Walrath and Fraumeni (1983, 1984), the potential for selection
23 bias is considered to be low. Further, the controls in Hauptmann et al. (2009) were carefully
24 selected to avoid individuals who died of any causes that were thought to even possibly be
25 related to formaldehyde exposure. Confounding is also unlikely to be an alternative explanation
26 for the observed results as there were clear and convincing exposure-responses and the
27 magnitude of the effect estimates were extremely large.
28 Given the consistency of the positive associations for formaldehyde with myeloid
29 leukemia cancer mortality across five of the six studies (Hauptmann et al., 2009; Pinkerton at al.,
30 2003; Hayes et al., 1990; Stroup et al., 1986; Walrath and Fraumeni, 1984, 1983; but not
31 Beane-Freeman et al., 2009), the statistically significant meta analysis by Zhang et al. (2009) and
32 the convincing results from Hauptmann et al. (2009), the human epidemiologic evidence is
33 sufficient to conclude that there is a causal association between formaldehyde exposure and
34 mortality from myeloid leukemia (also see Section 4.1.2.2.4 for further evaluation).
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-501 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.5.2.6. Solid Tumors ofLymphoid Origin
2 Multiple myeloma, Hodgkin's lymphoma, non-Hodgkin's lymphoma, lymphosarcoma,
3 reticulosarcoma, and other lymphomas may all be derived from immune cells outside of the bone
4 marrow compartment, in peripheral blood, in the gut and respiratory mucosa and immune tissues
5 at the POE (e.g., lymph nodes, mucosa-associated lymphoid tissue (MALT), gut-associated
6 lymphoid tissue (GALT) (Greaves, 2004). The only meta-analysis to specifically address
7 lymphoid malignancies found evidence for increased lymphoma (Hodgkin's lymphoma (pooled
8 RR= 1.23; 95% CI: 0.67-2.29) and multiple myeloma (1.31; 1.02-1.67), but not for non-
9 Hodgkin's lymphoma (1.08; 0.86-1.35) (Zhang et al., 2009). As seen in Table 4-93 below,
10 individual study results are mixed for these lymphoid cell-line malignancies, as they are for all
11 LHP cancers and all leukemia above. Although these tumors are from mature lymphocytes,
12 there is still variability in the etiology, natural history and risk factors for the many subtypes
13 which are included in these categories.
14
15 4.5.2.6.1. Hodgkin's lymphoma.
16 The only meta-analysis to specifically address Hodgkin's lymphoma was conducted by
17 Zhang et al. (2009) and included eight studies (Anjelkovich et al., 1995; Coggon et al., 2003;
18 Harrington and Shannon, 1975; Hauptmann et al., 2003; Hayes et al., 1990; Pinkerton et al.,
19 2004; Walrath and Fraumeni, 1983, 1984; and Wong, 1983). Zhang et al. (2009) reported a
20 summary RR = 1.23 (95% CI: 0.67-2.29). This elevated, but nonstatistically significant finding
21 is consistent with the large variance on reported results among the individual studies as well as
22 the wide confidence intervals of the results which were based on small numbers of cases—even
23 from the large cohort studies. Six of the eight studies observed three or fewer deaths from
24 Hodgkin's lymphoma. Coggon et al. (2003) reported 6 deaths from Hodgkin's lymphoma
25 against 8.5 expected for an SMT = 0.70 (95% CI: 0.26-1.53) and Hauptmann et al. (2003)
26 reported 21 observed deaths with 20 deaths among the exposed workers who has an SMR = 1.26
27 (95% CI: 0.81-1.95). However, the Beane-Freeman et al. (2009) update of the Hauptmann et al.
28 (2003) study had the largest number of observed cases (n = 27) and was not included in the
29 Zhang et al. (2009) meta-analysis. In fact, the Beane-Freeman et al. (2009) study describes more
30 deaths from Hodgkin's lymphoma than all the other studies in Zhang et al. (2009) combined.
31 Excluding the Hauptmann et al. (2003) results from the list of studies in the meta-analysis leaves
32 19 cases.
33
34
35
This document is a draft for review purposes only and does not constitute Agency policy.
4-502 DRAFT—DO NOT CITE OR QUOTE
-------
4
Table 4-93. Summary of mortality statistics for Hodgkin's lymphoma,
lymphoma and multiple myeloma from cohort analyses of formaldehyde
exposed workers
Study population
Hodgkin's
Lymphoma
Non-Hodgkin's
Lymphoma
Multiple
myeloma
Reference
SMR Analysis"
Pathologists (n = 2,079)
Pathologists (« = 4,5 12)
Male Anatomists (n = 2,327)
Male pathologists (n = 6,1 1 1)
Chemical workers (n = 2,026)
British Chemical plants (n = 14,014)
Swedish workers- abrasive
production plant (n = 91 1)
Industrial workers (n = 25,619)
Embalmers, New York (n = 1,132)
Embalmers, CA (n = 1,007)
Embalmers, U.S. (n = 4,046)
1.4
1.21 (0.03-6.71)
_ C
0.36(0.04-1.31)
2.94 (0.33-10.63)
0.36 (0.01-2.01)
NR
1.42(0.96-2.10)
0.87 (p< 0.05)
C
0.72(0.15-2.10)
2.0 (p< 0.05)
1.44 (0.69-2.63)
0.7 (.1-2.5) d
2.0 (0.7-4.4) e
1.31 (0.66-2.35) d
1.54 (0.82-2.63)e
NR
0.89(0.41-1.70)
2.0 (0.2-7.2)
0.86 (0.70-1.05)
1.08d
1.22e
3.10d
1.33e
1.26 (0.87-1.76)
1.12(0.58-1.96)d
1.35 (0.84-2.01)e
NRb
NR
NR
NR
NR
1.18(0.48-2.44)
4.0 (0.5-14)
0.94(0.71-1.25)
NR
NR
1.37(0.84-2.12
Harrington and
Shannon, 1975
Halletal., 1991
Stroupetal., 1986
Matanoski et al., 1989
Wongetal., 1983
Coggon et al., 2003
Edlingetal., 1987a
Beane-Freemen et al.,
2009
Walrath and
Fraumeni, 1983
Walrath and
Fraumeni, 1984
Hayes etal., 1990
Case-Control Studies"
Women in Connecticut (n = 601)
White men, Iowa and Minnesota (n =
622)
ACS Cancer Prevention Study II (n =
128)
Men, ACS Cancer Prevention Study
II (n = 45,399)
Danish workers (n = 1,098)
Danish women (607)
NR
NR
NR
NR
NR
NR
1.3 (1.0-1.7)
1.2 (0.9-1.7)
NR
0.92 (0.5-1.68)
2.88 (0.40-10.5/
NR
NR
NR
NR
1.8 (0.6-5.7)
0.74 (0.27-2.02)
1.6 (0.4-5.3)
Wang etal., 2009
Blair etal., 1993
Boffettaetal., 1999
Stellman et al., 1998
Heinemanetal., 1992
Potternetal., 1992
5
6
7
8
9
10
11
""Relative risk estimate (SMR or OR) presented with 95% confidence intervals, where available.
bNR is not reported.
°"—" no cases observed.
dLymphosarcoma and reticulosarcoma only.
e"other lymphoma."
formaldehyde exposure in a wood-related occupation. RR for wood-related occupation alone was not elevated 0.97
(0.55-1.73).
This document is a draft for review purposes only and does not constitute Agency policy.
4-503 DRAFT—DO NOT CITE OR QUOTE
-------
1 There is evidence for an exposure-response relationship for Hodgkin's lymphoma in the
2 NCI industrial cohort among exposed workers (Beane-Freeman et al., 2009). Clear exposure
3 response relationships for Hodgkin's lymphoma are defined with all three metrics of exposure,
4 peak average intensity and cumulative exposure (p = 0.01,/? = 0.05 andp = 0.08 respectively for
5 mortality through 2004). These associations have been evident from first follow-up through the
6 current publication, and statistically significant for the majority of the follow-up period
7 demonstrating that this is a strong and consistent finding in the NCI cohort (see Figures 4-3 and
8 4-4) (Beane-Freeman et al., 2009).
9 As the majority of the studies reported specific data on Hodgkin's lymphoma report on
10 just three or fewer cases, the best epidemiologic evidence is obtained from the most recent
11 evaluation of the NCI cohort by Beane-Freeman et al. (2009). This cohort study, on its own,
12 reported on more deaths from Hodgkin's lymphoma that the remainder of the epidemiologic
13 literature. Hodgkin's lymphoma was both shown to be at increased risk associated with peak
14 exposure concentrations. Peak exposures in the highest exposure category were associated with
15 a significant increase in Hodgkin's lymphoma deaths comparing death rates among workers with
16 peaks of >4 ppm to those with >0 to 2.0 ppm (RR = 3.96, 95% CI: 1.31-12.02). Across the three
17 categories of peak exposure, there was a statistically significant exposure-response trend
18 (p = 0.01). The exposure-response trend including the never-exposed workers was also
19 statistically significant (p = 0.004). The RR was also elevated for average intensity of
20 formaldehyde exposure with RR = 2.48 (95% CI: 0.84-7.32) and there were significant tests for
21 trend among only the exposed workers (p = 0.05) and all workers (p = 0.03). Similarly, there
22 were nearly significant tests for trend with cumulative exposure among only the exposed workers
23 (p = 0.08) and all workers (p = 0.06).
24 The majority of the studies reporting on Hodgkin's lymphoma did not have sufficient
25 statistical power describe any potential association with formaldehyde as the numbers of
26 observed and expect cases were small and the resulting effects estimates were imprecise. As the
27 Beane-Freeman et al. (2009) study reported on the largest number of cases and was the
28 individual study with the most detailed and objectively ascertained exposure assessment and
29 demonstrated significant exposure-response gradients, it is judged that this epidemiologic
30 evidence is supportive of an association between formaldehyde and Hodgkin's lymphoma (slso
31 see Section 4.1.2.2.4 for further evaluation).
32
33 4.5.2.6.2. Non-Hodgkin's lymphoma.
34 The only meta-analysis to specifically address non-Hodgkin's lymphoma was conducted
35 by Zhang et al. (2009) and included eleven studies. Zhang et al. (2009) reported a summary
This document is a draft for review purposes only and does not constitute Agency policy.
4-504 DRAFT—DO NOT CITE OR QUOTE
-------
1 RR = 1.08 (95% CI: 0.86-1.35). Hauptmann et al. (2009) did not specifically report on non-
2 Hodgkin's lymphoma. Beane-Freeman et al. (2009) did report on 106 deaths from non-
3 Hodgkin's lymphoma but did not identify any significant association in their categorical analyses
4 or in their tests for trend for either peak exposure, average intensity of exposure or for
5 cumulative exposure.
6 Wang et al. (2009) assessed the effect of formaldehyde exposure on the risk of non-
7 Hodgkin's lymphoma in a population-based case-control study. Semi quantitative exposure
8 metrics included average exposure intensity and average exposure probability which were
9 evaluated individually and together. Analyses use unconditional logistic regression and
10 controlled for age, family history of hematopoietic cancers, alcohol consumption, and race. For
11 the low average exposure intensity category the OR =1.4 (95% CI: 1.0-1.8), while for the
12 Medium-High category the OR =1.2 (95% CI: 0.8-1.7). For the Low average exposure
13 probability category the OR =1.3 (95% CI: 1.0-1.7), while for the Medium-High category the
14 OR =1.4 (95% CI: 0.9-2.3). The investigators also examined the risk of non-Hodgkin's
15 lymphoma among major subtypes. The risk of follicular lymphoma and chronic lymphocytic
16 leukemia/Small lymphocytic lymphoma was slightly elevated but the risk of diffuse large B-cell
17 lymphoma was OR =1.9 (95% CI: 1.3-2.6) for ever having been exposed to formaldehyde. For
18 Low average intensity exposure, the risk was OR = 2.1 (95% CI: 1.3-2.6) while for Medium-
19 High average intensity exposure, the risk was OR =1.5 (95% CI: 0.9-2.4). Even so, an
20 exposure-response relationship was demonstrated using the continuous parameterization of
21 average intensity rather that the categorical (p = 0.03). Likewise, an exposure-response
22 relationship was demonstrated using the continuous parameterization of average probability of
23 exposure rather that the categorical (p = 0.01).
24 The findings of Wang et al. (2009) provide some support for an association between
25 formaldehyde exposure and non-Hodgkin's lymphoma. It should be noted that a population-
26 based case-control study, where incidence rather than mortality defines the case—may be more
27 appropriate for cancers with relatively low mortality (i.e., CCL, large B-cell lymphoma, Small
28 lymphocytic lymphoma).
29 Aside from the semi quantitative study by Wang et al. (2009), non-Hodgkin's lymphoma
30 does not appear to be associated with formaldehyde exposure. There is not sufficient evidence of
31 a causal association between formaldehyde exposure and non-Hodgkin's lymphoma (also see
32 Section 4.1.2.2.4 for further evaluation).
33
This document is a draft for review purposes only and does not constitute Agency policy.
4-505 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.5.2.6.3. Multiple myeloma,
2 The only meta-analysis to specifically address Hodgkin's lymphoma was conducted by
3 Zhang et al. (2009) and included nine studies (Boffetta et al., 1989; Coggon et al., 2003; Dell and
4 Teta, 1995; Edling et al., 1987; Hauptmann et al., 2003; Hayes et al., 1990; Heineman et al.,
5 1982; Pottern et al., 1992; Stellman et al., 1998). Zhang et al. (2009) reported a summary
6 RR =1.31 (95% CI: 1.02-1.67). This statistically significant finding is consistent with the
7 findings of Beane-Freeman et al. (2009) who reported that peak exposures in the highest
8 exposure category were associated with a significant increase in multiple myeloma deaths
9 comparing death rates among workers with peaks of > 4 ppm to those with >0 to 2.0 ppm
10 (RR = 2.04, 95% CI: 1.01-4.12). Across the three categories of peak exposure, there was also
11 some evidence of an exposure-response trend (p = 0.08); however, there was no evidence of an
12 exposure-response trend including the never-exposed workers. The association of multiple
13 myeloma with formaldehyde exposure was also shown throughout the cohort experience (see
14 Figure 4-3 and 4-4) which adds strength to this finding.
15 The epidemiologic evidence for a causal association between formaldehyde and all
16 multiple myeloma as described by the statistically significant increased risk identified in the
17 meta-analysis of Zhang et al. (2009) and the most recently updated analysis of the NCI cohort by
18 Beane-Freeman et al. (2009) are considered to be supportive of an association between
19 formaldehyde exposure and mortality from multiple myeloma (also see Section 4.1.2.2.4 for
20 further evaluation).
21
22 4.5.2.7. Supporting Evidence from Animal Bio-Assays for Formaldehyde-Induced
23 Lymphohematopoietic Malignancies
24 Chronic animal studies provide limited supporting evidence for formaldehyde-induced
25 leukemia and lymphoma. Although the majority of chronic animal bioassays do not report either
26 leukemia or lymphoma, many studies focused primarily on the respiratory tract and did not
27 provide routine examination of other tissues, limiting the detection of leukemia and lymphoma
28 (Horten et al., 1963; Holmstrom et al., 1989; Wouterson et al., 1989; Appleman et al., 1988;
29 Monticello et al., 1996; Dalbey, 1982). Kamata et al. (1986) did examine additional tissues, but
30 there were only 5 animals at each sacrifice. Drinking water studies were similarly limited, where
31 Takahashi et al. (1986) only examined tissues from the stomach and intestines, and the study by
32 Tobe et al. (1989) only examined 20 Wistar rats per sex per exposure group including interim
33 sacrifices. Therefore, few studies have adequately evaluated the carcinogenic potential of
34 formaldehyde with respect to leukemia and lymphoma. [Complete study descriptions for all
35 experiments evaluated are found in Section 4.2.2.]
This document is a draft for review purposes only and does not constitute Agency policy.
4-506 DRAFT—DO NOT CITE OR QUOTE
-------
1 The two-year drinking water study by Til et al. (1989) in male and female Wistar rats, did
2 have adequate examination of tissues outside of the portal of entry, but found no increase in
3 leukemia or lymphoma, with only 4 tumor-bearing animals in all treatment groups sacrificed at
4 24 months (n = 200). Sellakumar et al. (1985) conducted a lifelong inhalation study in male
5 Sprague-Dawley rats exposed at 15 ppm formaldehyde. Some tissues outside of the respiratory
6 tract were routinely examined (liver, kidneys, and testes), including any organ exhibiting gross
7 pathology. However, spleen, thymus, and lymph nodes were not routinely examined, limiting
8 detection of leukemia and lymphoma, especially smaller lesions. Although Sellakumar et al.
9 (1985) was a lifelong study, there was a high mortality rate at 2 years (>80% from the figure),
10 again limiting the power of this study to detect late-in-life malignancies. Although, this study
11 did not indicate formaldehyde-induced lymphoma or leukemia, the limitations of the study
12 design should be acknowledged when interpreting these results.
13 The largest and most comprehensive cancer bioassay from formaldehyde inhalation
14 exposures is the study conducted at the Battelle Columbus Laboratory (1981). Although the
15 summary reports of this study do not discuss leukemia or lymphoma rates (Swenberg et al.,
16 1980; Kerns et al., 1983), mouse lymphoma and rat leukemia were selected by the study
17 pathologist and biostatistician for analysis and presented in the final laboratory report (Battelle
18 Columbus Laboratories, 1981).
19 Unadjusted leukemia incidence in male and female rats was similar between control and
20 formaldehyde exposed rats (9% and 6% in female rats; 9% and 4% in male rats respectively:
21 p> 0.05) (Battelle Columbus Laboratories, 1981). However, both male and female rats at the
22 highest exposure (15 ppm) exhibited significant early deaths due to nasal lesions (see
23 Figure 4-33). Additionally, the unadjusted leukemia rates included animals from all scheduled
24 sacrifices, with early time points representing 30% of the experimental animals (6, 12 and 18
25 months). Statistical analysis performed by Battelle (Tyrone extension of the Cox Test), which
26 accounted for time to lesion and survivorship rates, indicated a significant increase leukemia for
27 female rats but not male rats exposed at 15 ppm (p = 0.0003 and/? = 0.6891 respectively)
28 (Battelle Columbus Laboratories, 1981). Since only gross pathology was performed on tissues
29 outside of the respiratory tract of mid-dose animals, trend analysis was not performed (Battelle
30 Columbus Laboratories, 1981). As the first leukemia in unexposed rats was noted at 21 months,
31 the early deaths prior to that time in formaldehyde-exposed rats, reduced the number of animals
32 in which the leukemia could have been observed. Leukemia incidence in Fischer 344 rats
33 surviving at least 21 months are shown in Table 4-94. Although the histopathology was not as
34 rigorous in the mid-dose animals, leukemia was reported in these animals at a similar or greater
35 incidence than in control animals.
This document is a draft for review purposes only and does not constitute Agency policy.
4-507 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Figure 4-33. Unscheduled deaths in female F344 rats exposed to
formaldehyde for 24 months.
Source: Data extracted from the Final Report by Battelle Columbus Laboratories
(1981).
Table 4-94. Leukemia incidence in Fischer 344 rats surviving at least 21
months
Formaldehyde exposure
(8 hrs/day, 5 days a week)
Oppm
2ppm
6ppm
15 ppm
F344 male rats
8*772(11%)
10/72 (14%)
5/35 (14%)
4/39 (10%)
F344 female rats
11/72(15%)
17/72 (24%)
16/76 (21%)
6/34(18%)
aLeukemia incidence is given as the number of tumor bearing animals for the total number of
animals examined for each dose group.
Source: Data extracted from the Final Report by Battelle Columbus Laboratories, 1981
Male and female B6C3F1 mice exposed to formaldehyde for 24 months in the Battelle
Laboratory study did not experience the same rate of formaldehyde-related mortality as
formaldehyde-exposed rats (Kerns et al., 1983). However, significant early deaths were
This document is a draft for review purposes only and does not constitute Agency policy.
4-508 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
observed in male mice due to infighting and are not further considered here (Battelle Columbus
Laboratories, 1981). Statistical analysis performed by Battelle, which accounted for time to
lesion and survivorship rates, suggest an increase in lymphoma in female mice (p = 0.06, Tyrone
extension of the Cox Test) (Battelle Columbus Laboratories, 1981). However, this analysis
included animals from the 6 month sacrifice where tissues outside of the respiratory tract were
not examined. When these animals are removed from the analysis, the cumulative incidence of
lymphoma in formaldehyde-exposed female B6C3Flmice is (28%) versus controls (22%)
(p < 0.05) (see Figure 4-34). Results for the mid-dose groups are not shown here as tissues
outside of the respiratory tract were not routinely examined (e.g., spleen, liver, thymus, and
lymph nodes).
0%
12 Mo 18 Mo 24 Mo 30 Mo
Figure 4-34. Cumulative incidence of tumor bearing animals for lymphoma
in female B6C3F1 mice exposed to formaldehyde for 24 months (p < 0.05).
Note: Mice from the 6-month interim sacrifice are not included since only nasal
passages were examined.
Source: Data extracted from the Final Report by Battelle Columbus Laboratories
(1981).
This document is a draft for review purposes only and does not constitute Agency policy.
4-509 DRAFT—DO NOT CITE OR QUOTE
-------
1 Summary
2 There have been numerous chronic animal bioassays assessing lesions from inhalation
3 exposure to formaldehyde: however, few have adequate tissue pathology or study design (study
4 size and duration) to assess leukemia and lymphoma incidence. Of the available studies, the
5 study conducted at Battelle Columbus Laboratories (1981) provides the only evidence of
6 formaldehyde-induced leukemia or lymphoma. Although there were significant early deaths in
7 some of the exposure groups, formaldehyde exposure slightly increased leukemia incidence in
8 female but not male rats (Battelle Columbus Laboratories, 1981). Lymphoma incidence was also
9 elevated in female B6C3F1 mice exposed at 15 ppm versus control mice (see Figure 4-6).
10 Although one chronic inhalation study which examined tissues outside of the respiratory
11 tract did not find increased leukemia in Wistar rats (Sellakumar et al., 1985), this study did not
12 have the comprehensive histopathology, study size, or duration as the study conducted at Battelle
13 Columbus Laboratories (1981). Thus, the negative results in the Sellakumar et al. (1985) study
14 do not contradict the positive findings in fisher 344 Fisher female rats or B6C3F1 mice.
15 Therefore, the available evidence for increased Fisher 344 female rat leukemia and mouse
16 lymphoma from the Battelle study do provide limited support for the plausibility of
17 formaldehyde-induced leukemia and lymphoma.
18
19 4.5.3. Carcinogenic Mode(s) of Action
20 The EPA 2005 Guidelines for Carcinogen risk Assessment recommend a mode of action
21 (MOA) analysis when data are available for evaluation. The purpose of this MOA analysis is to
22 determine if sufficient data exist to adequately inform the exposure-response relationship for
23 cancer below the range of observed data in either human or animal studies. Since the majority of
24 the data supporting the carcinogenicity of formaldehyde comes from animal bio-assays and
25 epidemiological studies of workers, EPA must extrapolate from the observed risk of cancer
26 mortality/incidence in those studies to levels considered protective of human health for lifelong
27 environmental exposures. In this context, the EPA Cancer Guidelines provide a framework to
28 review MOA information for relevant data to establish an MOA informing appropriate low-dose
29 extrapolation.
30 The supporting data for the MOA evaluation of formaldehyde are complex, and presented
31 across multiple sections of a large document; therefore, this section includes a brief summary of
32 the biological actions of formaldehyde and key mechanistic data which are believed to be
33 relevant to the MOA evaluation (see Section 4.5.3.1). This information is not intended as a
34 stand-alone description of the evidence for a particular mechanism, but is intended to highlight
35 the major supporting arguments and direct the reader to text providing more detailed discussion.
This document is a draft for review purposes only and does not constitute Agency policy.
4-510 DRAFT—DO NOT CITE OR QUOTE
-------
1 The summary of data discussed below combines what is known about the human cancer
2 of concern (nasopharyngeal cancer, sinonasal cancer, leukemia and other lymphohematopoietic
3 cancers) with the potential formaldehyde-specific mechanisms of action to postulate
4 carcinogenic modes of action for each cancer or group of cancers (see Sections 4.5.3.2 and
5 4.5.3.2). The resulting evaluation provides multiple possible MO As for formaldehyde-induced
6 cancers where some key mechanistic events may be commonly at work in different tissues, and
7 some key events may be more relevant to a specific tissue/cancer type. Each of these MO As is
8 evaluated with respect to its relevance to human cancer, and the overall weight of evidence for
9 its relevance to formaldehyde-related human cancer.
10 Overall, multiple MO As considered relevant to humans are presented for each cancer
11 type. Although some MO As may have a greater level of supporting evidence, this reflects in part
12 how well a particular mechanism or key event may have been studied. For example, there are a
13 large number of studies across many testing systems, and levels of biological organization to
14 support the mutagenicity of formaldehyde. In contrast other likely MO As, such as viral
15 reactivation, have little direct mechanistic evidence, but the available evidence is supportive.
16 The MO As considered most relevant to upper respiratory tract cancers (e.g.
17 nasopharyngeal cancer and sinonasal cancer) are: (1) direct mutagenicity; (2) inhibition of DNA
18 repair mechanisms; (3) formaldehyde-induced cell proliferation; (4) cytotoxicity-induced cell
19 proliferation; (5) tumor promotion activity; and 6) localized immunosuppression/viral
20 reactivation (see Section 4.5.3.2). The majority of these MO As would apply equally to immune
21 cells present at the site of first contact and may also contribute to those lymphohematopoietic
22 cancers which arise from peripheral immune cells (e.g., Hodgkin's lymphoma, multiple
23 myeloma and some forms of leukemia). Additional MO As are considered, specifically for
24 formaldehyde-induced leukemia are: (1) damage of a circulating hematopoietic stem cell or
25 progenitor cell at the site of first contact; and (2) bone marrow toxicity (see Section 4.5.3.3).
26 In summary—no single MOA is singled out as the best explanation for cancer resulting
27 from formaldehyde exposure. Only one MOA—cytotoxicity induced cell proliferation—
28 suggests an exposure threshold below which the MOA would not be active. However this MOA
29 is the least applicable to humans and other MO As are considered operative at exposures below
30 exposures associated with cytotoxicity-induced cell proliferation. Therefore, multiple MO As are
31 considered supported by formaldehyde-specific mechanistic information which provide
32 biological plausibility for the cancers observed in formaldehyde exposed populations.
33
This document is a draft for review purposes only and does not constitute Agency policy.
4-511 DRAFT—DO NOT CITE OR QUOTE
-------
1 4.5.3.1. Mechanistic Data for Formaldehyde
2 4.5.3.1.1. DNA reactivity/genotoxicity/mutagenicity.
3 An agent's genotoxic potential and ability to induce mutations is a key consideration in
4 assessing a carcinogenic MO A, as cancer results from a series of genetic and epigenetic
5 alterations affecting genes that control cell growth, division, and differentiation (Hanahan and
6 Weinberg, 2000; Vogelstein et al., 1988; Kinzler and Vogelstein, 2002). The EPA Cancer
1 Guidelines suggest several lines of evidence which are key to evaluating a mutagenic MOA:
8 (1) Is the chemical under study DNA-reactive and/or has the ability to bind to DNA; (2) Does the
9 chemical generate positive results in in vitro mutagenic test systems (specifically gene mutations
10 and chromosomal aberrations); (3) Does the chemical induce manifestations of genetic damage
11 in in vivo tests (specifically gene mutations and chromosomal aberrations), and (4) Does the
12 chemical have properties and structure-activity relationships (SAR) similar to known mutagens
13 (U.S. EPA, 2005). As reviewed in Section 4.3 above, there is adequate evidence for
14 formaldehyde-induced genotoxicity and mutagenicity for consideration of these key events in
15 formaldehyde's carcinogenic MOA.
16 Formaldehyde, as a reactive chemical, forms DPX or DPC, DNA adducts and DDX or
17 DDC and may act to form adducts between other chemicals and DNA (Brutlag et al., 1969;
18 Donecke, 1978; Ohba et al., 1979; Fennel, 1999; Casanova-Schmitz and Heck, 1983, 1984; Heck
19 and Casanova, 1987; Casanova et al., 1989). The high reactivity of formaldehyde results in little
20 specificity indicating that a range of adducts and crosslinks might be expected. Formaldehyde
21 induces a variety of genotoxic and mutagenic events when tested both in vitro and in vivo
22 systems including DNA-protein crosslinks (DPC or DPX), point mutations, DNA single strand
23 breaks (SSB) and chromosomal aberrations (CAs) (see Section 4.3).
24 Numerous studies have shown that formaldehyde induces genotoxic and mutagenic
25 effects under a variety of experimental conditions (see Section 4.3 for a detailed discussion, also
26 reviewed by IARC 2006; Ma and Harris 1988; Auerbach et al., 1977). As discussed,
27 formaldehyde is known to directly react with DNA forming DPC and DNA adducts. A dose-
28 response in DPX formation has been described at exposure levels of 2-30 ppm formaldehyde in
29 rodent tissues (Casanova-Schmitz and Heck, 1983). Dose-dependent increases were also
30 observed for SSBs in peripheral blood lymphocytes and livers of rats (Im et al., 2006), and lung
31 epithelial cells of rats exposed to formaldehyde by inhalation (Sul et al., 2007), and for MN and
32 comet assay parameters in workers occupationally exposed to formaldehyde (Yu et al., 2005)
33 have also been reported. All these studies suggest that formaldehyde can induce dose-dependent
34 increase in the genotoxicity in both animals and humans. The DPX induced by formaldehyde are
35 bulky adducts which may induce distortion of the DNA helix and are likely to induce mutations.
This document is a draft for review purposes only and does not constitute Agency policy.
4-512 DRAFT—DO NOT CITE OR QUOTE
-------
1 Mutations may occur during repair of formaldehyde-induced DNA damage, or as a result of
2 replication errors during mitogenesis. Additionally, there is some evidence that DNA single
3 strand breaks (SSB) may be induced directly by formaldehyde reactivity (Grafstrom et al., 1984).
4 Clastogenic effects including increased micronuclei (MN), chromosomal aberrations (CAs) and
5 sister chromatid exchanges (SCEs) are also reported in a range of in vitro study systems.
6 Formaldehyde caused a concentration-dependent increase in mutations at the tk locus of
7 human lymphoblastoid cells (Craft et al., 1987) and clastogenicity (e.g., MN) in human cell lines
8 deficient in either DNA nucleotide excision repair (NER) or DDC repair systems even though
9 there is no change seen in DPC induction or removal between these cell lines (Speit et al., 2000).
10 These data suggest that alteration of DNA repair, not DPC removal, contributes to
11 formaldehyde-induced clastogenicity. Since DPC repair involves proteolytic removal of proteins
12 from the DNA, it has been hypothesized that single peptides or small peptide chains cross-linked
13 to the DNA as in the case of DPC are critical to formaldehyde-induced mutations.
14 Formaldehyde-induced MN and CAs are associated to concentration-dependent
15 mutagenic effects in L5178Y mouse lymphoma cells (Speit and Merk, 2002). Detailed analysis
16 of both spontaneous and formaldehyde-induced lesions indicate that recombination or deletion of
17 DNA from the thymidine kinase (tk) locus was primarily responsible for the loss of heterogeneity
18 leading to the observed mutant phenotype. Therefore, it is believed that formaldehyde is
19 mutagenic by a clastogenic mechanism, rather than through point mutations in the L5178Y
20 mouse lymphoma cell system. This finding is consistent with Craft et al. (1987) who
21 demonstrated formaldehyde-induced mutagenicity in the tk locus of TK6 human lymphoblastoid
22 cells, while Grafstrom et al. (1984) demonstrated increased SSBs in formaldehyde-exposed
23 human cell lines. The elegant series of experiments by Speit and Merk provide the possible links
24 between DPC, clastogenicity and locus-specific mutations firmly demonstrating formaldehyde-
25 induced mutations in the in vitro mouse lymphoma testing system.
26 Formaldehyde is genotoxic at the portal of entry (POE) in animal studies, resulting in
27 increased DPC formation in the nasal mucosa as discussed above. However, there are no animal
28 studies which directly examine the mutagenicity in nasal or respiratory epithelial cells in the
29 early stages of exposure. It is likely that the mutations are seen in advanced stage of tissue
30 transformation with formaldehyde exposure. With weak positive results in pulmonary lavage
31 cells (Dallas et al., 1992) and clastogenicity demonstrated in gastro-intestinal epithelial cells of
32 rats (Migliore et al., 1989), below exposure levels which trigger regenerative cell proliferation,
33 the existing evidence, although thin, supports clastogenic effects of formaldehyde.
34 Clastogenic effects are consistently reported in humans exposed to formaldehyde in the
35 industrial workplace or during anatomy or mortuary classes (see Section 4.3 for a full
This document is a draft for review purposes only and does not constitute Agency policy.
4-513 DRAFT—DO NOT CITE OR QUOTE
-------
1 discussion). Increased micronuclei have been reported in nasal epithelial cells from industry
2 workers (Ballarin et al., 1992; Ye et al., 2005), buccal epithelial cells from anatomy and
3 mortuary science students and/or staff (Kitaeva et al., 1996; Titenko-Holland et al., 1996; Burgaz
4 et al., 2001, 2002) compared to corresponding controls. Comparisons of micronuclei in nasal
5 and buccal cells of anatomy students before and after classes where they are exposed to
6 formaldehyde indicate an exposure-related increase in clastogenicity (Ying et al., 1997). An
7 examination of exfoliated buccal and nasal cells in mortuary students indicates greater increases
8 in centromere-negative micronuclei, suggesting the effects are due to chromosome breakage or
9 clastogenicity rather than aneuploidy (Titenko-Holland etal., 1996). Micronuclei were also
10 increased in a dose-dependent manner in buccal cells as well as peripheral blood lymphocytes
11 (PBLs) in mortuary students during the course of an embalming class; however, SCEs were
12 reduced in postexposure samples (Suruda et al., 1993). Buccal, oral, and nasal epithelial cells
13 present at the portal of entry may be directly exposed to formaldehyde and thus reports of
14 clastogenic effects are consistent with direct interaction of formaldehyde at the POE.
15 Mutagenicity in the form ofp53 mutations has been observed in the SCCs of chronically-
16 exposed rats, however, it is uncertain about the stage at which the mutations are induced (Recio
17 et al., 1992). There is some supporting evidence for the mutagenicity of formaldehyde in human
18 populations. Shaham et al. (2003) reported an increase in mutantp53 protein in the peripheral
19 blood lymphocytes of individuals with mean formaldehyde exposure duration of 16 years.
20 Additionally there was is a significant association between mutantp53 protein and DPC in this
21 study suggesting a relationship between the formaldehyde's genotoxic effects. More recently,
22 Zhang et al. (2010a) have reported aneuploidy in circulating hematopoietic stem cells in
23 formaldehyde exposed workers with increases in both monosomy7 and trisomy 8.
24 In summary, there are several lines of evidence supporting mutagenic effects of
25 formaldehyde exposure:
26
27 1) Formaldehyde directly interacts with DNA generating DPC, DNA adducts, and DDC
28 2) DPC in tissues at the POE exhibit a dose-response relationship to formaldehyde
29 exposure,
30 3) Formaldehyde-induced DPC are associated with formaldehyde-induced MN and CAs,
31 4) Mutations induced by formaldehyde due to small deletions and rearrangements in DNA
32 in various experimental systems are consistent with formaldehyde's observed clastogenic
33 effects (MN and CAs),
34 5) Formaldehyde-induced mutations and clastogenic effects occur at levels below where
35 significant cytotoxicity is detected, and
This document is a draft for review purposes only and does not constitute Agency policy.
4-514 DRAFT—DO NOT CITE OR QUOTE
-------
1 6) Formaldehyde exposure has been correlated to similar increased MN and CAs in human
2 buccal and oral cells corresponding to sites where formaldehyde-induced tumors arise.
3
4 4.5.3.1.2. Inhibition ofDNA repair.
5 Studies indicate that formaldehyde exposure may inhibit DNA repair mechanisms
6 directly (see Section 4.3.1.5). Grafstrom (1985) first documented formaldehyde effects on DNA
7 repair mechanisms, reporting that formaldehyde treatment of human bronchial fibroblasts in vitro
8 inhibited repair of O6-methyl-guuanine adducts induced by 7V-methyl-Nitrosurea (NMU).
9 Inhibition of DNA repair in human keratinocytes and fibroblasts cultured at 10 jiM
10 formaldehyde affected repair of DNA single strand breaks from ultraviolet light but was specific
11 to UVB and UVC, not impacting repair of single strand breaks from UVA (Emri et al., 2004).
12 To determine if formaldehyde may have similar effects in exposed humans, Hayes et al.,
13 (1997) assessed the activity O6-alkylguanine-DNA alkyltransferase (AGT) an enzyme critical in
14 repairing DNA damage induced by alkylating agents in formaldehyde-exposed mortuary students
15 previously shown to have increased micronuclei in both buccal cells and peripheral lymphocytes
16 (Suruda et al., 1993). AGT activity was lower in mortuary students with prior embalming
17 exposures versus students with no prior exposure (p = 0.08). Seventeen of 23 students had lower
18 AGT activity after the 9 week course (p < 0.05) with a larger proportion of naive students
19 demonstrating decreased activity (7 of 8) versus previously exposed students (10 of 15).
20 Although detailed exposure measurements were taken for each student, the changes in AGT
21 activity were not correlated to cumulative exposure (ppm-hrs).
22
23 4.5.3.1.3. Protein to protein cross-links.
24 Formaldehyde is a reactive molecule that is likely to interact with both low molecular
25 weight cellular components (e.g., reduced glutathione [GSH]) as well as high molecular weight
26 cellular components. Unlike nuclear DNA, which has additional membrane barriers to exposure
27 (i.e., nucleus), extracellular and intracellular proteins, are obvious primary targets for interacting
28 with formaldehyde. Formaldehyde is a well-known cross-linking agent that is used in the
29 fixation of tissues, inactivation of toxins and viruses (e.g., preparation of vaccines), and study of
30 protein-protein interactions (Metz et al., 2006). Using several identical synthetic polypeptides
31 differing on one amino acid, Metz et al. (2004) have shown that formaldehyde initially reacts
32 with the primary amino and thiol groups of amino acids forming unstable methylol adducts,
33 which later are partially dehydrated forming labile Schiff bases that are capable of forming
34 crosslinks with other amino acid residues, such as arginine, asparagine, glutamine, histidine,
35 tryptophan, and tyrosine through methylene bridges, but not between two primary amino groups.
36 The same group (Metz et al., 2006) has also shown that formaldehyde forms seven
This document is a draft for review purposes only and does not constitute Agency policy.
4-515 DRAFT—DO NOT CITE OR QUOTE
-------
1 intramolecular crosslinks in proteins with defined structure, such as insulin, involving arginine,
2 tyrosine and lysine and the TV-terminus of insulin was converted to a imidazolidinone adducts
3 similar to that observed with the synthetic peptide (Metz et al., 2004). (Figure 3-1 provides a
4 general reaction scheme for formaldehyde-mediated modifications of amino acids.)
5
6 4.5.3.1.4. Break-down of the mucociliary apparatus.
1 The mucociliary apparatus of the upper respiratory tract is the first line of defense against
8 airborne toxicants. Comprised of a thick mucus layer (epiphase), hydrophase and ciliated
9 epithelium, the mucociliary apparatus may entrain, neutralize and remove particulates and
10 airborne chemicals from inspired air (see Figure 4-7). Formaldehyde reacts with the components
11 of the mucous layer (proteins, glycoprotein, and lipids), crosslinking proteins. Formaldehyde
12 exposure induces slowing of the mucous flow, stiffing and breaking up of the mucous layer and
13 eventual mucostasis where gaps have been observed exposing the underlying hydrophase and
14 epithelium. Although ciliary beat first increases in response to formaldehyde exposure, perhaps
15 to compensate for reduced flow of the epiphase, ciliastasis ensues with both higher levels of
16 exposure and increased duration of exposure. Altered ciliary beat has been noted in as little as
17 15 minutes of exposure (1.25 ppm) with functional deficits in the mucociliary apparatus at 30
18 minutes. Altered ciliary beat has been reported at the lowest concentration tested (O.Sppm) for a
19 single 6 hour exposure. Severity of effects increase with both duration and level of exposure
20 (see Section 4.2.1.2.1).
21
22 4.5.3.1.5. Induced cell proliferation.
23 There are several reports apparently demonstrating formaldehyde-induced proliferation in
24 cells below cytotoxic levels of exposure. This phenomenon has been reported from studies
25 involving both in vitro and in vivo exposures. Tyihak et al. (2001) demonstrated significantly
26 increased cell proliferation in both HT-29 human colon carcinoma and human umbilical vein
27 endothelial cell (HUVEC) lines treated with O.lmM (the lowest dose) formaldehyde compared to
28 untreated controls (p < 0.0001). This effect was quantified as both an increase in cell number
29 over time (see Figure 4-35), and an increase in the percentage of cells undergoing mitosis at each
30 time point. The authors also report a significant (p < 0.01) inhibition of apoptosis in
31 formaldehyde-treated cells as compared to untreated cells (data not shown here). In a novel
32 system using xenotransplanted human tracheobronchial epithelial cells, formaldehyde was shown
33 to induce increased cell proliferation at doses below those required for a "massive toxic effect"
34 (Uraetal., 1989).
This document is a draft for review purposes only and does not constitute Agency policy.
4-516 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
.Q
s
3
e
"3
U
20
18
16
01
I 10
3
6
4
2-
0
B. HUVEC Cells
Day 1
Day 2
Post-treatment
Day 3
Day 1
Day 2
Post-treatment
Day 3
Figure 4-35. Effect of various doses of formaldehyde on cell number in (A)
HT-29 human colon carcinoma cells and in (B) human umbilical vein
epithelial cells (HUVEC).
Values are average of three samples ± SD; * p < 0.01 and ** p < 0.0001
compared to corresponding controls.
Source: Tyihak et al. (2001).
Some animal studies have demonstrated increased cell proliferation after formaldehyde
exposures by both inhalation and ingestion (see Section 4.2.1). However, whether sustained
increases in cell proliferation over baseline rates are observed upon exposure to subcytotoxic
doses of formaldehyde remains unclear. Several of the rodent inhalation studies demonstrate
increased cell proliferation in the nasal epithelium at formaldehyde exposures levels that were
subcytotoxic—i.e., in the absence of significant cell death. Acute formaldehyde exposures (1 to
3 days) induced increased cell proliferation at discrete locations in the nasal mucosa, where cell
proliferation was measured as a labeling index (percentage of cells pulse-labeled with tritiated-
thymidine). Reuzel et al. (1990) reported increased cell proliferation in the nasal passages
including the nasoturbinates, maxilloturbinates septum, and lateral wall in male Wistar rats
exposed at 3 ppm, but not at 0.3 or 1 ppm, formaldehyde for 22 hours/day for 3 days.
Zwart et al. (1988) reported increased cell proliferation after exposure to 1 or 3 ppm
formaldehyde, 6 hours/day for 3 days or 13 weeks in male and female albino Wistar rats. These
increases were transient at level 3 but sustained at level 2 of the nose and were not correlated
with cytotoxicity (see Table 4-17). In contrast, Wilmer et al. (1989), from the same group of
This document is a draft for review purposes only and does not constitute Agency policy.
4-517 DRAFT—DO NOT CITE OR QUOTE
-------
1 investigators using asimilar exposure regimen, reported no increase in cell proliferation after
2 repeated 8-hour exposures at 1 or 2 ppm formaldehyde for 3 days or 4 weeks. Swenberg et al.
3 (1986) demonstrated a transient increase in cell proliferation after a single 8-hour exposure to 0.5
4 or 2 ppm formaldehyde in male F344 rats but no increases after 3 days or repeated 8-hour
5 exposures. The authors suggest that adaptive responses of the nasal mucosa contribute to the
6 transient nature of formaldehyde-induced cell proliferation. After a series of acute studies at
7 various formaldehyde concentrations, Swenberg and coworkers concluded that, in addition to
8 cell proliferation being concentration-, dose- and time-dependent, the response varies by species
9 and by location of exposure in the nose (Swenberg et al., 1983, Swenberg et al., 1986).
10 Other methods of quantifying cell proliferation in the nasal mucosa have demonstrated
11 formaldehyde-induced cell proliferation at similar low exposure concentrations. For example,
12 Roemer et al. (1992) measured cell proliferation by flow-cytometry in epithelial cells harvested
13 from the nose and trachea of male Sprague-Dawley rats exposed to 2 ppm formaldehyde for 6
14 hours/day for 1 or 3 days and found increased cell proliferation after the 1-day exposure. These
15 increases were transient and were not evident after 3 days of exposure. Cassee and Feron (1994)
16 identified proliferating cells by staining for the presence of proliferating cell nuclear antigen
17 (PCNA). Formaldehyde exposure at 3.6 ppm for 6 consecutive periods of 12 hours (8-hour
18 exposures followed by 4-hour-periods of nonexposure) over three days, qualitatively increased
19 the expression of PCNA in respiratory epithelium at levels 2 and 3 of the nose in albino male
20 Wistar rats (Cassee and Feron, 1994). Hyperplasia, squamous metaplasia and frank necrosis
21 were also reported for these tissues.
22 Monticello et al. (1990, 1991, 1996) conducted in vivo cell proliferation studies in which
23 they exposed F344 rats for short durations (1, 4, 9 and 42 days) as well as much longer durations
24 (13, 26, 52 and 78 weeks) to exposure concentrations of 0, 0.7, 2.0, 6.0, 10.0, and 15.0 ppm.
25 These data are unique in that they also included low exposure concentrations. The authors
26 reported statistically significant increases in cell proliferation only at 6.0 ppm and higher
27 exposure concentrations in the short duration study and only at 10.0 ppm and higher
28 concentrations in the longer duration study. These data have undergone considerable statistical
29 analysis in several papers as well as in this document. Conolly et al. (2002, 2003) and Gaylor
30 and Conolly (2004) interpreted these data, when combined, as indicating a nonmonotonic
31 behavior at low dose. In other words, formaldehyde was judged to result in a reduction in cell
32 proliferation at low dose in comparison to baseline rates, with increased proliferation effect
33 kicking in only at exposures that were cytotoxic. However, as shown in Appendix C and in
34 Subramaniam et al. (2008), Crump et al. (2008), analysis of the individual animal data shows
35 considerable uncertainty and variability, both quantitative and qualitative, in the interpretation of
This document is a draft for review purposes only and does not constitute Agency policy.
4-518 DRAFT—DO NOT CITE OR QUOTE
-------
1 these cell proliferation data. (For example, even the control data vary over an order of
2 magnitude in some cases. See Figures 5-22 and 5-23 in Appendix C.) These analyses (which
3 were based on the replicate animal data used in the above studies) considered regional
4 formaldehyde dose to the tissue (flux), nasal site and duration of exposure, as well as the number
5 of cells at a given site. The overall conclusion in Section 5.3.3 (and detailed in Appendix C) is
6 that the cell proliferation dose-response at low dose could be reasonably described by both
7 monotonic (with and without a threshold) and nonmonotonic curves.
8 Only one study, by Monticello et al. (1989), quantified cell proliferation in primates after
9 formaldehyde exposure; this study, reported an 18-fold increase in cell proliferation in the nasal
10 epithelium (respiratory and transitional), larynx, trachea and carina of male Rhesus monkeys
11 exposed to 6 ppm formaldehyde compared to controls (see Section 4.2.1 for detailed study
12 description). The authors also noted that increased cell proliferation was seen in locations with
13 minimal histological changes, indicating proliferation may be a more sensitive predictor of
14 adverse health effects of formaldehyde exposure.
15
16 4.5.3.1.6. Cytolethality and resulting regenerative cell proliferation.
17 The toxic and cytolethal effects of formaldehyde exposure at the POE are well
18 documented after both inhalation and oral exposures (see Section 4.2.1). The nature and
19 progression of tissue injury has been best documented in rodent inhalation assays. Early effects
20 on the nasal mucosa include altered ciliary beat and mucus flow, hyperplasia and metaplasia of
21 nasal epithelium (Morgan et al., 1986a, b; Monteiro-Riviere and Popp, 1986; Maronpot et al.,
22 1986; Rusch et al., 1983; Monticello et al., 1986). These first changes may be considered
23 adaptive responses. Squamous epithelium may thicken and transitional epithelium may change
24 to squamous epithelium as evidenced by squamous hyperplasia, squamous metaplasia and
25 thickening of the epithelium in these anterior portions of the nose. Tissue damage may be
26 transient at lower formaldehyde exposures as these changes serve to protect tissue from
27 formaldehyde's reactivity. However, higher formaldehyde concentrations can overwhelm these
28 adaptive responses and result in gross tissue damage. Frank necrosis and focal erosions have
29 been reported in time- and concentration-dependent manner in rodent bioassays.
30 Both adaptive changes and cytolethality are associated with cell proliferation. However,
31 where adaptive changes are successful, e.g., prevent continued toxic insult to the tissue, cell
32 proliferation is transient. Exposure regimens where the adaptive changes are not adequate to
33 protect the tissue, would result in continued cytoxicity and cell death. Sustained damage to the
34 epithelium would result in sustained cell proliferation to compensate for cell death. A series of
35 rodent bioassays present convincing evidence that chronic inhalation exposures 6 hours a day,
This document is a draft for review purposes only and does not constitute Agency policy.
4-519 DRAFT—DO NOT CITE OR QUOTE
-------
1 5 days a week at 6, 10, and 15 ppm formaldehyde doses do result in sustained damage to the
2 nasal epithelium, sustained cell proliferation and tumor development (Kerns et al., 1983, Morgan
3 et al., 1986; Monticello et al., 1990, 1991, 1996). Work by Monticello and coworkers
4 demonstrate that chronic repeated inhalation exposures at 6, 10, or 15 ppm formaldehyde result
5 in sustained cell proliferation at the lateral meatus, mid-septum and maxilloturbinates of rat nasal
6 passages (Monticello et al., 1991, 1996).
7
8 4.5.3.1.7. Evidence for promotion.
9 There is some evidence, although mixed, that formaldehyde may promote tumor
10 development by other carcinogens, and known initiating agents by various routes of exposure.
11 Formaldehyde exposure in drinking water (0.5% formalin) increased glandular stomach
12 adenocarcinomas in male Wistar rats after initiation with 100 mg/L, 7V-methyl-
13 7V-nitrosoguanidine (MNNG), compared to MNNG-only-treated rats (Takahashi et al., 1986). In
14 white noninbred rats, inhalation exposures (3,30, or 300 ug/m3 formaldehyde 7hr/day,
15 5 days/week for 1 year) increased tumor multiplicity per animal and decreased latency of
16 benzo[a]pyrene induced tumors in white noninbred rats (Yanysheva et al., 1998.) Similarly,
17 formaldehyde skin application decreased tumor latency, in 7,12-dimethylbenz(a)anthracene
18 (DMBA) initiated hairless Oslo mice (Iversen, 1986). Although formaldehyde exposure also
19 increased the tumor multiplicity in Syrian golden hamsters where diethylnitrosamine (DEN)
20 (0.25mg IP.) was the tumor initiator, positive results were only reported for the exposure
21 regimen where hamsters were exposed to formaldehyde via inhalation 48 hours prior to DEN
22 injection, and then one a week thereafter for life. However, formaldehyde did not increase the
23 number of tumors per tumor bearing animals when only administered beginning one week after
24 all DEN injections. In contrast, bladder cancer was not enhanced by intravesical instillation of
25 0.5ml of 0.3% formalin, one week after instillation of 7V-methyl-7V-nitrosourea (MNU) in male
26 Fisher rats (Homma et al., 1986).
27 The observed promotion activity of formaldehyde has been tested in several systems, by
28 different routes of exposure. By several routes of exposure, formaldehyde enhanced tumor
29 development at a site where formaldehyde did not induce tumors alone, without the initiating
30 agent (Takahashi et al., 1986, Yanysheva et al., 1998; Iversen, 1986). Promotion activity in
31 these studies was evidenced by increased in tumor bearing animals (oral route), increase in
32 tumors per animal (inhalation routes) and decreased tumor latency compared to those animals
33 only exposed to the initiating agent (inhalation route) (Takahashi et al., 1986, Yanysheva et al.,
34 1998; Iversen, 1986). Although these experiments do not indicate how formaldehyde acts as a
This document is a draft for review purposes only and does not constitute Agency policy.
4-520 DRAFT—DO NOT CITE OR QUOTE
-------
1 promoter in these systems, it is possible formaldehyde-induced mutation, increased cell
2 proliferation or other toxic action could enhance tumor development from another agent.
3
4 4.5.3.1.8. Localized immunosuppression.
5 Formaldehyde exposure has induced localized immune suppression in experimental
6 animals (Dean et al., 1984) and in exposed workers (Lyapina et al., 2004). Repeated inhalation
7 exposures in rodents depopulated the URT and pulmonary tissues of resident macrophages,
8 resulting in a transient decrease in POE host defenses (Admas et al., 1987). After cessation of
9 exposure, the mononuclear phagocyte (MP) populations were replenished and there was a
10 subsequent increase in host defense representing both increased MP numbers and increased
11 bacteriocidal activity of the MPs. These data suggest that peak exposures of formaldehyde may
12 present localized immunosuppression for components of the mononuclear phagocyte system
13 (MPS) in tissues at the site of first contact.
14 A number of studies have evaluated the ability of formaldehyde to induce systemic
15 immunotoxic effects in humans (Ohtani et al., 2004a, b; Erdei et al., 2003; Thrasher et al., 1990,
16 1987; Pross et al., 1987). Some studies have reported altered innate immune responses
17 associated with formaldehyde exposure (Erdei et al., 2003), while others have noted adaptive
18 immune response suppression associated with formaldehyde exposure (Thrasher et al., 1990,
19 1987) and changes associated with alterations to a predominant T—lymphocyte helper 2 (Th2)
20 pattern (Ohtani et al., 2004a, b). In contrast, Pross et al. (1987) did not observe formaldehyde-
21 associated changes in systemic immune function.
22 Numerous studies have reported increased respiratory tract infections in formaldehyde
23 exposed individuals both in occupational and residential environments (Lyapina et al., 2004;
24 Krzyzanowski et al., 1990; Holness and Nethercott, 1989). Incidences of physician-diagnosed
25 chronic bronchitis were more prevalent in children (under age 15) living in homes with higher
26 formaldehyde (>60 ppb) readings in the kitchen (p < 0.001) but this effect was more pronounced
27 (p < 0.001) in children simultaneously exposed to environmental tobacco smoke (Kryzanowski
28 et al., 1990). The prevalence of chronic cough was also increased in adults living in homes with
29 measurable levels of formaldehyde, but data were not shown. Holness and Nethercott (1989)
30 assessed chronic bronchitis in 87 funeral workers, where the average formaldehyde exposure was
31 reported at 0.38 ± 0.19 ppm. Chronic bronchitis was observed in 20 funeral workers (n = 87)
32 exposed to formaldehyde compared with 3 cases of chronic bronchitis in nonexposed referent
33 controls (n = 38). A statistically significant association of self-reported chronic bronchitis and
34 decreased resistance to URT infection was reported in formaldehyde exposed workers compared
35 with controls (p = 0.02) (Lyapina et al., 2004). Of the workers, 41% had a history of chronic
This document is a draft for review purposes only and does not constitute Agency policy.
4-521 DRAFT—DO NOT CITE OR QUOTE
-------
1 respiratory infection and frequent long-lasting infectious inflammatory relapses (Group la).
2 Another group (group Ib) consisted of 17 exposed workers, 12 of whom had no history of
3 recurrent viral infections of the URT. There was a statistically significant association of
4 frequency and duration of inflammatory relapses between groups la and Ib.
5 Lyapina et al. (2004) also reported effects of formaldehyde exposure on neutrophil
6 respiratory burst activity (NRBA), the capacity of polymorphonuclear leukocytes to produce
7 reactive oxygen radicals in response to chemical or microbial stimuli using flow cytometry. A
8 suite of hematological tests and flow cytometric analysis for respiratory burst activity were
9 performed. Although no significant difference was observed in the spontaneous and stimulated
10 NRBA (median percentage of oxidizing cells) between the 29 exposed workers with URT
11 inflammation and the healthy controls (0.83 vs. 1.35, respectively), a separate comparison of the
12 NRBA of 12 workers with chronic, repeating URT infections and 17 workers with short,
13 infrequent episodes of URT inflammations was significant (0.45 vs. 1.00, p = 0.037). When the
14 NRBA of the group with chronic URT infections (n= 12) was separately compared with that of
15 the healthy controls (n = 21), the results were also significant (0.45 vs. 1.35,p = 0.012).
16 Individuals with chronic URT infections have reduced NRBA that could be due to formaldehyde
17 exposure. Neutrophils respond to tissue damage or local invasion of microorganisms and act to
18 phagocytize foreign cells. If neutrophilic activity is hampered or altered by formaldehyde
19 exposure, then the ability to fight infection will be diminished, leading to prolonged infection.
20 However, no dose-response pattern of formaldehyde exposure could be determined from this
21 study.
22
23 4.5.3.1.8. Potential for systemic transport of formaldehyde.
24 In aqueous solution formaldehyde exists in equilibrium with its hydrated form
25 methanediol (CH2OH2) (Kd = 5.5xlO-4). The equilibrium favors methanediol at physiological
26 temperature and pH (>99.9%) and is readily reversible. In biological systems, as free
27 formaldehyde is removed from aqueous solution through binding with serum proteins and
28 cellular components, the equilibrium is reestablished by dehydration of methanediol to free
29 formaldehyde. The reversible nature of this hydration reaction describes how a pool of free
30 formaldehyde may be sustained in biological systems.
31 There is strong and consistent evidence in biological testing systems in vitro that treating
32 cells with formaldehyde in an aqueous media results in significant cytoxicity, cell proliferation,
33 clastogenic effects and clear evidence of mutational events (see Section 4.3). Similarly, animal
34 bioassays where formaldehyde is administered in drinking water report portal of entry toxicity
35 including hyperplasia, increased cell proliferation, focal lesions, and tumors (see Section 4.2.1).
This document is a draft for review purposes only and does not constitute Agency policy.
4-522 DRAFT—DO NOT CITE OR QUOTE
-------
1 It should be noted that URT tissues are covered by an aqueous mucous layer, through which
2 formaldehyde must pass to react the cellular components of the URT. It has been postulated that
3 formaldehyde transports through this mucous layer and the underlying tissues as methanediol
4 (Georgieva et al., 2003).
5 The dynamic equilibrium between the hydrated and unhydrated forms of formaldehyde in
6 biological systems is well understood. Since the hydration reaction favors methanediol, it is
7 expected that exogenous formaldehyde which reaches the blood will primarily exist as
8 methanediol and is subject to physiological elimination. As free, unhydrated formaldehyde
9 continues to react with serum proteins and cellular components, the blood levels of methanediol
10 are expected to reduce as it is dehydrated to maintain equilibrium. Although some attempts to
11 measure significant changes in free formaldehyde levels in blood after inhalation exposure have
12 not been successful, the half-life in blood has been measured after i.v. injection at approximately
13 2 minutes (McMartin et al., 1979). Additionally, the detection of antibodies to formaldehyde-
14 hemoglobin adducts and formaldehyde-albumin adducts in exposures workers, smokers and
15 laboratory animals exposed via inhalation provides direct evidence that formaldehyde is able to
16 react with serum albumin and hemoglobin in biological systems (Thrasher et al., 1990, Grammer
17 et al., 1990, 1993; Dykewicz et al., 1991; Varro et al., 1997; Li et al., 2007). These data support
18 the hypothesis that exogenous formaldehyde may reach and transport through the blood. If so,
19 formaldehyde (or methanediol) may reach sites distal to the portal of entry.
20
21 4.5.3.2. Mode of Action Evaluation for Upper Respiratory Tract Cancer (Nasopharyngeal
22 Cancer, Sino-nasal)
23 From the above discussion, it can be seen that numerous mechanisms of action for
24 formaldehyde-induced cancer can be reasonably supported based on various known biological
25 actions of formaldehyde (e.g., mutation, cell proliferation, cytotoxicity, and regenerative cell
26 proliferation). Additionally, alternative actions, such as immunosuppression or viral
27 reactivation, are possible, although less data exist to evaluate these MO As. Rather than a single
28 MO A, it is plausible that a combination of these factors contribute to cancer incidence in an
29 exposed population. Considering multiple factors may help to better understand the biological
30 and mechanistic basis for the increases in cancer incidence observed in exposed human
31 populations. Unlike animal bioassays, human epidemiological studies may reflect not only the
32 effects of the agent of concern but also numerous other risk factors (e.g., viral status, diet,
33 smoking, etc.). Additionally, human studies may be impacted by biological human variability
34 across individuals, cancer biology (subtypes), wide variability in exposure regimens in human
35 populations, etc. Therefore, if the purposes of exploring the carcinogenic MO A of an agent are
36 to better understand the relevance of a given carcinogen to human populations and to inform the
This document is a draft for review purposes only and does not constitute Agency policy.
4-523 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure-response analysis, then discussions of MO As which recognize the interaction of an
2 agent with human variability and various risk factors is an appropriate analysis. For each of the
3 below postulated MO As, the detailed supporting evidence for the referenced key events are
4 provided above in Section 4.5.3.1 (e.g., mutation, cell proliferation, cytotoxicity, and
5 regenerative cell proliferation), and is not repeated here in detail each time a key event is
6 proposed as part of a postulated MOA.
7
8 a) Direct mutagenicity of formaldehyde in cells at the site of first contact: Mutations,
9 the permanent heritable changes in the genome of the cell, are a primary mechanism for
10 the activation of oncogenes or the inactivation of tumor suppressor genes. Mutagenicity
11 is the most widely recognized determinant of chemical-induced carcinogenicity, and it is
12 difficult to set aside the relevance of direct formaldehyde-induced mutations from its
13 demonstrated carcinogenicity. Formaldehyde-induced mutation in mucosal cells of the
14 URT, throat and buccal cavity may serve to initiate cells, or provide subsequent
15 mutagenic events to already initiated cells. Since the mucosal cells have proliferative
16 capacity, and cell proliferation is a normal tissue function, mutations may be fixed and
17 passed to daughter cells due to baseline cell proliferation of the tissue.
18 Relevance to humans: This MOA is relevant to humans. The well-documented DNA
19 reactivity (e.g., DPC and DNA adducts) and clastogenicity of formaldehyde in the URT
20 of laboratory animals is a direct effect of formaldehyde on tissues of first contact. As this
21 is a direct acting agent—no distribution or metabolism is required for the genotoxic
22 action—there is little expected species variability. As discussed in Chapter 3, there are
23 species differences in flux of formaldehyde into the respiratory mucosal tissues, but this
24 introduces species differences in dosimetry—not mechanism. Finally, the clastogenic
25 effects in nasal and buccal epithelial cells in formaldehyde- exposed workers confirms
26 the direct genotoxic effects of formaldehyde at the first site of contact in humans.
27 b) Decrease in DNA repair function within cells at the site of first contact: A decrease in
28 DNA repair capacity in these tissues by formaldehyde may increase total mutations over
29 time due to either endogenous or exogenous sources of mutation. Although there are
30 only a few studies which have explored the potential for formaldehyde to reduce DNA
31 repair capacity, the evidence is positive, both in vitro testing systems, and in one study of
32 occupationally exposed humans (Grafstrom, 1985; Hayes et al., 1997).
33 Relevance to humans: This MOA is considered relevant to humans. The general
34 population is exposed to various carcinogens, many with mutagenic potential, at sites of
35 first contact including; air pollution, tobacco products, nitrosamines and viruses.
36 Additionally, there are endogenous sources of DNA damage and mutagenicity in humans
37 (e-g-, lipid peroxidation, oxidative stress). The demonstration of reduced DNA repair
38 activity (O6-alkylguanine-DNA alkyltransferase activity) in formaldehyde-exposed
39 mortuary students suggests this toxic action of formaldehyde is possible in humans.
40 c) Formaldehyde-induced cell proliferation: Formaldehyde-induced cell proliferation in
41 the oral and respiratory mucosa may be considered a key event in conjunction with the
42 genotoxic effects, and induced mutational events observed with formaldehyde exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
4-524 DRAFT—DO NOT CITE OR QUOTE
-------
1 This MOA is intended to describe events which occur below exposure levels which
2 induce cell death and mucosal lesions. Therefore this MOA is comprised of two key
3 events:
4 a. Formaldehyde-induced genotoxicity or mutation
5 b. Formaldehyde-induced cell proliferation
6
7 DNA replication during cell proliferation may serve to translate DNA damage or a
8 formaldehyde-related DNA lesion into a permanent change in the sequence of nucleic
9 acids during replication of the DNA—e.g. 'fix' a mutation from DNA damage.
10 Additionally formaldehyde-induced cell proliferation may provide an opportunity for
11 initiated cells to proliferate, increasing the potential for additional mutation events and
12 transformation. The increased cell proliferation observed in the mucosal tissues in direct
13 contact with formaldehyde during inhalation exposures may serve to amplify the risk of
14 cell transformation from mutation alone. Researchers have noted that increased cell
15 proliferation may be transient in some locations as adaptive responses compensate
16 (Swenberg, 1983). However, evidence in both monkeys and rodents indicate that
17 increased cell proliferation in repeated exposures across time do result in sustained cell
18 proliferation. Data in Rhesus monkeys indicates increased cell proliferation is observed
19 beyond the nasal cavities to the larynx, trachea and carina (first tracheal branching)
20 (Monticello et al., 1989). Additionally, the authors note that cell proliferation is a more
21 sensitive indicator of effects on the epithelium, observed even when minimal histological
22 changes were present.
23 Human Relevance: Both formaldehyde-induced mutation and cell proliferation are
24 direct effects on the oral and nasal mucosa, well documented in rodent models with
25 supporting evidence in human epidemiological studies. Therefore both key events are
26 relevant to humans. As noted above, there are species differences in localized flux of
27 formaldehyde into the tissues of the oral and respiratory tract based on structural
28 differences in the airways, as well as breathing patterns. Although these differences may
29 effects the dosimetry of the formaldehyde absorption into the tissues, this only influences
30 the magnitude of response at any given location. Data from exposed Rhesus monkeys
31 which documents formaldehyde-induced cell proliferation in tissues beyond the nasal
32 cavity, and tissues with minimal histological changes supports a role for cell-proliferation
33 in the observed cancers in humans, which occur beyond the nasal cavities, and in tissues
34 without formaldehyde-related focal lesions.
35 d) Cytotoxicity-induced cell proliferation (CICP): Cell death followed by compensatory
36 cell proliferation is a reasonable MOA for agent-induced cancer. It should be noted that
37 the exposure conditions which result in CICP in rodents is known to result in significant
38 DNA reactivity and genotoxicity. Therefore, formaldehyde-induced mutations cannot be
39 excluded from this MOA. The animal bioassays support the carcinogenic potential of
40 formaldehyde in this context (Kerns et al., 1983; Selkemur et al., 1983; Monticello et al.,
41 1986). The majority of squamous cell carcinomas (SCCs) seen in formaldehyde-exposed
42 rats have been localized to the lateral meatus and mid-septum in the nasal passages
43 (Morgan et al., 1986; Monticello et al., 1996), while polyploid adenomas have
This document is a draft for review purposes only and does not constitute Agency policy.
4-525 DRAFT—DO NOT CITE OR QUOTE
-------
1 predominantly been reported at the maxilloturbinates (Morgan et al., 1986; Monticello et
2 al., 1996). Morgan et al. (1986) speculated that the maxilloturbinate was less susceptible
3 to SCC due to metabolic differences. However, Monticello et al. (1996) later suggested
4 that the smaller population of cells available at the maxilloturbinate accounted for fewer
5 SCCs observed at that site. Regardless, for those locations where SCCs do arise in rats
6 chronically exposed to formaldehyde, a clear temporal relationship can be demonstrated
7 for dose regimens capable of producing sustained epithelial damage and sustained cell
8 proliferation to eventual tumor development. Conversely, tumors are not observed in
9 these rodent models at those sites in the nasal passages without sustained cell
10 proliferation.
11 Relevance to humans: Human exposure to formaldehyde would most likely involve
12 chronic exposures to indoor levels of formaldehyde, and episodic exposures in the
13 environment or from an occupational exposure (see review in Chapter 2). An exposure
14 scenario parallel to that used in chronic rodent bioassays is unlikely (e.g. 2-15 ppm
15 6-8 hours/day, 5 days/week for 10-30 months). Exposure conditions are difficult to
16 assess especially in retrospective studies. However, only the most extreme industrial
17 work conditions would result in human exposures similar to those that produce sustained
18 compensatory cell proliferation in animal studies (i.e. 6-15 ppm 6 hours/day, 5 days per
19 week). Gross tissue lesions as reported in rodents from repeated chronic exposures at
20 6 and 10 ppm formaldehyde have not been reported from workplace exposure, and only
21 minor histopathological changes have been noted (Boysen et al., 1990; Holmstrom and
22 Wilhelmsson et al., 1989). It is possible that workers were episodically exposed to
23 formaldehyde levels which resulted in cell death and focal or gross lesions requiring cell
24 proliferation for tissue remodeling or repair. However, it is unexpected that these
25 conditions would be relevant to human environmental exposures. Therefore, although
26 regenerative cell proliferation is retained as a reasonable MOA for formaldehyde
27 carcinogenicity in experimental animals, it is unclear whether it is relevant to the
28 extrapolation of health risks to formaldehyde exposures in the general environment.
29 e) Promotion: Several animal studies indicate that formaldehyde exposure may promote
30 tumor formation due to other carcinogenic or initiating agents. There are positive data by
31 several routes of exposure (oral, dermal and inhalation) and promotion has been reported
32 as an increase in tumor bearing animals, an increase in tumors multiplicity or a decrease
33 in tumor latency with formaldehyde exposure in conjunction with the initiating agent
34 compared to tumors from the initiating agent along, or formaldehyde alone. The specific
35 key events which may explain this promotion effect are unknown but may include several
36 of the mechanisms discussed as potential MOAs for formaldehyde: mutagenicity,
37 mitogenesis, cocarcinogenicity, immunosuppression. Promotion is considered here as a
38 separate MOA, since these activities are noted for experimental conditions and tumor
39 sites where formaldehyde did not induce tumors in the absence of the initiating agent.
40 Relevance to humans: Although the human epidemiologic literature doesn't address
41 issues of tumor promotion, the nature of the cancers of concern indicate that chemical
42 promotion may be relevant to cancer incidence for these sites. Many of the risk factors
43 for nasopharyngeal cancer and other mouth and oral and URT cancers include direct
44 mutagens (e.g., smoking, dietary nitrosamines) where a promoting agent would be
45 expected to increase cancer incidence with these other risk factors. Additionally, the well
This document is a draft for review purposes only and does not constitute Agency policy.
4-526 DRAFT—DO NOT CITE OR QUOTE
-------
1 known viral risk factors for cancers of the mouth and URT also suggest a role for
2 promoting agents to human cancer incidence. Although only tangential evidence, this
3 does suggest that the promoting activity of a chemical agent, would be relevant to the
4 agent's carcinogenicity at these sites. Therefore, the potential for formaldehyde to act as
5 a promoter with other initiators—is considered relevant to formaldehyde's carcinogenic
6 MOA.
7 f) Increased URT infections/viral reactivation: Inhalation exposure to formaldehyde has
8 been shown to decrease the defenses of the body against infection through two
9 mechanisms: (1) damage to the protective mucous barrier and function of the mucociliary
10 apparatus; and (2) localized immunosuppression. These effects have been demonstrated
11 in both exposed humans and controlled animal experiments. Additionally, increased
12 respiratory tract infections are associated with formaldehyde exposure in several
13 populations. Common viral agents (e.g., Epstein barr virus) are known risk factors for
14 nasopharyngeal cancer, sinonasal cancers, and other URT cancers. Although direct
15 evidence does support increased URT infections due to formaldehyde exposure, and URT
16 infections are considered risk factors for URT cancers, direct evidence for formaldehyde-
17 related infections leading to cancer is lacking. There is however one epidemiological
18 study which finds the association between formaldehyde and nasopharyngeal cancer is
19 strengthened in Epstein barr virus sero-positive cases versus sero-negative cases. These
20 data suggest a possible role for formaldehyde in infection, viral reactivation, or
21 cocarcinogenicity with a viral agent.
22 Relevance to humans: The potential role of increased URT infections and
23 immunosuppression at the portal of entry is considered to relevant to humans. Data in
24 humans are available to support both key events in this MOA. Additionally,
25 epidemiological studies are conducted in human populations where individuals may be
26 exposed to various viral agents across the study period. Therefore, toxic actions by
27 formaldehyde which may increase URT infections, or viral-reactivation at the site of first
28 contact, could be acting in conjunction with viral agents to contribute, in part, to observed
29 associations between formaldehyde exposure and increased URT cancer.
30
31 Summary and integration of key events
32 Each of the hypothesized MO As discussed above to better understand the carcinogenic
33 potential of formaldehyde is supported by formaldehyde-specific evidence, either in animal
34 studies, human studies, or both. For those key events studied in animal models such as cell
35 proliferation, genotoxicity, degradation of the mucociliary apparatus and CICP, supporting
36 evidence is available in more than one species, multiple strains (e.g., rats) and has been reported
37 by multiple researchers. Therefore the overall database supporting these key events in laboratory
38 studies, and their corresponding MO As is fairly large. In contrast, some key events relevant to
39 humans, but less studied in animal models may have a small supporting database (e.g., increased
40 respiratory tract infections). These alternative MO As are retained as potentially relevant to the
41 carcinogenic action of formaldehyde as the intent of this discussion is to identify modes of action
This document is a draft for review purposes only and does not constitute Agency policy.
4-527 DRAFT—DO NOT CITE OR QUOTE
-------
1 will may contribute to the observation of increased upper respiratory tract cancers in exposed
2 human populations. It is noted that additional study is needed to better understand the range of
3 effects formaldehyde may have at sites of first contact in humans.
4 The MO As which include genotoxicity, mutation, decreased DNA repair, increased cell
5 proliferation and CICP are interrelated. Conditions which provide both a source of cell
6 proliferation and increased mutation would be expected to increase neoplastic transformation.
7 Formaldehyde acts on the target tissue, the respiratory epithelium, to induce each of these events.
8 However, these key events operate across different exposure ranges and present different
9 exposure response relationships. For example, formaldehyde-induced mutations would be
10 expected across the exposure range, where any incremental increase in genotoxicity and
11 formaldehyde-related mutation would contribute to background levels, with the potential to
12 increase cancer risk incrementally. In contrast, focal and gross lesions to the respiratory mucosa
13 due to cytolethality are not observed unless exposure concentrations are sufficient to provide
14 localized tissue doses (flux) required to result in cell death and related compensatory cell
15 proliferation. Since tissue dose (flux) is dependent on not only exposure concentration but also
16 duration of exposure and location in the respiratory tract (see Section 3.4), and varies by species,
17 correlation of exposure concentrations to tissue responses directly are complex. Exposure
18 response relationships for the key events (cell proliferation, genotoxicity, degradation of the
19 mucociliary apparatus and CICP) are reported by exposure concentration, not tissue flux, which
20 would be a more biologically relevant measure.
21 Although the tissue dose-response relationships for formaldehyde induced mutation,
22 mitogenesis and cytolethality are different, the effects at the tissue level cannot be easily
23 disaggregated. At any given exposure concentration, target cells in the respiratory tract will
24 experience different effective tissue concentrations of formaldehyde. Measurement of cell
25 proliferation, DNA protein crosslinks, or genotoxicity may require examining a population of
26 cells which would have been subject to different flux rates of formaldehyde (see Chapter 3).
27 Similarly, when evaluating the tumor dose response, cells within the target tissue will represent a
28 range of target tissue formaldehyde concentrations. Therefore, an integrated MOA scheme is
29 hypothesized where key events may influence the observed tumor response differentially across
30 the exposure response range (see Figure 4-36). This schematic illustrates the potential for
31 genotoxicity and formaldehyde-induced mutation to occur where tissue dose (flux of
32 formaldehyde into the tissue is minimal). Where tissue dose is increased, formaldehyde-induced
33 cell proliferation is observed in addition to genotoxicity. As tissue dose increases and
34 formaldehyde effects on the respiratory mucosa are more severe, gross pathology including focal
35 and gross lesions due to cell death are noted. Therefore, several of the MO As presented above
This document is a draft for review purposes only and does not constitute Agency policy.
4-528 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Tumor
Response
Integrated Dose Response Curve for/Formaldehyde
A
Exposure Concentration
Mutagenesis
Basal Cell Proliferation
Induced mitogenesis
>
Adaptive Cell Proliferation
Regenerative Proliferation
Figure 4-36. Integrated MOA scheme for respiratory tract tumors.
may be operative and relevant to human exposures at exposure levels resulting in minimal tissue
flux—(a) direct formaldehyde genotoxicity and resulting mutation, (b) inhibition of DNA repair
and c) formaldehyde induced cell proliferation in conjunction with mutation. CICP, which
involves localized and gross tissue lesions, would be operative at higher exposure levels. There
is little data to inform the dose range over which the remaining hypothesized MO As may operate
(promotion and increased respiratory tract infections/viral action).
4.5.3.3. Mode(s) of Action for Lymphohematopoietic Malignancies
4.5.3.3.1. MOA evaluation for Leukemia.
Leukemia may arise from stem cells or progenitor cells in the bone marrow (e.g., acute
and chronic myeloid leukemia) or from mature lymphocytes (e.g., chronic lymphatic leukemia,
hairy cell leukemia) (see Figure 4-32, Section 4.5.2). Although there is a consistent association
between formaldehyde exposure and forms of leukemia when considered as group of diseases
(see Table 4-91, Section 4.5.2), the strongest and most consistent associations are seen
specifically with myeloid leukemia. Little evidence supports an association between
formaldehyde exposure and other specific leukemia subtypes, although two studies support a
strong association between formaldehyde and "other leukemia and unspecified leukemia (ICD-9
This document is a draft for review purposes only and does not constitute Agency policy.
4-529 DRAFT—DO NOT CITE OR QUOTE
-------
1 code 207). Therefore, this MOA evaluation will focus on mechanisms which may impact all
2 forms of leukemia (e.g., bone marrow toxicity) or those specific to myeloid leukemia. The
3 mechanistic data supporting the key events in this analysis are presented in Section 4.5.3.1. For
4 each of the below postulated MO As, the detailed supporting evidence for the referenced key
5 events are provided above in Section 4.5.3.1 (e.g., mutation, cell proliferation, cytotoxicity, and
6 regenerative cell proliferation), and is not repeated here in detail each time a key event is
7 proposed as part of a postulated MOA.
8
9 a) Direct effects of formaldehyde on a circulating stem cell or progenitor cell present at
10 the portal of entry: Hematopoietic stem cells do circulate throughout the body and can
11 be harvested from peripheral blood. Formaldehyde exhibits a range of toxic effects at the
12 site of first contact including genotoxic effects believed to be mediated by direct DNA
13 reactivity (see Section 4.3). Formaldehyde is known to directly react with blood
14 components in formaldehyde exposed humans and animals resulting in both hemoglobin
15 and albumin adducts (Thrasher et al., 1990, Grammer et al., 1990; Grammer et al., 1993;
16 Dykewicz et al., 1991; Varro et al., 1997; Li et al., 2007). Therefore, it has been
17 hypothesized that formaldehyde could react with DNA in circulating hematopoietic stem
18 cells (Zhang et al., 2009) resulting in heritable mutations which may contribute to
19 leukemia incidence. Recently, Zhang et al. (2010a) have tested the hypothesis that
20 exogenous formaldehyde may damage circulating stem cells. Clastogenic effects were
21 found in circulating hematopoietic stem cells cultured from formaldehyde exposed
22 workers. The reported aneuploidy was demonstrated as significant increases in both
23 monosomy 7 and trisomy 8. These specific chromosomal changes are consistent with
24 those reported for agent-induced myeloid leukemia (Zhang et al., 2010a).
25 Relevance to Humans: This hypothesized MOA is considered relevant to humans.
26 Supporting evidence is found in humans for formaldehyde direct reactivity with blood
27 proteins (e.g., albumin and hemoglobin) as well as clastogenic effects in circulating
28 hematopoietic stem cells in formaldehyde exposed workers.
29 b) Bone marrow toxicity: Direct bone marrow toxicity is the most studied leukemogenic
30 action for an endogenous agent (e.g., benzene, ionizing radiation). It is believed that an
31 agent which exerts its toxicity on the bone marrow, resulting in translocations and
32 heritable mutations in hematopoietic stem cells may cause leukemia. It has been
33 hypothesized that formaldehyde may transport to the bone marrow in its hydrated form
34 (methandiol) and react with cellular proteins, and DNA causing direct effects on
35 components of the bone marrow. A study by Ward et al. (1983) has shown that mice
36 given a single dose of formaldehyde, or methanol by gavage were able to develop
37 cytogenetic changes (SCEs, CAs and aneuploidy) in bone marrow cells suggesting that
38 formaldehyde can reach to bone marrow and induce toxicity. Pancytopenia (a reduction
39 in blood borne cells formed in the bone marrow) is a symptom of direct bone marrow
40 toxicity and is observed with other leukemogenic agents (e.g., benzene, ionizing
41 radiation). A recent review of 8 published studies of formaldehyde exposed workers in
42 China by Tang et al. (2009) indicates 7 of the studies provide evidence of reduced white
43 blood cell counts, platelet levels and hemoglobin levels associated with formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
4-530 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure. A study of occupationally exposed nurses provided a correlation between
2 decreased white blood cell counts and formaldehyde exposure (Kuo et al., 1997). A
3 recent study by Zhang et al. (2010a) provides the best evidence for bone marrow toxicity,
4 where they report not only a reduction in white blood cell counts, but reductions in cell
5 counts of all the blood cells, as well as increased mean cell volume. Although these
6 reductions did not meet the clinical definition of pancytopenia (when averaged across the
7 study population), reduction of all blood borne cells formed in the bone marrow is
8 consistent with the bone marrow toxicity associated with pancytopenia seen with other
9 leukemogens (Zhang etal., 201 Ob).
10 Relevance to Humans: This hypothesized MOA is considered relevant to humans.
11 Supporting evidence is found in humans for bone marrow toxicity in formaldehyde
12 exposed workers.
13
14 4.5.3.3.2. MOA evaluation for Lymphomas (e.g., Hodgkin's lymphoma, Multiple myeloma).
15 The general MOA for formaldehyde is based on direct chemical reactivity and toxic
16 effects at the portal of entry (POE). Formaldehyde is directly and indirectly genotoxic, and
17 reacts with cellular proteins and DNA in cells which it comes into contact. Additionally,
18 immunosuppression, viral reactivation and promotion effects are relevant to lymphoma and
19 related malignancies. Therefore, the key events for the adult cell lymphoid cancers would
20 include these actions. For each of the below postulated MO As, the detailed supporting evidence
21 for the referenced key events are provided above in Section 4.5.3.1 (e.g., mutation, cell
22 proliferation, cytotoxicity, and regenerative cell proliferation), and is not repeated here in detail
23 each time a key event is proposed as part of a postulated MOA.
24 Lymphoid tumors (e.g., lymphocytic leukemia, B-cell lymphoma, mantle cell lymphoma
25 [a rare form on non-Hodgkin's lymphoma] and myeloma) may arise from cells present at the
26 portal of entry (POE) (see Figure 4-32). The location and function of mature lymphocytes
27 contribute to their vulnerability to transformation by agents at the POE. Therefore, a brief
28 summary of the immuno-biology of these cells is provided in order to provide context for the
29 MOA evaluation:
30
31 Location: Lymphocytes are present in the oral and respiratory tract epithelium, as well as
32 in cell aggregates and tertiary immune structures (e.g., germinal centers) in the mucosal
33 tissues (Zuercher and Cebra, 2002; Zuercher et al., 2002; Wu et al., 1997; Kupper et al.,
34 1990). These mucosa-associated lymphoid tissues (MALT) provide the opportunity for
35 formaldehyde to directly interact with components of the immune system present at the
36 POE (Wu et al., 1997, Claeys et al., 1996, Park et al., 2003; Fujimura, 2000).
37 Intraepithelial lymphocytes are present in the pseudostratified epithelium of the
38 nasopharyngeal passages and there are aggregates of immune cells and germinal cells
39 present in these tissues. Crypts containing mature lymphocytes exist at the surface of the
40 nasal epithelium (Fujimura, 2000). Microfold cells or M-cells form the crypts, where the
This document is a draft for review purposes only and does not constitute Agency policy.
4-531 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
lymphocytes are covered by a thin membrane (see Figure 4-37). Functionally, these
lymphocytes identify and process foreign antigens at the POE (Fujimura, 2000).
Therefore the mature lymphocytes within these crypts, exposed to exogenous agents, are
involved in active immune responses to foreign antigens.
M
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Figure 4-37. Location of intraepithelial lymphocytes along side epithelial
cells in the human adenoid.
Source Fujimura et al., 2000
Clonal expansion: Mature lymphocytes (both B and T-cells) clonally expand their
populations in response to an exogenous antigen stimulation when a humoral immune
response is stimulated. Therefore cell proliferation is a normal function of these mature
lymphocytes and occurs every time there is an infection. Cell proliferation of mature B
and T-cells, responsive to a particular antigen, occurs in active germinal centers
(including those within the respiratory tract). Cells may be exposed to exogenous agents
during the immune response, or cells responding in the germinal center may have
previously been in the epithelium or M-cell crypt.
Somatic hypermutation: Normal immune function includes the process of somatic
hypermutation where B-cells undergo DNA rearrangement of the variable region genes to
produce novel antibodies specific to a given antigen. This process is key to adaptive
immunity and demonstrated by the basic principles of immuno-biology which underlie
vaccination theory. Gene sequencing of adult B-cell lymphomas and leukemias indicate
that the chromosomal regions involved in somatic hypermutation correspond to known
oncogenes in these cancers. The vulnerability of these processes is evidenced by the
This document is a draft for review purposes only and does not constitute Agency policy.
4-532 DRAFT—DO NOT CITE OR QUOTE
-------
1 observation that approximately 90-95% of adult lymphomas and leukemias are of B-cell
2 origin (Gordon et al., 2003). Formaldehyde-induced protein-protein crosslinking could
3 disrupt cellular processes including somatic hypermutation and cell mitosis, resulting in
4 agent-induced translations similar to those found in spontaneous B-cell malignancies.
5 PostulatedMOAs for lymphomas (e.g., Hodgkin's lymphoma, multiple myeloma): For
6 each of the below postulated MOAs, the detailed supporting evidence for the referenced
7 key events are provided above in Section 4.5.3.1 (e.g., mutation, cell proliferation,
8 cytotoxicity, and regenerative cell proliferation), and is not repeated here in detail each
9 time a key event is proposed as part of a postulated MOA.
10 a) Direct or indirect formaldehyde-induced mutations in cells at the site of first
11 contact: Immune cells including intraepithelial lymphocytes, and cells in MALT are
12 collocated with the epithelial cells from which URT cancers arise (see Figure 4-37).
13 Therefore direct and indirect mutagenic potential of formaldehyde is equally applicable
14 to components of the immune system present in these tissues. Mutations, the permanent
15 heritable changes in the genome of the cell, are a primary mechanism for the activation of
16 oncogenes or the inactivation of tumor suppressor genes. Mutagenicity is the most
17 widely recognized determinant of chemical-induced carcinogenicity, and it is difficult to
18 set aside the relevance of direct formaldehyde-induced mutations from its demonstrated
19 carcinogenicity. Formaldehyde-induced mutation in immune cells present at the site of
20 first contact, may provide initiated cells or subsequent mutagenic events to already
21 initiated cells. The competence of the immune system relies on the proliferation of
22 peripheral blood lymphocytes in response to immune challenge (e.g. infection).
23 Additionally, heritable changes to the variable gene regions in B-cells generated during
24 somatic hypermutation are essential to adaptive immunity (e.g., immunization)
25 demonstrating that permanent heritable changes in the DNA of peripheral B-cells are
26 passed to daughter cells and retained in the body for decades. Any agent-induced
27 mutations would be similarly propagated and retained with the potential to contribute to
28 the transformation of mature lymphocytes.
29 Relevance to humans: This MOA is relevant to humans. The well-documented DNA
30 reactivity (e.g., DPC and DNA adducts) and clastogenicity of formaldehyde at the POE in
31 laboratory animals is a direct effect of formaldehyde on tissues of first contact, and these
32 mechanisms are considered relevant to humans. As with epithelial cells, clastogenic
33 effects in peripheral lymphocytes are documented in formaldehyde-exposed students and
34 workers, confirming the genotoxic effects of formaldehyde in immune cells, from which
35 lymphomas and related hematopoietic diseases may arise (see Section 4.5.2,
36 Figure 4-32).
37 b) Formaldehyde-induced protein-protein crosslinks may disrupt somatic-
38 hypermutation: Although not as well studied as DNA-protein crosslinks, formaldehyde
39 also forms crosslinks between amino acids on proteins (see Section 4.5.3.1.3 for details).
40 Specific oncogenes for malignancies which arise from mature B-cells are linked to errors
41 in the process of somatic hyper-mutation (Greaves et al., 2004). If formaldehyde creates
42 protein crosslinks in competent B-cells which affects the process of DNA rearrangement,
43 formaldehyde may generate translocations and related oncogenes similar to those
44 observed in spontaneous B-cell malignancies.
This document is a draft for review purposes only and does not constitute Agency policy.
4-533 DRAFT—DO NOT CITE OR QUOTE
-------
1 Relevance to humans: This hypothesis has not been tested in either exposed human or
2 animal test systems. However, the link between somatic-hypermutation and B-cell
3 oncogenes is well established and perturbation of this process by an exogenous agent is a
4 reasonable extension of the current understanding of the etiology of B-cell malignancies.
5 c) Increased URT infections/viral reactivation: Inhalation exposure to formaldehyde has
6 been shown to decrease the defenses of the body against infection through two
7 mechanisms: (1) damage to the protective mucous barrier and function of the mucociliary
8 apparatus in the nasal passages; and (2) localized immunosuppression (see
9 Section 4.5.3.1). These effects have been demonstrated in both exposed humans and
10 controlled animal experiments. Additionally, increased respiratory tract infections are
11 associated with formaldehyde exposure in several populations. Common viral agents
12 (e-g-, Epstein barr virus) are known risk factors for malignancies which arise from mature
13 lymphocytes. Thus, increased URT infections or viral reactivation due to formaldehyde
14 exposure may influence the incidence of these cancers.
15 Relevance to humans: The potential role of increased URT infections and
16 immunosuppression at the portal of entry is considered to be relevant to humans. Data in
17 humans are available to support both key events in this MOA. Additionally, coexposure
18 to infectious agents (including viruses) would be expected in participants in an
19 epidemiological study, suggesting an MOA which acted in conjunction with infectious
20 agents may be relevant to agent-induced cancer. Therefore, immunotoxic actions by
21 formaldehyde which may increase URT infections, or viral-reactivation at the site of first
22 contact, could be acting in conjunction with viral agents to contribute, in part, to observed
23 associations between formaldehyde exposure and increased lymphoma and related
24 diseases.
25
26 4.5.3.3.3. Summary and evaluation of hypothesized MOA(s) for Lymphohematopoietic
27 Malignancies.
28 The well-documented direct toxic action of formaldehyde on cells at the site of first
29 contact is a general effect based on the reactivity of formaldehyde with cellular components
30 (e.g., proteins and DNA) (see Section 4.5.3.1). As a general effect, it is reasonable that these
31 toxic effects would be relevant to all cells which come into contact with formaldehyde. The
32 current debate regarding the biological plausibility of formaldehyde-induced
33 lymphohematopoietic malignancies centers around a perspective that the diseases within this
34 general grouping are systemic cancers arising only out of bone marrow toxicity (Heck et al.,
35 2006, Pyatt et al., 2008) and that it is implausible for formaldehyde to induce bone marrow
36 toxicity. The above MOA evaluation expands the current debate by considering the impact of
37 POE toxicity on elements of the immune system and cancers might arise from these cells (see
38 Section 4.5.3.3.2) and by presenting data which support the observation that formaldehyde is
39 associated with bone marrow toxicity and damage to circulating stem cells in exposed humans
40 (see Section 4.5.3.3.1).
This document is a draft for review purposes only and does not constitute Agency policy.
4-534 DRAFT—DO NOT CITE OR QUOTE
-------
1 As significant increases in free formaldehyde in peripheral blood from exogenous
2 exposure has not been detected (Heck et al., 1987), it has been hypothesized that formaldehyde
3 does not transport and therefore cannot exert toxic effects outside of the tissues at the site of first
4 contact (Heck et al., 2006, Pyatt et al., 2008). In contrast to this hypothesis, at least one study by
5 Ward et al. (1983) in rats has shown that orally administered formaldehyde was able to induce
6 bone marrow genotoxicity, and also effects are seen in formaldehyde-exposed humans, which
7 indicate systemic effects on the hematopoietic system including reduced white blood cell counts,
8 clastogenic effects in peripheral blood lymphocytes and aneuploidy in circulating stem cells
9 (Tang et al., 2009, Zhang et al., 2010a; Section 4.5.3.1). These observed effects in humans are
10 consistent with agent-induced bone marrow toxicity and are observed with other well-studied
11 exogenous leukemogens (e.g., benzene and ionizing radiation). It is unknown if formaldehyde is
12 distributed systemically to exert its effects directly on cells in the bone marrow or if damage to
13 circulating stem cells or progenitor cells would be sufficient to result in the observed effects in
14 humans (Zhang et al., 2010b). Additional research is needed to better determine the potential for
15 systemic transport of formaldehyde considering both detection of its hydrated form (methylene
16 glycol) as well as formaldehyde protein adducts (e.g., formaldehyde-GSH, formaldehyde-Hb and
17 formaldehyde-albumin). Similarly the results of Zhang et al. (2010a) need to be extended
18 (analysis for additional chromosomal aberrations) and repeated. Although further evidence is
19 needed to better understand the hypothesized mechanisms for formaldehyde-induced effects on
20 hematopoietic stem cells, the observed hematologic effects in humans cannot be set aside.
21 Therefore, however unlikely, the current data support the biological plausibility of formaldehyde
22 effects on the hematopoietic system.
23
24 4.5.4. Hazard Characterization for Formaldehyde Carcinogenicity
25 Formaldehyde is Carcinogenic to Humans by the Inhalation Route of Exposure
26 Human epidemiological evidence is sufficient to conclude a causal association between
27 formaldehyde exposure and nasopharyngeal cancer, nasal and paranasal cancer, all leukemias,
28 myeloid leukemia and lymphohematopoietic cancers as a group. Epidemiological evidence is
29 also strongly supportive of, but in itself not sufficient for, a conclusion of causal association for
30 other upper-respiratory tract cancers, Hodgkins lymphoma, or multiple myeloma. Animal
31 bioassays consistently demonstrate formaldehyde-induced nasal cancers in rodents which
32 provide strong support for the observed upper respiratory tract cancers in humans. Limited
33 evidence from animal bioassays is available to support the conclusion from human
34 epidemiologic data that formaldehyde causes some types of lymphohematopoietic cancers.
This document is a draft for review purposes only and does not constitute Agency policy.
4-535 DRAFT—DO NOT CITE OR QUOTE
-------
1 The epidemiologic evidence is sufficient to characterize the association between
2 formaldehyde exposure and nasopharyngeal cancer as causal in humans, based on the results
3 from a large and well-followed longitudinal cohort study of industrial workers and several case-
4 control studies (Hauptmann et al., 2004; Vaughan et al., 2000; West et al., 1993; Vaughan et al.,
5 1986b). The epidemiologic evidence of association between formaldehyde exposure and other
6 upper respiratory tract cancers (see Section 4.1.2.1.3) is consistent with, and supportive of, a
7 causal association but insufficient on its own to reach a causal conclusion. Case-control studies
8 have demonstrated associations between formaldehyde exposure and rare cancers of the URT.
9 Luce et al. (2002) evaluated pooled data from 12 case-control studies and demonstrated a
10 statistically significant increased risk between formaldehyde exposure and sinonasal cancer.
11 Hypopharyngeal cancer was linked with formaldehyde exposure with an OR of 3.78 (95% CI:
12 1.50-9.49) in another case-control study (Laforest et al., 2000). Hauptmann and colleagues
13 (2004) concluded that in spite of the small numbers of deaths from cancers of the URT, the
14 positive associations with average intensity and peak exposure were consistent with the
15 carcinogenicity of formaldehyde at these sites of first contact. The finding that formaldehyde
16 inhalation causes nasal squamous cell carcinoma in animals (see Section 4.2.1.2) further supports
17 a causal association of formaldehyde exposure and increased risk of upper respiratory tract
18 cancer in humans. Both humans and animals developed tumors within the upper respiratory
19 tract, the POE site expected to receive direct exposure to formaldehyde.
20 Also, overall, there is a consistent association between formaldehyde exposure and
21 various forms of lymphohematopoietic (LHP) cancers, with all leukemias, myeloid leukemia
22 specifically, Hodgkin's lymphoma and multiple myeloma demonstrating the greatest strength
23 and consistency of results. Where exposure-response data exist, exposure-response trends have
24 been seen for all LHP malignancies, all leukemia, myeloid leukemia and Hodgkin's lymphoma
25 (Pinkerton et al., 2004; Beane-Freeman et al., 2009). Taken together, the data demonstrate a
26 consistent association, across various worker populations, with the expected temporal association
27 to exposure and defined exposure-response relationships in two different worker cohorts. The
28 strongest associations tend to be with myeloid leukemia and Hodgkin's lymphoma. The criterion
29 of reasonable biological plausibility is easily met for the majority of the diseases which
30 contribute to an observation of all LHP cancers, specifically the cancers derived from mature
31 lymphocytes. Although it is largely unknown how inhalation exposure to formaldehyde would
32 influence bone-marrow derived malignancies, new evidence supports formaldehyde-induced
33 bone marrow toxicity as well as damage to circulating stem cells which may be of importance to
34 formaldehyde's leukemic potential (Zhang et al., 2010, Tang et al., 2009). Limited support for
35 the potential for formaldehyde-induced LHP cancer is found when considering animal bioassays,
This document is a draft for review purposes only and does not constitute Agency policy.
4-536 DRAFT—DO NOT CITE OR QUOTE
-------
1 where formaldehyde exposure influenced leukemia and lymphoma incidence in female animals
2 of two species (rats and mice) in a long-term bioassay (Battelle Laboratories, 1981).
O
4 4.6. SUSCEPTIBLE POPULATIONS
5 "Susceptible subpopulations" is used here to refer to factors, such as life stage, genetics,
6 health status, etc., that may predispose individuals to greater response to an exposure. This
7 greater response could be achieved either through differences in exposure to the chemical or
8 differences in underlying toxicokinetic (TK) and toxicodynamic (TD) differences between the
9 susceptible and other populations. For example, life stages may include the developing
10 individual before and after birth up to maturity (e.g., preconception, embryo, fetus, young child,
11 adolescent), adults, or aging individuals. Another susceptibility factor is genetics. Specifically,
12 susceptible subpopulations may also include people with specific genetic polymorphisms that
13 render them more vulnerable to a specific agent or people with specific diseases or pre-existing
14 conditions (e.g., asthmatics). The term may also refer to gender differences, lifestyle choices, or
15 nutritional state (U.S. EPA, 2002, Section 4.3.2.3).
16 A discussion of a comprehensive list of all possible susceptibility factors affecting
17 exposure and response to formaldehyde, or any chemical, is not possible. Therefore, the
18 discussion of susceptibility factors focuses on (1) factors hypothesized to be of importance to
19 formaldehyde; and (2) factors for which there are available formaldehyde data. A partial list of
20 these factors includes gender, genetic polymorphisms, preexisting disease status, nutritional
21 status, diet, and previous or concurrent exposures to other chemicals. Qualitatively, the presence
22 of multiple susceptibility factors will increase the variability that is seen in a population response
23 to formaldehyde toxicity.
24
25 4.6.1. Life Stages
26 Individuals at different life stages are physiologically, anatomically, and biochemically
27 different. Examples include physiological changes that occur through the lifespan (Selevan et
28 al., 2000). They may also have distinctive exposure pathways (i.e., transplacental, breast milk
29 ingestion), and exhibit differences in behavior (U.S. EPA, 2006b; NRC, 1993). Early life stages
30 (i.e., during development, prior to mature adulthood) and the later life stages (i.e., aging) differ
31 greatly from mature adulthood in body composition, organ function, and many other
32 physiological parameters that can impact the TK and/or TD of chemicals and their metabolites
33 (TLSI, 2003). This section presents and evaluates the pertinent published literature available to
34 assess whether and how individuals of differing life stages may respond differently to
35 formaldehyde.
This document is a draft for review purposes only and does not constitute Agency policy.
4-537 DRAFT—DO NOT CITE OR QUOTE
-------
1
2 4.6.1.1. Early Life Stages
3 4.6.1.1.1. Factors influencing exposure and dosimetry.
4 For all life stages, the primary exposure routes for formaldehyde include inhalation and,
5 in some cases, ingestion (see Chapter 5). Some exposure scenarios may be child specific. For
6 example, to the extent that the presence of baby furniture produced with formaldehyde in a
7 child's house contributes to greater concentrations in a child's room, exposures for very young
8 children in those circumstances may be increased (Environment California, 2008). As with all
9 chemicals, placental transfer and breast milk ingestion are exposure pathways that are unique to
10 early life stages. Studies assessing early life stage exposure pathways to formaldehyde have not
11 been performed. Presumably, unmetabolized formaldehyde reacts too quickly to be effectively
12 transported from mother to fetus by placental transfer; in addition, formaldehyde is not lipophilic
13 and is therefore unlikely to accumulate in breast milk. However, the relevant dose metric for
14 formaldehyde-related effects may vary depending on the specific target of concern (e.g., direct
15 toxicity at the portal of entry versus systemic effects); insufficient information is currently
16 available to determine whether individuals in different life stages are at higher risk for exposure
17 to specific target tissues.
18 There are some calculations however which shed light on lifestage differences in the
19 inhaled tissue dose at the portal of entry. Using respiratory tract surface areas and ventilation
20 rates reported in the literature and the scheme in EPA (1994), Ginsberg et al. (2005) calculated
21 that overall extrathoracic absorption of highly reactive and soluble gases is similar in adults and
22 children. These results are in agreement with that of Garcia et al. (2009) who used
23 computational fluid dynamics to study differences in the nasal dosimetry of reactive, water-
24 soluble gases between 5 adults and 2 children, aged 7 and 8 years old. Overall uptake efficiency,
25 average flux (rate of gas absorbed per unit surface area of the nasal lining) and maximum flux
26 levels over the entire nasal lining did not vary substantially between adults and children (1.6-fold
27 difference in average flux and much less in maximum flux). On the other hand, the local flux of
28 formaldehyde varies between the two children by a factor of 2 to 4 at various distances along the
29 septal axis of the nose. The results in Garcia et al. (2009) have been further described and
30 evaluated in Appendix C.I. Under normal resting breathing conditions, it is expected that very
31 little formaldehyde is delivered to the lung. However, under higher activity as well as mouth
This document is a draft for review purposes only and does not constitute Agency policy.
4-538 DRAFT—DO NOT CITE OR QUOTE
-------
1 breathing scenarios, both of which appear likely to happen more regularly in children,
2 formaldehyde dose to the lung will be substantial.7
3 The toxicokinetic characteristics of formaldehyde are described in Chapter 3, with
4 absorption and distribution studies discussed in Sections 3.2 and 3.3. Studies to assess
5 differential absorption or distribution of formaldehyde in early life stages have not been
6 performed and represent a significant data gap. The metabolism of formaldehyde is described in
7 Section 3.4. Expression of the enzymes that metabolize formaldehyde (ALDH2 and FALDH,
8 and specifically ADH3) is known to be developmentally regulated and thus may alter the TK of
9 formaldehyde in early life stages. ADH3 is ubiquitously expressed and is present in the human
10 fetus, neonate, and 1- to 10-year-old children (Hines and McCarver, 2002; Estonius et al., 1996).
11 During early development in rodents, when neurulation first begins and forms collections of
12 somites along the neural tube, ADH3 activities are significantly lower (at 8-10 and 11-13
13 somite stages) and suggest a decreased ability to detoxify formaldehyde in the early embryo
14 (Harris et al., 2003). ADH mRNA expression levels appear to be age related, with decreased
15 expression of ADH common in premature neonates and infants up to 5 months old. Thereafter,
16 ADH expression increases and is dependent on body weight (Ginsberg et al., 2004). Benedetti et
17 al. (2007) reported that decreased ADH expression persisted until age 2 to 5 years. Westerlund
18 et al. (2005) tracked the ontogeny of ADH3 specifically and reported that ADH3 expression was
19 ubiquitous in mouse and rat embryos and was the only ADH enzyme to be consistently localized
20 to brain tissue, suggesting a housekeeping function. Thus, neonates and very young children
21 may have a decreased ability to metabolize formaldehyde due to differential expression of ADH3
22 in development compared that of with adults; however, activity levels of this enzyme and
23 alternate pathways specific to children are not available in the literature.
24
25 4.6.1.1.2. Life-stage exposure and adverse health outcomes.
26 In general, exposure to toxic agents during early development (i.e, preconception,
27 prenatal stages, or postnatal development) may affect organ development and may also lead to
28 increased disease susceptibility later in life. Following early life stage exposure to
29 formaldehyde, a number of adverse health outcomes have been observed, including alterations in
30 the respiratory, reproductive, and neurological systems. For example, the developing respiratory
31 tract may be more vulnerable to insult compared with an adult respiratory tract, and thus,
32 increase the severity of response. The potential for reproductive and developmental toxicity of
33 formaldehyde is discussed in detail in Sections 4.1.1.7 (human studies) and 4.2.7 (animal
7 For example, in the case of ozone concentrations of 0.1 ppm, a moderately reactive gas, Ginsberg et al. (2008)
predict a 5-fold variation in the dose to the deep lung between quiet and heavy breathing conditions for an 8-year old
child.
This document is a draft for review purposes only and does not constitute Agency policy.
4-539 DRAFT—DO NOT CITE OR QUOTE
-------
1 studies), while effects on the nervous system are discussed in Sections 4.1.1.6 and 4.2.6 (human
2 and animal studies, respectively). The specific case of formaldehyde exposure and pulmonary
3 effects is discussed in detail in Sections 4.1.1.1 to 4.1.1.4 and 4.2.1.1 to 4.2.1.4. A brief
4 summary of identified effects of formaldehyde that may indicate susceptibility during particular
5 life stages is provided below.
6
7 4.6.1.1.2.1. Preconception.
8 Exposure prior to conception may damage reproductive organs and/or germ cells that
9 could affect reproduction and/or damage the genetic makeup of the offspring. Effects on
10 reproduction are discussed in Sections 4.1.1.7 and 4.2.7. In summary, an epidemiological study
11 (Taskinen et al., 1999) reported significantly delayed conception among female workers exposed
12 to formaldehyde at average daily ambient formaldehyde levels; these effects could be consistent
13 with adverse effects on either preconceptional and/or gestational exposure. One animal study
14 (Maronpot et al., 1986) reported endometrial hypoplasia and lack of ovarian luteal tissue in
15 female mice exposed for 13 weeks to 40 ppm formaldehyde via inhalation, suggesting the
16 potential for treatment-related alterations to the female reproductive system. Since the exposure
17 was to the adult, the findings suggest that preconceptional formaldehyde exposure caused female
18 reproductive system effects that in turn could affect pregnancy. In the rodent study of Kitaev et
19 al. (1984), a three-fold increase in embryo degeneration on gestational days 2-3 was observed
20 after formaldehyde exposure to the dams during premating. Since the exposure was to the adult
21 in these three studies, the findings suggest that preconceptional formaldehyde exposure caused
22 female reproductive system effects and/or affected the gametes.
23
24 4.6.1.1.2.2. Prenatal
25 A population-based study (Grazuleviciene et al., 1998) found an association between
26 atmospheric formaldehyde exposure and low birth weight, yielding an adjusted OR of 1.37 (95%
27 CI: 0.90-2.09). Three studies (Dulskiene and Grazuleviciene, 2005; Taskinen et al., 1994;
28 Hemminki et al., 1985) that examined the effect of occupational exposures on the incidence of
29 congenital malformation produced mixed results.
30 Results from Taskinen et al. (1999) support associations between formaldehyde exposure,
31 subfertility, and spontaneous abortion. Subfertility and spontaneous abortion are biologically
32 linked (subclinical pregnancy losses are increased among women with fertility problems) (Gray
33 and Wu, 2000; Hakim et al., 1995), and both subfertility and spontaneous abortion may be
34 related to sensitivity to environmental agents (Correa et al., 1996).
This document is a draft for review purposes only and does not constitute Agency policy.
4-540 DRAFT—DO NOT CITE OR QUOTE
-------
1 Two experimental animal studies (Martin, 1990; Saillenfait et al., 1989) evaluated a
2 standard battery of developmental endpoints following inhalation exposure on GDs 6-10, but
3 effects were minimal. Similarly, Chernoff and Kavlock (1982), Marks et al. (1980), and Hurni
4 and Ohder (1973) reported minimal reproductive or developmental effects in rodents in studies
5 in which dams were exposed orally during early gestation. When formaldehyde was
6 administered via inhalation throughout gestation in female rats, some developmental effects,
7 including increased pup weight and decreases in lung and liver weight in newborns, were
8 reported at 0.01 and 0.4 ppm (Senichenkova and Chebotar, 1996; Senichenkova, 1991; Kitaev et
9 al., 1984; Gofmekler and Bonashevskaya, 1969; Gofmekler, 1968; Pushkina et al., 1968). Two
10 studies also reported changes in motor activity in offspring of dams exposed via inhalation to 0.4
11 ppm formaldehyde during gestation (Senichenkova, 1991; Sheveleva, 1971).
12
13 4.6.1.1.2.3. Postnatal
14 Following early life stage exposure to formaldehyde, a number of adverse postnatal
15 outcomes are possible, including effects on the developing and adult respiratory, reproductive,
16 and neurological systems. The potential for increased risk of childhood cancer is also discussed
17 below.
18
19 4.6.1.1.2.3.1. Respiratory toxicity. Formaldehyde is known to induce changes in
20 pulmonary function and cause pulmonary irritation in human studies (Rumchev et al., 2002;
21 Garrett et al., 1999; Krzyzanowski et al., 1990; Holmstrom et al., 1989; Holmstrom and
22 Wilhelmsson, 1988; Ritchie and Lehnen, 1987) and animal studies (Ohtsuka et al., 2003, 1997;
23 Riedel et al., 1996; Swiecichowski et al., 1993; Lee et al., 1984). Exposure to formaldehyde in
24 early life can cause damage to the lungs and permanently influence airway function, resulting in
25 increased vulnerability to toxicants later in life. Thus, young children may demonstrate
26 increased susceptibility to formaldehyde-related health effects. Krzyzanowski et al. (1990)
27 reported an association between physician-diagnosed asthma and chronic bronchitis in children
28 who lived in homes with formaldehyde levels that were higher than 60 ppb, after controlling for
29 socioeconomic status and ethnicity. Rumchev et al. (2002) reported a statistically significant
30 increased risk of asthma with increased residential concentrations of formaldehyde. Garrett et al.
31 (1999) found an increased association between bedroom concentration of formaldehyde and
32 increased risk of atopy in children. These studies suggest that formaldehyde exposure may
33 exacerbate responses in sensitive airways, particularly in children. Exacerbation of response has
34 also been noted in asthmatic adults and will be discussed below.
This document is a draft for review purposes only and does not constitute Agency policy.
4-541 DRAFT—DO NOT CITE OR QUOTE
-------
1 Another child-specific concern is that respiratory irritation may have greater impact on
2 lung function in children due to their smaller lung size; this is true even if the lung development
3 is normal. Irritation is commonly accompanied by inflammation, which can have a greater
4 impact on children's airways because they are narrower than adult airways. Thus, less
5 inflammation is required for significant airway obstruction in children than in adults.
6
7 4.6.1.1.2.3.2. Developmental neurotoxicity. In neonatal exposure paradigms, changes in
8 brain structure (Sarsilmaz et al., 2007; Asian et al., 2006), and brain chemistry (Songur et al.,
9 2008) were seen in young rats following inhalation exposures (6,000 or 12,000 ppb, 5 days per
10 week from postnatal day 0-30). In addition, Weiler and Apfelbach (1992) found juvenile
11 animals to be more sensitive to formaldehyde-induced changes in olfactory thresholds when
12 compared with adult animals (shifts in olfactory thresholds were greater when exposure [at
13 250 ppb] was initiated at PND 30 than at adult ages). These studies are consistent with the
14 hypothesis that early life exposure to formaldehyde can lead to long-lasting neurological effects.
15 Exposure levels in these studies (250-6,000 ppb) were in the same range as those producing the
16 behavioral effects in adults (as low as 100 ppb), but provide limited information regarding
17 relative sensitivity as no NOAELs were identified, and (with the exception of Weiler and
18 Apfelbach), similar parameters were not measured in adult animals using the same exposure
19 paradigms.
20
21 4.6.1.2. Later Life Stages
22 In general, older adults may be at risk for increased susceptibility to exposure to
23 environmental chemicals by virtue of their slower metabolism and increased incidence of altered
24 health status (Benedetti et al., 2007; Ginsberg et al., 2005; U.S. EPA, 2005a). Additionally,
25 adverse effects of earlier exposure to some toxicants may be observed in older adults as a result
26 of latency in expression of the effect (Olsen et al., 1997; Sweeney et al., 1986). No studies have
27 examined the differential effects of formaldehyde exposure for elderly adults (>65 years old) as
28 compared to other age groups.
29
30 4.6.1.3. Conclusions on Life-Stage Susceptibility
31 In summary, timing both of the exposure and of the assessment of health outcomes may
32 be important for understanding the relative risk of adverse effects from formaldehyde exposure
33 during different life stages. There are known developmental differences in kinetics across life
34 stages, including differences in enzymes involved in formaldehyde metabolism, but the
35 contribution of these differences to formaldehyde-related health effects is unknown. Similarly,
This document is a draft for review purposes only and does not constitute Agency policy.
4-542 DRAFT—DO NOT CITE OR QUOTE
-------
1 information regarding life-stage differences in respiratory physiology raises possible concern
2 regarding increased exposure to children, but studies for formaldehyde are not available.
3 Available data do support an increased risk for adverse effects on lung function in children. The
4 overall body of evidence shows some support, although minimal, for susceptibility in
5 reproductive or developmental endpoints associated with exposure to formaldehyde. Some
6 studies observed altered development of the nervous system following formaldehyde exposure
7 during early life. Older adults may be at risk for increased susceptibility to formaldehyde
8 because of slower metabolism and clearance rates. Elderly adults have an increased probability
9 of having both altered health status and altered metabolism, which could impact their ability to
10 process and recover from an adverse effect. The available data are consistent with some life-
11 stage susceptibility differences for formaldehyde at the level of TD or TK differences, the results
12 are nonetheless inconclusive due to the number of data gaps.
13
14 4.6.2. Health/Disease Status
15 The factor for which we have the greatest evidence is pre-existing disease, and
16 specifically asthma. Numerous studies have assessed the potential for increased susceptibility to
17 formaldehyde in asthmatics. Formaldehyde does not induce airway hyperreactivity directly
18 (Sheppard et al., 1984) and has not been shown to increase airway hyperreactivity in either
19 asthmatics or nonasthmatics (Pazdrak et al., 1993; Harving et al., 1991; Kulle et al., 1987).
20 Significantly decreased forced expiratory volume (FEVi) measurements were reported among
21 asthmatics in two studies (Casset et al., 2006; Green et al., 1987), while others did not find any
22 significant change in FEVi following formaldehyde exposure (Ezratty et al., 2007; Frigas et al.,
23 1984).
24 A few available case reports of bronchial asthma do suggest direct respiratory tract
25 sensitization to formaldehyde gas (Lemiere et al., 1995; Burge et al., 1985; Hendrick et al., 1982;
26 Hendrick and Lane, 1977, 1975). All cases displayed marked changes in FEVi or pulmonary
27 airflow rate in response to acute challenges with formaldehyde gas at exposure levels <3 ppm.
28 Formaldehyde-induced IgE production has been reported in some studies (Vandenplas et al.,
29 2004; Wantke et al., 1996a).
30 There is a large quantity of human data providing evidence of an association between
31 formaldehyde exposure and increased incidence of asthma or exacerbation of asthmatic
32 responses in compromised individuals. For example, Krzyzanowski et al. (1990) reported an
33 association between physician-diagnosed asthma and chronic bronchitis in children who lived in
34 homes with formaldehyde levels that were higher than 60 ppb, after controlling for
35 socioeconomic status and ethnicity. Rumchev et al. (2002) reported a statistically significant
This document is a draft for review purposes only and does not constitute Agency policy.
4-543 DRAFT—DO NOT CITE OR QUOTE
-------
1 increased risk of asthma with increased residential concentrations of formaldehyde. Garrett et al.
2 (1999) found an increased association between bedroom concentration of formaldehyde and risk
3 of atopy in children. These studies suggest that formaldehyde exposure may exacerbate
4 responses in sensitive airways, particularly in children. Exacerbation of response has also been
5 noted in adults. Kriebel et al. (1993) reported a greater decrease in peak expiratory flow (PEF)
6 in asthmatic medical students (7.3% decrement) compared with nonasthmatic medical students
7 (2% decrement) after 2 weeks exposure to formaldehyde (average concentration 0.73 ppm) in an
8 anatomy lab. This effect does not appear to be immunogenic in nature (Fujimaki et al., 2004a;
9 Lee etal., 1984).
10 Several animal studies document increased airway resistance and bronchial constriction
11 following inhalation exposure to formaldehyde (Nielson et al., 1999; Swiecichowski et al., 1993;
12 Biagini et al., 1989; Amdur et al., 1960). Sadakane et al. (2002) demonstrated that formaldehyde
13 exposure exacerbated sensitization and challenge with a common dust mite allergen (Der f) as
14 measured by increased eosinophil infiltration into the interstitium around the bronchi and
15 bronchioles as well as goblet cell proliferation in the bronchial epithelium; they suggested that
16 formaldehyde exposure may aggravate eosinophilic infiltration and goblet cell proliferation that
17 accompanies allergic responses. The MOA underlying this response is unknown. These
18 decrements may occur indirectly in response to formaldehyde and may be mediated via
19 neurogenic potentiation (Sadakane et al., 2002; Riedel et al., 1996; Tarkowski and Gorski, 1995).
20 In particular, Tarkowski and Gorski (1995) suggest that formaldehyde may increase permeability
21 of respiratory epithelium and destruction of immunologic barriers. Thus, the respiratory tract
22 may become vulnerable to inhaled allergens after formaldehyde exposure (Tarkowski and
23 Gorski, 1995).
24 In summary, the data indicate that formaldehyde exposure can aggravate a type I
25 hypersensitivity response and that this hypersensitivity may in turn increase the susceptibility to
26 formaldehyde exposure in these individuals. Formaldehyde exposure may predetermine an
27 asthmatic phenotype or may induce new incidences of asthma via indirect mechanisms, though
28 definitive evidence and a proposed mechanism remain to be determined. Individuals that exhibit
29 chemically induced sensitivity and are exposed acutely or chronically to formaldehyde in
30 residential and occupational settings might exhibit adverse responses at lower concentrations of
31 formaldehyde than the average healthy person.
32
33 4.6.3. Nutritional Status
34 Limited available data indicate that certain types of malnutrition may increase
35 susceptibility to formaldehyde exposure. Senichenkova and Chebotar (1996) reported increased
This document is a draft for review purposes only and does not constitute Agency policy.
4.544 DRAFT—DO NOT CITE OR QUOTE
-------
1 fetal anomalies in fetuses from iron-deficient pregnant mice after formaldehyde exposure
2 compared with anemic mice that had not been exposed to formaldehyde. Forced iron reduction
3 (induced by addition of bipyridyl treatment in pregnant mice) in utero increased the overall
4 incidence of fetal anomalies when paired with formaldehyde exposure (Senichenkova and
5 Chebotar, 1996). The findings are difficult to evaluate due to poor reporting and have not been
6 substantiated by other laboratories.
7
8 4.6.4. Gender Differences
9 Males and females can differ greatly in body composition, organ function, and many
10 other physiological parameters that may influence the toxicokinetics of chemicals and their
11 metabolites in the body (Gochfeld, 2007; Gandhi et al., 2004).
12 The human epidemiology data set does not support any specific gender susceptibilities
13 for noncancer effects due to formaldehyde exposure. In general, data suggest that nonpregnant
14 women, on a per kg body weight basis, may have slightly lower air intake than men, which
15 would suggest that women may be less susceptible to inhaled pollutants like formaldehyde than
16 men, but this has not been investigated in the available formaldehyde literature.
17 A few isolated reports have investigated potential gender differences in development of
18 nasal pharyngeal carcinomas following exposure to formaldehyde. One case-control study
19 identified a higher OR for sinonasal adenocarcinomas in women (OR = 6.2 [95% CI: 2.2-19.7])
20 compared with the OR observed in men (OR = 3.0 [95% CI: 1.5-5.7]) following exposure to
21 formaldehyde (Luce et al., 2002). However, the overall body of evidence remains scant.
22 There are a few reports concerning differential formaldehyde-induced effects on the male
23 and female reproductive systems. Ozen et al. (2002), Sarsilmaz et al. (1999), and Woutersen et
24 al. (1987) reported reduced Ley dig cell numbers in adult male rats exposed by inhalation. In
25 female mice, inhalation exposure to formaldehyde resulted in endometrial hypoplasia and lack of
26 ovarian luteal tissue (Maronpot et al., 1986). The clinical significance of these effects in humans
27 is unknown, and due to limited data it is unclear whether the female or male reproductive system
28 is more susceptible to perturbation by formaldehyde.
29
30 4.6.5. Genetic Differences
31 There are some data for polymorphisms in humans that affect formaldehyde TK. As
32 discussed in Section 3.4, the primary metabolizing enzymes of formaldehyde are ALDH2 and
33 ADH3, with the latter enzyme considered more relevant to low exposures. Polymorphisms in
34 ALDH2 have been shown to have implications in human risk assessment, specifically in regard
35 to acetaldehyde metabolism (Ginsberg et al., 2002). Teng et al. (2001) demonstrated the
This document is a draft for review purposes only and does not constitute Agency policy.
4-545 DRAFT—DO NOT CITE OR QUOTE
-------
1 importance of ALDH2 for formaldehyde metabolism in rat hepatocytes at fairly high
2 formaldehyde concentrations (2.5 mM and greater). Cheng et al. (2008) investigated the
3 relationship between occupational exposure to formaldehyde and genetic polymorphisms of
4 ALDH2 and CYP2E1. There was a positive relationship between the concentration of formic
5 acid in the urine and ALDH2 genotypes (%2 = 9.241, p < 0.05). Urinary formic acid
6 concentration may be affected by formaldehyde exposure concentration and ALDH2 genotype
7 (Cheng et al., 2008) for individuals that have high exposure levels. Thus, although ALDH2 may
8 not be involved in formaldehyde metabolism if exposure levels are low, polymorphisms of this
9 enzyme may lead to differences in response at higher exposure levels.
10 Wu et al. (2007) looked for and identified two SNPs in ADH3 among a population of
11 Mexican asthmatic children 4 to 17 years of age. Carrying one or two copies of the minor allele
12 for one SNP resulted in a decreased RR of asthma (RR = 0.66-0.77). For the second SNP,
13 homozygotes for the minor allele had an RR of 1.6 for asthma. The functional characteristics of
14 these SNPs are unknown. Studies evaluating whether any of the polymorphisms affect the
15 expression, regulation, stability, or activity of the enzyme in vivo are lacking; therefore, the
16 relative susceptibility of individuals with different polymorphisms cannot be characterized at this
17 time.
18 One study (Hedberg et al., 2001) identified three polymorphisms in human ADH3
19 involving four base-pair substitutions in the promoter region of which one (C—»T) showed
20 reduced activity (-50-70% of control). Hedberg et al. (2001) reported differences in allele
21 frequencies among Chinese, Spanish, and Swedish groups, consisting of Asian-Caucasian
22 differences and ethnic subgroups among Caucasians. Their results suggest that a small
23 percentage of Caucasians may have decreased ADH3 expression and thus, be more susceptible to
24 formaldehyde exposure. Additional studies to validate these findings have not been performed.
25 The relative activity level of these enzymes may also impact the metabolism of
26 formaldehyde. In pharmacokinetic studies, deletion of ADH3 increased the sensitivity of mice to
27 formaldehyde (Deltour et al., 1999) and was deleterious to yeast (Achkor et al., 2003). These
28 results suggest that deficiencies in ADH3 may confer an increased susceptibility to formaldehyde
29 toxicity (Teng et al., 2001). The importance of properly functioning enzymes also suggests that
30 genetic differences in ADH3 or ALDH2 may affect the response to formaldehyde exposure.
31 However, comparable human data are not available.
32 Race/ethnicity may be a surrogate for genetic differences but racial or ethnic groups may
33 also reflect socioeconomic, and/or cultural factors that are distinct from genetics. Possible ethnic
34 differences may be related to genetic polymorphisms of enzymes ALDH2 and ADH3, relevant
35 for formaldehyde metabolism. ALDH2 variants, present primarily in East Asians, are known to
This document is a draft for review purposes only and does not constitute Agency policy.
4-546 DRAFT—DO NOT CITE OR QUOTE
-------
1 have protective effects against alcoholism but were not found in the people of Indo-Trinidadian
2 descent (Moore et al., 2007) or in American Indians or Alaska natives (Ehlers, 2007). However,
3 there is no direct evidence to associate these variants to differential susceptibility to
4 formaldehyde exposure, nor is there direct evidence to associate these ethnic groups specifically
5 with differential susceptibility to formaldehyde. Further, no studies have specifically assessed
6 ethnic variability in responses to formaldehyde.
7 There are complex pathways through which genetic polymorphisms in ADH3 can
8 potentially affect differential susceptibility to formaldehyde. Firstly, ADH3 is central to the
9 metabolism of formaldehyde. However, ADH3 itself may indirectly contribute to the adverse
10 effects of formaldehyde on pulmonary physiology (Thompson et al., 2009; Staab et al., 2008a, b;
11 Thompson and Grafstrom, 2008). Exposure to formaldehyde is itself thought to alter the activity
12 of ADH3 resulting in the perturbation of critical metabolic pathways. ADH3 participates in the
13 oxidation of retinol and long-chain primary alcohols, as well as the reduction of S-
14 nitrosoglutathione (GSNO). The activity of ADH3 toward some of these substrates has been
15 shown to be significantly increased in the presence of formaldehyde. ADH3 has recently also
16 been shown to contribute to NO signaling through its dual role in metabolizing GSNO, an
17 endogenous bronchodilator and reservoir of NO (Staab et al., 2008a; Hess et al., 2005; Jensen et
18 al. 1998). Through its regulatory function on GSNO, ADH3 may thus play a central role in
19 regulating bronchial tone allergen-induced hyperresponsiveness (Gerard, 2005; Que et al., 2005).
20 As concluded by California Environmental Protection Agency (CalEPA) (2008), "the
21 dysregulation of NO by formaldehyde [in this manner] helps to explain the variety and
22 variability in the toxic manifestations following formaldehyde inhalation."
23
24 4.6.6. Coexposures
25 4.6.6.1. Cumulative Risk
26 When considering health risks, it is important to consider the impact of coexposures to
27 other agents that may interact with the chemical under evaluation. Coexposure to other
28 pollutants, particularly those that produce some of the same metabolites and similar health
29 effects as formaldehyde, is likely to occur in both occupational and nonoccupational settings.
30 Due to effects on metabolic enzymes (inducing and/or inhibition) as well as direct effects
31 on organ system function, coexposures may alter the way in which formaldehyde is metabolized
32 and cleared from the body. Inhibition or induction of the enzymes responsible for metabolism of
33 chemicals may alter susceptibility to toxicity (Lash and Parker, 2001; IARC, 1995; U.S. EPA,
34 1985a). Smokers may be at increased risk for effects of formaldehyde exposure, because
35 formaldehyde is one of the components of cigarette smoke and is likely to heighten the point-of-
This document is a draft for review purposes only and does not constitute Agency policy.
4-547 DRAFT—DO NOT CITE OR QUOTE
-------
1 entry effect when combined with occupational or residential exposures to inhaled formaldehyde.
2 However, no evidence is available to evaluate the potential aggregate effects.
O
4 4.6.6.2. Aggregate Exposure
5 In addition, multiple routes of exposure to a single agent may increase the cumulative
6 risk by increasing the overall body burden of the chemical. A human aggregate exposure model
7 developed by McKone and Daniels (1991) incorporated likely exposures from air, water, and soil
8 media through inhalation, ingestion, and dermal contact. The authors hypothesized that the
9 aggregate exposure could be age-dependent but did not present any data for persons of differing
10 life stages. The role of multiple exposures on different genders, genetic susceptibility, or altered
11 health and nutrition status has not been investigated. The available database regarding the
12 potential for multiple routes of exposure (or aggregate exposure) formaldehyde is limited.
13 Guseva (1972) specifically assessed the reproductive and developmental effects caused
14 by coexposure to formaldehyde via both inhalation (0.25 mg/m3) and ingestion (0.01 mg/L)
15 routes in male rats. The authors reported reduced nucleic acid levels in testes to 88 and 92% of
16 controls, which suggests a possible toxic gonadotropic effect. The ability of male rats (receiving
17 combined exposure to formaldehyde at a low concentration level for a long period of time) to
18 reproduce was preserved since all the cohabited females were impregnated. The number and
19 weight of the fetuses and newborn rat pups in the experimental coexposure groups did not differ
20 substantially from those figures observed in the control group. No developmental defects or
21 anomalies were observed in the offspring for up to 1 month postnatally. Thus, at low exposures,
22 the reproductive effects due to combined ingestion and inhalation exposure are unknown.
23
24 4.6.7. Uncertainties of Database
25 There is a need to better characterize the implications of formaldehyde exposures to
26 susceptible populations. A number of areas where the database is currently insufficient are
27 identified below.
28
29 4.6.7.1. Uncertainties of Exposure
30 Although information exists on early life exposure to formaldehyde, a number of
31 uncertainties regarding children's susceptibility remain. First, inhalation is believed to be of
32 most concern for formaldehyde, since formaldehyde vapors are released from insulation or from
33 ambient sources of formaldehyde, including secondary production from other pollutants involved
34 in photo-oxidant reactions. Any additional pathways of exposure for children have not been
35 characterized. Since formaldehyde is nearly ubiquitous in the environment, it is difficult to
This document is a draft for review purposes only and does not constitute Agency policy.
4-548 DRAFT—DO NOT CITE OR QUOTE
-------
1 quantify the total exposure. Second, children have different respiratory, metabolic, and activity
2 rates compared with healthy adults, potentially influencing ADME and target tissue exposure to
3 formaldehyde. However, studies to identify the specific changes in absorption of formaldehyde
4 and its metabolites across developmental stages and across organs have not been performed. In
5 addition, exposure prenatally may be altered based on whether formaldehyde or its metabolites
6 pass through the placenta, but placental transfer data are not available. Third, no quantitative
7 models have been developed to characterize these differences for formaldehyde. Formaldehyde-
8 specific PBPK models and their validation will aid in understanding the uncertainties associated
9 with formaldehyde exposure in children.
10 Given the large proportion of time that most individuals in the U.S. spend indoors,
11 exposure scenarios where indoor concentrations to formaldehyde are high (e.g., in homes or in
12 trailers; see Section 2.3.1) may play a significant role and may be of particular concern to the
13 elderly or health-impaired individuals who spend relatively more time at home. Further
14 evaluation of the effects of coexposures and pathways of exposure and aggregate risk is needed.
15 An estimate of the multiple exposure pathways is needed to know where along the dose-response
16 curve to place an incremental exposure to formaldehyde.
17
18 4.6.7.2. Uncertainties of Effect
19 Studies specifically designed to evaluate effects after early and later life stage exposure
20 are needed in order to more fully characterize potential life-stage-related differences in
21 formaldehyde toxicity, including the defining of critical windows during development. For
22 example, life-stage-specific neurotoxic and pulmonary effects, particularly in the developing
23 fetus, need further evaluation. The preconceptional period may be a critical window for
24 formaldehyde exposure and reproductive and developmental effects, based on rodent studies of
25 reproductive, embryonic and gamete effects. Data specific to the carcinogenic effects of
26 formaldehyde exposure during early life stages do not exist. The reduction in fertility seen in
27 some studies (Gray and Wu, 2000; Taskinen et al., 1999; Hakim et al., 1995) is not adequately
28 described and a well-established MOA has not been identified, but some have been hypothesized
29 including altered sperm quality (Ozen et al., 2002; Sarsilmaz et al., 1999; Woutersen et al.,
30 1987). Further, spontaneous abortion/fetal loss occurring early in gestation, prior to maternal
31 knowledge of the pregnancy, can lead to misclassification of the effect as infertility (see Sections
32 4.1.1.7 and 4.2.7).
33 More research is needed to clarify the role of genetic polymorphisms in formaldehyde
34 metabolism. Similarly, data gaps pertaining to gender differences remain. A potential impact of
35 nutritional status and iron deficiency on formaldehyde toxicity needs further investigation.
This document is a draft for review purposes only and does not constitute Agency policy.
4-549 DRAFT—DO NOT CITE OR QUOTE
-------
1 A fair body of evidence suggests that asthmatics are more susceptible to formaldehyde exposure
2 than the general population, however, the mechanism of action for this increased susceptibility is
3 unknown.
4 In the studies discussed above, there are a number of examples of studies that assessed
5 multiple susceptibility factors that are worth noting. For example, the Krzyzanowski et al.
6 (1990) study reported asthma and chronic bronchitis cases for two interacting potential
7 susceptible groups, in children and those with high exposure (due to living in homes with
8 formaldehyde levels that were higher than 60 ppb). Similarly, the Garrett et al. (1999) study
9 assessed the same two interacting potential susceptible groups.
10 The study of Senichenkova and Chebotar (1996) assessed developmental effects in
11 mouse fetuses after in utero iron-deficiency and formaldehyde exposure. Thus, the study
12 findings must be considered in light of possible interactions between life stage exposure
13 differences and nutritional status differences.
14 Studies to understand the nature of the interactions between the various susceptibility
15 factors for formaldehyde have not been performed.
16
17 4.6.8. Summary of Potential Susceptibility
18 There is some evidence to demonstrate susceptibility for various populations exposed to
19 formaldehyde. Available data are summarized in Table 4-95 where formaldehyde susceptibility
20 factors are presented by those with data for increased formaldehyde susceptibility and those with
21 data for differences but with an unknown impact on formaldehyde susceptibility.
22 Exposure to formaldehyde during early developmental and later life stages may be of
23 concern. However, human exposure to the developing fetus is unknown since it is not known
24 whether formaldehyde or one of its metabolites crosses the placenta. However, there is very
25 limited life-stage-specific information regarding the TK of formaldehyde. Life-stage-specific
26 TK has not been characterized, and, thus, no PBPK models exist to effectively evaluate the risk
27 to early life stages. Children may be more susceptible to noncancer health effects as a result of
28 inhalation exposure to formaldehyde due to increased respiratory rates. There are no studies to
29 evaluate whether formaldehyde exposure in early life (e.g., pregnancy) is associated with an
30 increased risk of childhood cancer.
31 The weight of evidence supports a plausible association between formaldehyde exposure
32 and aggravated asthmatic responses in humans and this association is corroborated by limited
33 evidence from animal studies. Formaldehyde does not appear to directly induce airway
34 hyperreactivity but may sensitize airways to subsequent exposures. One issue in interpreting the
35 available studies that assessed the relationship between asthma and formaldehyde could not
36
This document is a draft for review purposes only and does not constitute Agency policy.
4-550 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 4-95. Available evidence for susceptibility factors of concern for
formaldehyde exposure
Factor
Evidence that factors increase
susceptibility to formaldehyde
Evidence that factors show differences but
unknown impact on susceptibility
Life stage
Preconception
Prenatal
Postnatal
Developmental effects reported
suggesting that critical windows of
exposure may be relevant:
• Reproductive outcomes (Taskinen et
al., 1999; Maronpot et al., 1986)
• Embryo effects (Kitaev et al., 1984)
• Structural- and functional
developmental outcomes (Martin, 1990;
Saillenfait et al., 1989; Sheveleva,
1971; Seninchenkova, 1991)
Lung function outcome (Krzyzanowski
et al., 1990; Rumchev et al., 2002;
Garrett et al., 1999)
Developmental neurotoxicity (Weiler
and Apfelbach, 1992)
Possible life stage level differences in
some enzymes involved in formaldehyde
metabolism (Harris et al., 2003; Ginsberg
et al., 2004; Westerlund et al., 2005;
Benedetti et al., 2007)
Mixed reports of associations between
prenatal exposure and developmental
outcomes in human studies (positive
association: Grazuleviciene et al., 1998)
Possible life stage level differences in
some enzymes involved in formaldehyde
metabolism (e.g., J.ADH expression over
first 5 months; Ginsberg et al., 2004)
Developmental neurotoxicity (Sarsilmaz
et al., 2007; Asian et al., 2006; Songur et
al., 2008)
Disease status
Bronchial asthma (Lemiere et al., 1995;
Burge et al., 1985; Hendrick et al.,
1982; Hendrick and Lane, 1977, 1975)
Increased airway resistance and
bronchial constriction (Nielson et al.,
1999; Swiecichowski et al., 1993;
Biagini et al., 1989; Amdur et al., 1960)
Mixed results for forced expiratory
volume (FEV1) measures affected by
formaldehyde exposure in asthmatics
(Casset et al., 2006; Green et al., 1987;
Ezratty et al., 2007; Frigas et al., 1984)
Nutritional status/diet
Iron-deficiency in utero (Senichenkova
and Chebotar, 1996).
Genetics
• Polymorphisms
For high formaldehyde exposure:
Urinary formic acid levels affected by
ALDH2 genotype (Cheng et al., 2008)
In mice, ADH3 increased sensitivity to
formaldehyde (Achkor et al., 2003)
Differences among ADH3 alleles and
asthma outcome (Wu et al., 2007)
Differences among ethnic groups in
ADH3 alleles (Hedberg et al., 2001)
Gender
Gender differences in incidence of
nasopharyngeal carcinoma following
formaldehyde exposure (Luce et al., 2002)
4
5
6
This document is a draft for review purposes only and does not constitute Agency policy.
4-551 DRAFT—DO NOT CITE OR QUOTE
-------
1 distinguish between the cases of asthma that were due to earlier formaldehyde exposure vs. those
2 without a direct link to formaldehyde exposure.
3 No direct link exists between formaldehyde exposure and differential susceptibility in
4 different ethnic groups, although genetic polymorphisms in the enzymes involved with
5 formaldehyde metabolism, ADH3 and ALDH2, provide some support for differential
6 susceptibility to alcoholism in a number of ethnic groups. The evidence for differential gender
7 responses to formaldehyde exposure is equivocal. Coexposures may result in altered metabolism
8 and clearance, but there is no evidence that coexposures are a critical part of formaldehyde-
9 mediated differential susceptibility.
10 Thus, given the available data, increased susceptibility to adverse effects of
11 formaldehyde is most strongly supported for three populations: (1) Preconception and perinatal
12 exposure based on reproductive and developmental effects; (2) children, whose exposure may be
13 higher by virtue of their increased activity level and respiratory rate; and (3) asthmatics who may
14 exhibit exacerbation of response to formaldehyde.
-End of Volume II -
This document is a draft for review purposes only and does not constitute Agency policy.
4-552 DRAFT—DO NOT CITE OR QUOTE
-------
DRAFT-DO NOT CITE OR QUOTE EPA/635/R-10/002A
www.epa.gov/iris
f/EPA
TOXICOLOGICAL REVIEW OF
FORMALDEHYDE -
INHALATION ASSESSMENT
(CAS No. 50-00-0)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
VOLUME III of IV
Quantitative Assessment, Major Conclusions in
the Characterization of Hazard and Dose
Response, and References
June 2, 2010
NOTICE
This document is an External 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.
IH-ii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS IN BRIEF
VOLUME I: INTRODUCTION, BACKGROUND, AND TOXICOKINETICS
Chapter 1: Introduction
Chapter 2: B ackground
Chapter 3: Toxicokinetics
VOLUME II: HAZARD CHARACTERIZATION
Chapter 4: Hazard Characterization
VOLUME III: QUANTITATIVE ASSESSMENT, MAJOR CONCLUSIONS IN THE
CHARACTERIZATION OF HAZARD AND DOSE RESPONSE, AND
REFERENCES
Chapter 5: Quantitative Assessment: Inhalation Exposure
Chapter 6: Major Conclusions in the Characterization of Hazard and Dose-Response
References
VOLUME IV: APPENDICES
Appendix A: Summary of External Peer Review and Public Comments and Dispositions
Appendix B: Simulations of Interindividual and Adult-to-Child Variability in Reactive
Gas Uptake in a Small Sample of People (Garcia et al., 2009)
Appendix C: Lifetable Analysis
Appendix D: Model Structure & Calibration in Conolly et al. (2003, 2004)
Appendix E: Evaluation of BBDR Modeling of Nasal Cancer in the F344 Rat: Conolly
et al. (2003) and Alternative Implementations
Appendix F: Sensitivity Analysis of BBDR Model for Formaldehyde Induced
Respiratory Cancer in Humans
Appendix G: Evaluation of the Cancer Dose-Response Modeling of Genomic Data for
Formaldehyde Risk Assessment
Appendix H: Expert Panel Consultation on Quantitative Evaluation of Animal
Toxicology Data for Analyzing Cancer Risk Due to Inhaled Formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
IH-iii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS—TOXICOLOGICAL REVIEW OF FORMALDEHYDE
(CAS No. 50-00-0)
LIST OF TABLES Ill-xi
LIST OF FIGURES III-xx
1 5. QUANTITATIVE ASSESSMENT: Inhalation Exposure 5-1
2 5.1. INHALATION REFERENCE CONCENTRATION (RfC) 5-2
3 5.1.1. Candidate Critical Effects by Health Effect Category 5-3
4 5.1.1.1. Sensory Irritation of the Eyes, Nose, and Throat 5-3
5 5.1.1.2. Upper Respiratory Tract Pathology 5-5
6 5.1.1.3. Pulmonary Function Effects 5-6
7 5.1.1.4. Asthma and Allergic Sensitization (Atopy) 5-10
8 5.1.1.5. Immune Function 5-16
9 5.1.1.6. Neurological and Behavioral Toxicity 5-17
10 5.1.1.7. Developmental and Reproductive Toxicity 5-26
11 5.1.2. Summary of Critical Effects and Candidate RfCs 5-34
12 5.1.2.1. Selection of Studies for Candidate RfC Derivation 5-34
13 5.1.2.2. Derivation of Candidate RfCs from Key Studies 5-40
14 5.1.2.3. Evaluation of the Study-Specific Candidate RfCs 5-66
15 5.1.3. Database Uncertainties in the RfC Derivation 5-70
16 5.1.4. Uncertainties in the RfC Derivation 5-72
17 5.1.4.1. Point of Departure 5-73
18 5.1.4.2. Extrapolation from Laboratory Animal Data to Humans 5-73
19 5.1.4.3. Human Variation 5-74
20 5.1.4.4. Subchronic-to-Chronic Extrapolation 5-74
21 5.1.5. Previous Inhalation Assessment 5-75
22 5.2. QUANTITATIVE CANCER ASSESSMENT BASED ON THE NATIONAL
23 CANCER INSTITUTE COHORT STUDY 5-75
24 5.2.1. Choice of Epidemiology Study 5-76
25 5.2.2. Nasopharyngeal Cancer 5-78
26 5.2.2.1. Exposure-Response Modeling of the National Cancer Institute
27 Cohort 5-78
28 5.2.2.2. Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer
29 Mortality 5-81
30 5.2.2.3. Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer
31 Incidence 5-83
32 5.2.2.4. Sources of Uncertainty 5-85
33 5.2.3. Lymphohematopoietic Cancer 5-90
34 5.2.3.1. Exposure-Response Modeling of the National Cancer Institute
35 Cohort 5-90
36 5.2.3.2. Prediction of Lifetime Extra Risks for Hodgkin Lymphoma and
37 Leukemia Mortality 5-93
38
This document is a draft for review purposes only and does not constitute Agency policy.
IH-iv DRAFT—DO NOT CITE OR QUOTE
-------
1 CONTENTS (continued)
2
3
4 5.2.3.3. Prediction of Lifetime Extra Risks for Hodgkin Lymphoma and
5 Leukemia Incidence 5-96
6 5.2.3.4. Sources of Uncertainty 5-98
7 5.2.4. Conclusions on Cancer Unit Risk Estimates Based on Human Data 5-102
8 5.3. DOSE-RESPONSE MODELING OF RISK OF SQUAMOUS CELL
9 CARCINOMA IN THE RESPIRATORY TRACT USING ANIMAL DATA 5-104
10 5.3.1. Long-Term Bioassays in Laboratory Animals 5-108
11 5.3.1.1. Nasal Tumor Incidence Data 5-108
12 5.3.1.2. Mechanistic Data 5-110
13 5.3.2. The CUT Biologically Based Dose-Response Modeling 5-111
14 5.3.2.1. Major Results of the CUT Modeling Effort 5-115
15 5.3.3. This Assessment's Conclusions from Evaluation of Dose-Response
16 Models of DPX, Cell-Replication and Genomics Data, and of BBDR
17 Models for Risk Estimation 5-115
18 5.3.4. Benchmark Dose Approaches to Rat Nasal Tumor Data 5-122
19 5.3.4.1. Benchmark Dose Derived from BBDR Rat Model and Flux as
20 Dosimeter 5-122
21 5.3.4.2. Comparison with Other Benchmark Dose Modeling Efforts 5-129
22 5.3.4.3. Kaplan-Meier Adjustment 5-132
23 5.3.4.4. EPA Time-to-Tumor Statistical Modeling 5-132
24 5.4. CONCLUSIONS FROM THE QUANTITATIVE ASSESSMENT OF CANCER
25 RISK FROM FORMALDEHYDE EXPOSURE BY INHALATION 5-138
26 5.4.1. Inhalation Unit Risk Estimates Based on Human Data 5-138
27 5.4.2. Inhalation Unit Risk Estimates Based on Rodent Data 5-138
28 5.4.3. Summary of Inhalation Unit Risk Estimates 5-139
29 5.4.4. Application of Age-Dependent Adjustment Factors (ADAFs) 5-140
30 5.4.5. Conclusions: Cancer Inhalation Unit Risk Estimates 5-142
31
32 6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND
33 DOSE-RESPONSE 6-1
34 6.1. SUMMARY OF HUMAN HAZARD POTENTIAL 6-1
35 6.1.1. Exposure 6-1
36 6.1.2. Absorption, Distribution, Metabolism, and Excretion 6-1
37 6.1.3. Noncancer Health Effects in Humans and Laboratory Animals 6-4
38 6.1.3.1. Sensory Irritation 6-4
39 6.1.3.2. Respiratory Tract Pathology 6-6
40 6.1.3.3. Effects on Pulmonary Function 6-8
41 6.1.3.4. Asthmatic Responses and Increased Atopic Symptoms 6-9
42 6.1.3.5. Effects on the Immune System 6-11
43 6.1.3.6. Neurological Effects 6-11
44 6.1.3.7. Reproductive and Developmental Effects 6-12
45 6.1.3.8. Effects on General Systemic Toxicity 6-13
This document is a draft for review purposes only and does not constitute Agency policy.
III-v DRAFT—DO NOT CITE OR QUOTE
-------
1 CONTENTS (continued)
2
3
4 6.1.3.9. Summary 6-14
5 6.1.4. Carcinogen!city in Humans and Laboratory Animals 6-14
6 6.1.4.1. Carcinogenicity in Humans 6-14
7 6.1.4.2. Carcinogenicity in Laboratory Animals 6-21
8 6.1.4.3. Carcinogenic Mode(s) of Action 6-21
9 6.1.5. Cancer Hazard Characterization 6-25
10 6.2. DOSE-RESPONSE CHARACTERIZATION 6-25
11 6.2.1. Noncancer Toxicity: Reference Concentration (RfC) 6-25
12 6.2.1.1. Assessment Approach Employed 6-25
13 6.2.1.2. Derivation of Candidate Reference Concentrations 6-26
14 6.2.1.3. Adequacy of Overall Data Base for RfC Derivation 6-27
15 6.2.1.4. Uncertainties in the Reference Concentration (RfC) 6-29
16 6.2.1.5. Conclusions 6-32
17 6.2.2. Cancer Risk Estimates 6-33
18 6.2.2.1. Choice of Data 6-33
19 6.2.2.2. Analysis of Epidemiologic Data 6-34
20 6.2.2.3. Analysis of Laboratory Animal Data 6-36
21 6.2.2.4. Extrapolation Approaches 6-37
22 6.2.2.5. Inhalation Unit Risk Estimates for Cancer 6-41
23 6.2.2.6. Early-Life Susceptibility 6-41
24 6.2.2.7'. Uncertainties in the Quantitative Risk Estimates 6-42
25 6.2.2.8. Conclusions 6-44
26 6.3. SUMMARY AND CONCLUSIONS 6-45
REFERENCES R-l
This document is a draft for review purposes only and does not constitute Agency policy.
IH-vi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES
1 Table 5-1. Candidate points of departure (PODs) including duration adjustments for
2 nervous system toxicity in key human and animal studies 5-18
3 Table 5-2. Effects of formaldehyde exposure on completion of the labyrinth test by
4 male and female LEW. IK rats 5-24
5 Table 5-3. Candidate PODs including duration adjustments for developmental and
6 reproductive toxicity in key animal studies 5-28
7 Table 5-4. Summary of candidate studies for formaldehyde RfC development by health
8 endpoint category 5-37
9 Table 5-5. Adjustment for nonoccupational exposures to formaldehyde 5-63
10 Table 5-6. Summary of reference concentration (RfC) derivation from critical study
11 and supporting studies 5-67
12 Table 5-7. Relative risk estimates for mortality from nasopharyngeal malignancies
13 (ICD-8 code 147) by level of formaldehyde exposure for different exposure
14 metrics 5-80
15 Table 5-8. Regression coefficients from NCI log-linear trend test models for NPC
16 mortality from cumulative exposure to formaldehyde 5-81
17 Table 5-9. Extra risk estimates for NPC mortality from various levels of continuous
18 exposure to formaldehyde 5-82
19 Table 5-10. ECooos, LECooos, and inhalation unit risk estimates for NPC mortality from
20 formaldehyde exposure based on the Hauptmann et al. (2004) log-linear
21 trend analyses for cumulative exposure 5-83
22 Table 5-11. ECooos, LECooos, and inhalation unit risk estimates for NPC incidence from
23 formaldehyde exposure based on the Hauptmann et al. (2004) trend
24 analyses for cumulative exposure 5-84
25 Table 5-12. Relative risk estimates for mortality from Hodgkin lymphoma (ICD-8
26 code 201) and leukemia (ICD-8 codes 204-207) by level of formaldehyde
27 exposure for different exposure metrics 5-91
28 Table 5-13. Regression coefficients for Hodgkin lymphoma and leukemia mortality
29 from NCI trend test models 5-92
30 Table 5-14. Extra risk estimates for Hodgkin lymphoma mortality from various levels of
31 continuous exposure to formaldehyde 5-94
This document is a draft for review purposes only and does not constitute Agency policy.
IH-vii DRAFT—DO NOT CITE OR QUOTE
-------
1 LIST OF TABLES (continued)
2
3
4 Table 5-15. Extra risk estimates for leukemia mortality from various levels of
5 continuous exposure to formaldehyde 5-94
6 Table 5-16. ECooos, LECooos, and inhalation unit risk estimates for Hodgkin lymphoma
7 mortality from formaldehyde exposure based on Beane Freeman et al. (2009)
8 log-linear trend analyses for cumulative exposure 5-95
9 Table 5-17. ECoos, LECoos, and inhalation unit risk estimates for leukemia mortality
10 from formaldehyde exposure based on Beane Freeman et al. (2009)
11 log-linear trend analyses for cumulative exposure 5-96
12 Table 5-18. ECooos, LECooos, and inhalation unit risk estimates for Hodgkin lymphoma
13 incidence from formaldehyde exposure, based on Beane Freeman et al.
14 (2009) log-linear trend analyses for cumulative exposure 5-97
15 Table 5-19. ECoos, LECoos, and inhalation unit risk estimates for leukemia incidence
16 from formaldehyde exposure based on Beane Freeman et al. (2009)
17 log-linear trend analyses for cumulative exposure 5-97
18 Table 5-20. Calculation of combined cancer mortality unit risk estimate at 0.1 ppm 5-103
19 Table 5-21. Calculation of combined cancer incidence unit risk estimate at 0.1 ppm 5-103
20 Table 5-22. Summary of tumor incidence in long-term bioassays on F344 rats 5-109
21 Table 5-23. BMD modeling of unit risk of SCC in the human respiratory tract 5-129
22 Table 5-24. Human benchmark extrapolations of nasal tumors in rats by using
23 formaldehyde flux and DPX 5-133
24 Table 5-25. Formaldehyde-induced rat tumor incidences 5-134
25 Table 5-26. Summary of inhalation unit risk estimates 5-140
26 Table 5-27. Extra risk estimates per ppm based on ECs 5-141
27 Table 5-28. Total cancer risk from exposure to a constant formaldehyde exposure level
28 of 1 ug/m3 from ages 0-70 years 5-141
29
30
31
Table 6-1.
Table 6-2.
32 Table 6-3.
Summary of candidate reference concentrations (RfC) for cocritical studies... 6-28
Effective concentrations (lifetime continuous exposure levels) predicted for
specified extra cancer risk levels for selected formaldehyde-related cancers... 6-36
Inhalation unit risk estimates based on epidemiological and experimental
This document is a draft for review purposes only and does not constitute Agency policy.
IH-viii DRAFT—DO NOT CITE OR QUOTE
-------
animal data 6-41
This document is a draft for review purposes only and does not constitute Agency policy.
IH-ix DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
1 Figure 5-1. Change in number of additions made in 10 minutes following
2 formaldehyde exposure at 0.04, 0.21, 0.48, or 1.1 mg/m3 (32, 170, 390,
3 or890ppb) 5-22
4 Figure 5-2. Effects of formaldehyde exposure on the error rate of female LEW. IK rats
5 performing the water labyrinth learning test 5-25
6 Figure 5-3. Fecundity density ratio among women exposed to formaldehyde in the high
7 exposure index category with 8-hour time-weighted average formaldehyde
8 exposure concentration of 219 ppb 5-31
9 Figure 5-4. Estimated reduction in peak expiratory flow rate (PEFR) in children in
10 relation to indoor residential formaldehyde concentrations 5-42
11 Figure 5-5. Odds ratios for physician-diagnosed asthma in children associated with
12 in-home formaldehyde levels in air 5-46
13 Figure 5-6. Prevalence of asthma and respiratory symptom scores in children
14 associated with in-home formaldehyde levels 5-50
15 Figure 5-7. Prevalence and severity of allergic sensitization in children associated with in-
16 home formaldehyde levels 5-50
17 Figure 5-8. Positive exposure-response relationships reported for in-home
18 formaldehyde exposures and sensory irritation (eye irritation) 5-53
19 Figure 5-9. Positive exposure-response relationships reported for in-home
20 formaldehyde exposures and sensory irritation (burning eyes) 5-54
21 Figure 5-10. Age-specific mortality and incidence rates for myeloid, lymphoid, and all
22 leukemia 5-101
23 Figure 5-11. Schematic of integration of pharmacokinetic and pharmacodynamic
24 components in the CUT model 5-106
25 Figure 5-12. Fit to the rat tumor incidence data using the model and assumptions in
26 Conolly et al. (2003) 5-110
27 Figure 5-13. Spatial distribution of formaldehyde over the nasal lining, as characterized
28 by partitioning the nasal surface by formaldehyde flux to the tissue per
29 ppm of exposure concentration, resulting in 20 flux bins 5-124
30 Figure 5-14. Distribution of cells at risk across flux bins in the F344 rat nasal lining 5-124
This document is a draft for review purposes only and does not constitute Agency policy.
IH-x DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
1
2 Figure 5-15. MLE and upper bound (UB) added risk of SCC in the human nose for
3 two BBDR models 5-128
4 Figure 5-16. Replot of log-probit fit of the combined Kerns et al. (1983) and
5 Monticello et al. (1996) data on tumor incidence showing BMC10 and
6 BMCL10 5-131
7 Figure 5-17. EPA Multistage Weibull modeling: nasal tumor dose response 5-136
8 Figure 5-18. Multistage Weibull model fit 5-137
9 Figure 5-19. Multistage Weibull model fit of tumor incidence data compared with KM
10 estimates of spontaneous tumor incidence 5-137
This document is a draft for review purposes only and does not constitute Agency policy.
IH-xi DRAFT—DO NOT CITE OR QUOTE
-------
This page intentionally left blank.
1
2
3
This document is a draft for review purposes only and does not constitute Agency policy.
IH-xii DRAFT—DO NOT CITE OR QUOTE
-------
1 5. QUANTITATIVE ASSESSMENT: INHALATION EXPO SURE
2
3
4 This chapter presents the quantitative assessments conducted by EPA for both cancer and
5 noncancer health effects associated with formaldehyde exposure. The quantitative assessment is
6 focused on the inhalation route of exposure. The current IRIS reference dose (RfD) is not
7 reevaluated in this assessment. Formaldehyde's carcinogenicity via the oral route of exposure is
8 not evaluated herein nor is an oral slope factor considered at this time. Therefore, the following
9 sections address derivation of a reference concentration (RfC) and cancer unit risk estimate for
10 inhalation exposures.
11 For noncancer effects, the RfC is an estimate (with uncertainty spanning perhaps an order
12 of magnitude) of a continuous inhalation exposure to the human population (including sensitive
13 subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime.
14 It can be derived from a NOAEL, LOAEL, or benchmark concentration, with uncertainty factors
15 generally applied to reflect limitations of the data used. Data from the previous chapters are
16 evaluated to determine the health effects associated with formaldehyde exposure and which
17 studies may best inform the exposure response relationship for RfC derivation. Section 5.1
18 summarizes the observed noncancer health effects, selecting key studies and critical effects for
19 consideration. Candidate RfCs are derived for each identified key study. Several alternatives
20 are considered for uncertainty factors addressing human variability for key studies and
21 alternatives presented (see Section 5.1.2.3). Options for addressing the overall database
22 uncertainty factor are provided which may modify the final RfC (see Section 5.1.3).
23 The derivation of the cancer inhalation unit risk estimate considered data regarding both
24 respiratory tract cancers and lymphohematopoietic malignancies. Exposure-response modeling
25 from epidemiologic studies was used to derive a combined unit risk estimate for nasopharyngeal
26 cancer and lymphohematopoietic cancers (see Section 5.2). This unit risk estimate is supported
27 by an analysis of exposure-response modeling of respiratory tract cancer risk using data from
28 experimental animal studies (see Section 5.3). Analysis of the animal bioassays includes an
29 evaluation of a published biologically based dose-response model as well as an appraisal of
30 published dose-response modeling of genomics data and a presentation of benchmark dose
31 modeling approaches. Finally, Section 5.4 provides a summary and conclusions from the cancer
32 exposure-response modeling, presenting the final unit risk estimate based on the combined risk
33 of nasopharyngeal cancer and lymphohematopoietic cancers observed in the epidemiology
34 studies.
35
This document is a draft for review purposes only and does not constitute Agency policy.
5-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 5.1. INHALATION REFERENCE CONCENTRATION (RFC)
2 Prior to the current assessment, the EPA IRIS file for formaldehyde did not provide an
3 inhalation RfC. As presented in the hazard identification in Chapter 4, a number of noncancer
4 health effects are associated with formaldehyde exposure. Section 5.1.1 describes each of the
5 health effect categories considered for RfC derivation and the specific endpoints considered for
6 each category. The identified effect categories are: sensory irritation (eye, nose, and throat);
7 upper respiratory tract (URT) pathology; pulmonary function; increased asthma and atopic
8 sensitization; altered immune function; neurotoxicity and reproductive and developmental
9 toxicity. For each health effect category, studies that may adequately inform the
10 exposure-response relationship for specific critical effects are identified for consideration in RfC
11 derivation.
12 EPA employed a screening process across the different health effect categories to select
13 key studies that would best support the derivation of an inhalation RfC (as described in
14 Section 5.1.2.1). The following factors were considered in this evaluation: characteristics of the
15 study population, exposure regimen, quality of exposure assessment, quality of
16 exposure-response assessment, exposure levels at which effects were seen and statistical power
17 of the study. Based on this analysis, seven studies were considered for RfC derivation.
18 Candidate RfC derivation from a key study includes the following steps: 1) define the critical
19 effect(s); 2) determine appropriate point(s) of departure (PODs) on the basis of inhaled
20 concentration; 3) adjust each POD by endpoint/study-specific uncertainty factors (UFs), to
21 account for uncertainties in the extrapolation of study results to conditions of human
22 environmental exposure. All of the identified key studies were observational epidemiology
23 studies of people and several studies included potentially susceptible individuals (e.g., children,
24 asthmatics). The uncertainty factor for human variability has sometimes been reduced for
25 studies of susceptible populations or lifestages. However, for five of the seven key studies it was
26 unclear if an uncertainty factor of 3 or 1 for human variability was most appropriate. Therefore,
27 alternatives are presented for consideration. Candidate RfCs (cRfCs) are derived for sensory
28 irritation, decreased pulmonary function in children, increased asthma incidence in children,
29 increased allergic sensitization to common allergens in children, and decreased fecundability
30 density ratio (FDR) in women (increased time to pregnancy) (see Table 5-7). All of these cRfCs
31 are derived from endpoints identified in residential studies, with the exception of decreased FDR
32 (observed in an occupational study of women in the woodworking industry).
33 The overall literature database of both human and laboratory animal studies examining
34 the health effects from formaldehyde exposure is large; however, the available studies for some
35 types of effects are limited. Limitations in the existing database are discussed in Section 5.1.3,
This document is a draft for review purposes only and does not constitute Agency policy.
5-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 specifically regarding understanding the reproductive and developmental effects and the
2 exposure-response relationship for the observed neurological and behavioral effects from
3 formaldehyde exposure. EPA considers 3 options for addressing these database uncertainties in
4 the final RfC: (1) providing an RfC derived from studies of respiratory and allergenic responses
5 and protective of sensory irritation effects, without further adjustment for uncertainties in the
6 database (noting the need for further research to elucidate reproductive, developmental and
7 neurotoxic effects); (2) providing an RfC with a database uncertainty factor incorporated to
8 reflect the potential that reproductive, developmental, or neurotoxic effects might occur at lower
9 doses; or (3) provide a range for the RfC which encompasses the above two options for the
10 database uncertainty factor.
11
12 5.1.1. Candidate Critical Effects by Health Effect Category
13 The following subsections describe the best available studies and endpoints for
14 quantitative RfC derivation within each health effect category. These studies are considered
15 representative of the health effects attributed to formaldehyde exposure. For more details on
16 specific studies discussed here, see Sections 4.1.1 and 4.2.1. The identified health effect
17 categories are: sensory irritation (eye, nose, and throat); upper respiratory tract (URT) pathology;
18 pulmonary function; increased asthma and allergic sensitization; altered immune function;
19 neurotoxicity and reproductive and developmental toxicity. Discussions in each subsection
20 below describe the various health effects observed in human and animal studies for each
21 category.
22 For each health effect category, specific studies that may adequately inform the exposure-
23 response relationship for critical effects are identified for consideration in RfC derivation. In
24 general, studies are included where study quality and ability to define exposures are considered
25 adequate for RfC derivation. Whenever possible, greater consideration is typically given to
26 human data from observational epidemiology studies for derivation of an RfC.
27
28 5.1.1.1. Sensory Irritation of the Eyes, Nose, and Throat
29 Eye, nose, and throat irritation are common effects of chemically induced sensory
30 irritation; specific effects include lacrimation, burning of the eyes and nose, rhinitis, burning of
31 the throat, and cough (Feron et al., 2001). Chemical irritants such as formaldehyde bind to
32 protein receptors of the trigeminal nerve, triggering a burning and painful sensation. This
33 process is distinct from taste and smell (Cometto-Muniz and Cain, 1992; Nielsen, 1991). The
34 trigeminal nerve has three branches (ophthalmic, maxillary, and mandibular) and not only acts as
35 an afferent nerve relaying these sensations to the central nervous system but has efferent nerve
This document is a draft for review purposes only and does not constitute Agency policy.
5.3 DRAFT—DO NOT CITE OR QUOTE
-------
1 activity as well (Meggs, 1993). Stimulation of the trigeminal nerve may result in reflex
2 responses, including lacrimation, coughing, and sneezing. In this assessment, both the reflex
3 responses and the sensations (such as burning, pain, and itching of the eyes, nose, and throat) are
4 considered adverse effects (see Section 4.1 for a full discussion of available human data).
5 There are studies noting irritant effects in rodents (Sarsilmaz et al., 1999; Holmstrom
6 et al., 1989; Dubreuil et al., 1976) and monkeys (Monticello et al., 1989; Rusch et al., 1983).
7 These animal studies are supportive of the health effects reported in humans. However, given
8 the uncertainties in extrapolation from responses in laboratory animals to expected responses in
9 humans, the available human studies are preferred.
10 In human studies, the endpoints for assessing irritation include subjective self reporting
11 of symptoms (e.g., pain, burning, itching, increased cough) via questionnaires or objective
12 measures of irritation that can be assessed during controlled acute exposures (e.g., eye-blink
13 counts, lacrimation). Several acute chamber studies support development of a concentration-
14 response relationship for sensory irritation, identifying an effect level for various exposure
15 durations (Kulle, 1993; Andersen and M01have, 1983; Bender et al., 1983; Weber-Tschopp et al.,
16 1977). Arts et al. (2006b) reviewed several studies and performed BMD analyses, reporting
17 10% extra risk BMCL values for reported eye discomfort of 560 and 240 ppb for 3 and 5 hour
18 exposures, respectively. LOAELs of 1,000 ppb and 1,700 ppb were reported for 1-2 minute
19 exposures (Bender et al., 1983; Weber-Tschopp et al., 1977). These acute studies support a role
20 for both concentration and duration in the effect level for eye irritation. Although exposure
21 concentrations are well-defined in these chamber studies, the chamber studies are not appropriate
22 for RfC derivation because they are of acute duration and the exposure levels used are much
23 higher than those reported for chronic exposure scenarios, both occupational and residential.
24 A study of industrial workers assessed sensory irritation and provided an average
25 exposure derived from in-plant exposure measurements and the work history of each study
26 participant (Holmstrom and Wilhelmsson, 1988). Although average daily exposures were
27 estimated for each employee, these data were not used to explore an exposure-response
28 relationship within the worker cohort. The symptom prevalence for sensory irritation (e.g., nasal
29 discomfort, eye discomfort, and airway discomfort) relative to the referent group was reported
30 for the cohort as a whole, where worker exposure ranged from 0.05 to 0.5 mg/m3 formaldehyde
31 8-hour time-weighted average (TWA), with a mean of 0.26 mg/m3 (210 ppb). The daily TWA
32 does not reflect the peak exposures experienced during specific work tasks. Although this study
33 demonstrated marked increases in symptoms of sensory irritation in the workplace due to
34 formaldehyde exposure, it provided little data to inform the exposure-response relationship,
35 especially in the range of environmental exposures.
This document is a draft for review purposes only and does not constitute Agency policy.
5.4 DRAFT—DO NOT CITE OR QUOTE
-------
1 There are three studies that report sensory irritation in humans from chronic exposures in
2 a residential environment and provide sufficient exposure data to support quantitative assessment
3 (Liu et al., 1991; Ritchie and Lehnen, 1987; Hanrahan et al., 1984). Each study reports
4 site-specific exposure measurements and presents some metric of individual exposure. These
5 residential studies employ in-home measurements for each study participant, either as average
6 exposure level (Ritchie and Lehnen, 1987; Hanrahan et al., 1984) or as calculated cumulative
7 exposure based on the time in the home (Liu et al., 1991). Eye irritation is reported at similar
8 levels of residential formaldehyde exposure in the three studies (see Section 5.1.2.2.4). Each
9 study provides an exposure-response relationship for prevalence of sensory irritation in relation
10 to in-home formaldehyde exposure based on individual-level data. The detailed exposure
11 information and chronic nature of the exposures support the selection of these studies as
12 potential principal studies for RfC derivation. Each of these studies is further evaluated and a
13 cRfC developed for consideration (see Section 5.1.2).
14
15 5.1.1.2. Upper Respiratory Tract Pathology
16 Formaldehyde-induced respiratory tract pathology includes inflammation, rhinitis, goblet
17 cell hyperplasia, metaplastic changes, squamous cell hyperplasia, and impaired mucociliary
18 transport. A series of laboratory animal studies assessing formaldehyde-induced changes in the
19 nasal mucosa suggests that these changes may be a protective or adaptive response and that
20 increased mucus flow and metaplastic changes will progress in relation to the concentration and
21 duration of exposure to protect the underlying tissue (Swenberg et al., 1983). The degree of
22 inflammation, hyperplasia, and metaplastic change that is due to sensory irritation-induced
23 inflammatory responses versus inflammation and tissue remodeling from formaldehyde-induced
24 direct cell damage cannot be distinguished. These changes have been noted as sensitive
25 indicators of formaldehyde-induced effects, occurring before gross cellular damage and focal
26 lesions (Monticello et al., 1989). These responses are considered for RfC derivation, especially
27 for exposure concentrations where gross damage of the underlying tissue is not expected.
28 Although well-documented studies demonstrating formaldehyde-induced upper respiratory tract
29 (URT) pathology have been performed in laboratory animals, including the rat (Zwart et al.,
30 1988; Woutersen et al., 1987; Morgan et al., 1986a, b, 1983; Swenberg et al., 1986, 1983) and
31 monkey (Rusch et al., 1983), robust human data from epidemiologic studies are available, and
32 these human data are preferred for RfC derivation.
33 Six epidemiology studies examined the effects of formaldehyde exposure on URT
34 pathology (Pazdrak et al., 1993; Boysen et al., 1990; Holmstrom et al., 1989; Edling et al., 1988;
35 Holmstrom and Wilhelmsson, 1988; Andersen and M01have, 1983). Of these studies,
This document is a draft for review purposes only and does not constitute Agency policy.
5.5 DRAFT—DO NOT CITE OR QUOTE
-------
1 Holmstrom and Wilhelmsson (1988) and Holmstrom et al. (1989) were identified as the most
2 robust and sensitive and are included as candidate studies for RfC derivation. Both studies
3 address the same cohort and, thus, were considered together. The Holmstrom and Wilhelmsson
4 (1988) study is discussed above under sensory irritation effects. In this study of 70 factory
5 workers exposed to a TWA formaldehyde concentration of 210 ppb, impaired mucociliary
6 clearance was reported in 20% of the exposed workers and 3% of the 36 nonexposed workers.
7 Using rhinomanometry, Holmstrom and Wilhelmsson (1988) also found an increase in nasal
8 resistance due to mucosal swelling, though this increase was not statistically significant. In
9 Holmstrom et al. (1989), nasal biopsy samples were collected from 62 of the 70 formaldehyde-
10 exposed factory workers (these 62 had been exposed to a TWA formaldehyde concentration of
11 240 ppb) and also from 32 of the nonexposed workers. A pathologist scored each sample by
12 using a scale of 0 (normal respiratory epithelium) to 8 (carcinoma). Biopsy scores for both the
13 exposed and control groups ranged from 0 (normal respiratory epithelium) to 4 (stratified
14 squamous epithelium with marked horny layer). Although the mean biopsy scores for the
15 two groups were similar—2.16 for the formaldehyde-exposed workers and 1.56 for the
16 unexposed workers—the difference was statistically significant and the authors reported that the
17 loss of cilia, goblet cell hyperplasia, and the incidence of cuboidal and squamous cell metaplasia
18 replacing the columnar epithelium were more frequent in the group exposed to formaldehyde.
19 There was no correlation between the duration of exposure and histologic changes or between
20 smoking habits and biopsy scores. The URT effects, taken together (decreased mucous flow,
21 increased inflammation, decreased nasal flow, and degradation of the respiratory epithelium),
22 demonstrate a range of formaldehyde-induced URT pathology consistent with effects observed
23 in controlled animal studies.
24
25 5.1.1.3. Pulmonary Function Effects
26 A synthesis of the literature evaluating formaldehyde exposure and pulmonary function is
27 provided in Section 4.4.2. The potential effects of formaldehyde exposure on pulmonary
28 function in humans can be examined on several time-scales of interest. There are reports
29 examining effects from acute exposures among naively exposed anatomy graduate students
30 (Kriebel et al., 1993; 2001), anatomy graduate students with several weeks of episodic exposure
31 (Kriebel et al., 1993), as well as postshift versus preshift differences in pulmonary function in
This document is a draft for review purposes only and does not constitute Agency policy.
5-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 workers with regular occupational exposure (Malaka and Kodama, 1990; Herbert et al., 1994;
2 Alexandersson et al., 1982; Alexandersson and Hedenstierna, 1989). Depending on whether the
3 exposures are naive, the epidemiologic studies that assessed the pulmonary effects of acute
4 exposures to formaldehyde may be assessing different biological responses, namely, the acute
5 effect alone or the acute effect(s) in people who may have already been sensitized to
6 formaldehyde effects.
7 Pulmonary effects of acute formaldehyde exposure have been studied in both healthy
8 volunteers and sensitive populations under controlled conditions (e.g., acute chamber studies).
9 Although acute chamber studies have the advantage of measured controlled exposures, other
10 factors can limit the usefulness of the studies for RfC derivation including: acute duration, small
11 study populations and lack of statistical power to assess the measured parameters. The acute
12 chamber studies are more fully evaluated in Section 4.1.1 and will not be further considered here
13 for RfC derivation.
14 The observed effects in the previously unexposed anatomy students provide additional
15 information on acute exposures in two naive populations (Kriebel et al., 1993; 2001), as well as
16 insight into the intermediate stages of possible sensitization (Kriebel et al., 1993). Kriebel and
17 colleagues (1993) examined the prelaboratory and postlaboratory peak expiratory flow (PEF) in
18 students attending anatomy classes once per week. They found the strongest pulmonary
19 response when examining the average cross-laboratory decrement in peak expiratory flow in the
20 first 2 weeks of the study when formaldehyde concentrations collected in the breathing zones
21 had a geometric average concentration of 0.73 ppm. Overall, the students exhibited a
22 2% decrement in PEF, while the students with any history of asthma showed a 7.3% decrement
23 in PEF. These findings of acute decreases in PEF following students' initial anatomy sessions
24 were corroborated by the Kriebel et al. (2001) study, which used a similar study design applied
25 to another class of anatomy students.
26 The first Kriebel et al. (1993) study also shows how the acute effects of formaldehyde
27 exposure were altered following several weeks of episodic exposure. By the 5th week of class,
28 the pre- and post-laboratory measurements of PEF were no longer reflecting a clearly
29 demonstrated acute effect but following the 7th week of episodic exposure, both pre-and
30 post-laboratory PEF continued to drop steadily until the class adjourned after 10 weeks time.
31 While the acute effects of formaldehyde exposure appeared to diminish after several weeks of
32 exposure, the intermediate effect across 10 weeks was a 27 liter/minute drop in PEF that was
33 statistically significant (p<0.01) after statistical control for random person effects, asthma, an
34 interaction between time and asthma and eye and nose symptoms of irritation. The Kriebel et al.
35 (1993) study is considered of sufficient quality to support an acute RfC but the quantitative
This document is a draft for review purposes only and does not constitute Agency policy.
5.7 DRAFT—DO NOT CITE OR QUOTE
-------
1 details on the initial acute effects among the naively exposed students are not adequately
2 provided. The findings of the Kriebel et al. (2001) study may have been influenced by decreased
3 class attendance, which dropped from 37 in the first week to 20 in week 6 and to just 10 students
4 by week 10. While the Kriebel et al. (2001) study could be useful as a supportive study for
5 naively exposed students, the longitudinal component is not strong enough to support RfC
6 development.
7 Several studies of workers assess both cross-shift and chronic effects of formaldehyde
8 exposure (Malaka and Kodama, 1990; Herbert et al., 1994; Alexandersson et al., 1982;
9 Alexandersson and Hedenstierna, 1989). Since formaldehyde exposure may have cumulative
10 effects over chronic exposures, occupational studies generally showed clinically small but
11 statistically significant decrements in pulmonary function across shifts. In general these studies
12 did not identify, have information on or have appropriate statistical control of, potential
13 confounding coexposures. While these occupational studies provide evidence that is clearly and
14 consistently supportive of an acute effect on pulmonary function, they do not directly support
15 RfC development of an acute effect divorced of the concomitant chronic effects.
16 Several studies allowed for the examination of potential chronic effects of formaldehyde
17 exposure. These included an occupational study (Malaka and Kodama, 1990) that reported
18 preshift pulmonary function as a percentage of expected among the formaldehyde exposed
19 compared to comparable people not exposed to formaldehyde. Studies that did not report
20 preshift pulmonary function as a percentage of expected function (Herbert et al., 1994;
21 Alexandersson et al., 1982) contribute less to an assessment of potential chronic effects because,
22 post hoc, it is difficult to calibrate for cross-study comparison the multiple pulmonary function
23 data without knowledge of the age, gender, smoking status, height, year of birth, etc. that are
24 important determinants of the pulmonary function metrics of concern. The single study (Malaka
25 and Kodama, 1990) that did report functional measures in relation to expected value, found that
26 an average 8-hour time weighted average formaldehyde exposure of 1.13 ppm from area samples
27 was associated with statistically significant decrements in FEVi, FEVi/FVC and FEF2s 75
28 compared to a referent population. The strongest response was for FEF2s 75, which showed a
29 12% drop in observed function compared to expected function in the unexposed, but it is unclear
30 how to interpret the potential chronic health effect(s) with just the magnitude of the decrement
31 and the length of the average occupational tenure at this plywood facility (6.5 years), which was
32 not reported by exposure status.
33 One study reported on the longitudinal follow-up of workers exposed to formaldehyde
34 (Alexandersson and Hedenstierna, 1989). This investigation not only examined the acute effects
35 of exposure across shift, but was able to do so among some of the same workers that had been
This document is a draft for review purposes only and does not constitute Agency policy.
5-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 studied five years earlier (Alexandersson et al., 1982). Statistically significant decreases in
2 FEVi/FVC and FEF25-75 were noted over the intervening five years in nonsmokers after
3 correction for normal aging. The decrease in FEF25_75 was 0.212 liters/s (SD = 0.066 liters/s) for
4 each year of exposure and was highly significant (p < 0.01). For comparison with the 12% drop
5 in the same pulmonary metric reported by Malaka and Kodama (1990) over an estimated
6 6.5 years, EPA computed the extrapolated percentage decrease in FEF25_75 for the Alexandersson
7 and Hedenstierna (1989) using the reported yearly decrement applied to the preshift values at the
8 time of the initial study period. EPA calculated that from the predicted value of 4.57 liters/s, a
9 decrease of 0.168 liters/s could be estimated for each year of exposure regardless of smoking
10 status. For 6.5 years of exposure, this would result in a 24% drop in FEF25-75. Formaldehyde
11 concentrations were estimated at 0.42 ppm in the first Alexandersson et al. (1982) study and at
12 0.50 ppm in the second study, but without better exposure measures, the results of the
13 longitudinal follow-up cannot support quantitative RfC development.
14 Information is lacking in these studies such as length or tenure of employment associated
15 with the preshift pulmonary function or how long the residents had lived in their homes.
16 Likewise, knowledge of how occupational or residential exposure may have changed over time
17 would have allowed for an examination of the progression of any decrement in function
18 associated with long-term episodic exposure. Among these studies, the best designed and
19 executed of the cross-sectional studies was that of Krzyzanowski and colleagues (1990).
20 Municipal employees and their children (613 adults and 298 children) were randomly sampled
21 and were considered to be representative of a diverse local population. Residential exposures to
22 formaldehyde were based on repeated samples from each individual's kitchen, living area and
23 bedroom. The average formaldehyde concentration was 26 ppb, with a maximum sample value
24 of 140 ppb. The majority of subjects (83%) lived in homes with 2-week average concentrations
25 below 40 ppb. Subjects' peak expiratory flow rates (PEFR) were determined 4 times daily in the
26 morning, at noon, in the early evening and before bed for 2 weeks. A statistically significant
27 linear relationship between increased formaldehyde exposure and decreased peak expiratory
28 flow rate was reported in children but not adults. All statistical models controlled for
29 socioeconomic status, tobacco smoking (current active or environmental tobacco smoking) and
30 nitrogen dioxide concentrations. In children, formaldehyde concentrations of 60-140 ppb
31 increased the prevalence of physician-diagnosed asthma and bronchitis. Among adults who
32 smoked, there was a statistically significant nonlinear relationship with decreased morning PEFR
33 for formaldehyde concentration > 40 ppb. This well-conducted study had only minor
34 weaknesses such as non-differential measurement error. However, random measurement error
35 tends to attenuate any true effect and is unlikely to have produced a spurious effect. It is
This document is a draft for review purposes only and does not constitute Agency policy.
5.9 DRAFT—DO NOT CITE OR QUOTE
-------
1 unlikely that these findings were the product of unmeasured or residual confounding as the
2 analyses controlled for smoking as well as nitrogen dioxide levels and there is no evidence of
3 alternative factors that were temporally correlated with formaldehyde concentrations and more
4 strongly associated with decrements in pulmonary function. This study of a large and
5 representative sample from a diverse study population with a well-quantified concentration-
6 response function and is further considered for RfC derivation.
7
8 5.1.1.4. Asthma and Allergic Sensitization (Atopy)
9 Sensitization to inhalational chemical exposure may manifest as an allergic or asthmatic
10 response that is characterized by bronchial constriction (BC) or bronchial hyperresponsiveness
11 (BHR). This Sensitization may be a result of immune involvement, as in the case of
12 hypersensitivity, or a neurogenic Sensitization, where a chemical may directly stimulate
13 inflammation. Asthma is a specific manifestation of IgE-mediated hypersensitivity,
14 characterized by BHR and airway inflammation, resulting in lower airway obstruction (Fireman,
15 2003; Kuby, 1991).
16 A variety of hypersensitivity reactions have been reported following exposure to
17 formaldehyde. Rashes and skin reactions have been reported in some individuals after dermal
18 exposures to formaldehyde. Increased expression of Th-2 cytokines in the lymph nodes of mice
19 given dermal applications of formaldehyde indicates the involvement of an immune component
20 to the observed Sensitization (Dearman et al., 2005; Hilton et al., 1998; Arts et al., 1997).
21 However, the response does not appear to be IgE mediated (Arts et al., 1997; Lee et al., 1984).
22 Gorski et al. (1992) observed an increase in formaldehyde-mediated neutrophil burst in
23 dermatitis patients exposed in a controlled chamber study and suggests a putative role of
24 oxidative stress and reactive oxygen species (ROS).
25
26 5.1.1.4.1. Epidemiologic studies.
27 A synthesis of the literature evaluating formaldehyde exposure and asthma and allergic
28 Sensitization is provided in Section 4.4.3. Inhalation exposure has been associated with
29 increased asthmatic responses in asthmatics in occupational settings. While few available case
30 reports of bronchial asthma suggest direct respiratory tract Sensitization to formaldehyde gas
31 (Lemiere et al., 1995; Burge et al., 1985; Hendrick et al., 1982; Hendrick and Lane, 1977, 1975),
32 a greater body of epidemiological data provides evidence of an association between
33 formaldehyde exposure and exacerbation of asthmatic responses in compromised individuals
34 (Kriebel et al., 1993) and particularly in children (Rumchev et al., 2002; Garrett et al., 1999a,b;
35 Krzyzanowski et al., 1990). Asthma incidence in children increased with in-home exposure to
This document is a draft for review purposes only and does not constitute Agency policy.
5-10 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde (Rumchev et al., 2002). Similarly, the frequency of respiratory symptoms
2 associated with asthmatic responses and measures of allergic sensitization in children increased
3 with in-home formaldehyde exposure (Garrett et al., 1999 a,b).
4 The association between formaldehyde and asthma has been studied by examining
5 occupational exposures (Fransman et al., 2003; Malaka and Kodama, 1990), school-related
6 exposures (Zhao et al., 2008; Smedje and Norback, 2001; Norback et al., 2000) and residential
7 exposures (Matsunaga et al., 2008; Tavernier et al., 2006; Gee et al., 2005; Delfino et al., 2003;
8 Rumchev et al., 2002; Garrett et al., 1999 a,b; Palczynski et al., 1999; Norback et al., 1995;
9 Krzyzanowski et al., 1990). The two occupational studies examined the respiratory health of
10 plywood workers (Fransman et al., 2003; Malaka and Kodama, 1990). The most recent of these
11 was conducted in New Zealand by Fransman et al. (2003). Personal samples of formaldehyde
12 exposure were taken. The mean level of exposure was 0.08 mg/m3 (65 ppb) and the majority of
13 samples were below the limit of detection which was reported to be 0.03 mg/m3 (24 ppb).
14 Compared with those with low levels of formaldehyde exposure, workers with high levels of
15 exposure were more likely to report having asthma (OR = 4.3 [95% CI: 0.7-27.7]). The
16 association was not seen when examining formaldehyde exposure and use of asthma medication.
17 The second study of plywood workers was completed in Indonesia. Background levels of
18 formaldehyde ranged from 0.003 to 0.07 ppm. The highest concentration of formaldehyde
19 detected in an air sample was in the particleboard unit (range 1.16 to 3.48 ppm). The occurrence
20 of asthma was found to be positively associated with formaldehyde exposure, where asthma was
21 defined as, "Have you ever had an attack of wheezing that made you feel short of breath?",
22 (Malaka and Kodama, 1990).
23 Studies of exposure to formaldehyde at schools have been performed in China (Zhao
24 et al., 2008) and in Sweden (Smedje and Norback, 2001). In the study from China (Zhao et al.,
25 2008), mean levels of formaldehyde were reported to be 2.3 |ig/m3 (range 1.0-5.0 |ig/m3)
26 indoors and 5.8 |ig/m3 (range 5.0-7.0 |ig/m3) outdoors. Cumulative asthma (i.e., physician-
27 diagnosed asthma since birth) and daytime attacks of breathlessness were found to be associated
28 with outdoor formaldehyde levels. Neither of these outcomes was associated with indoor
29 concentrations of formaldehyde; however, indoor levels were found to be associated with
30 nocturnal attacks of breathlessness. In Sweden (Smedje and Norback, 2001), the levels of
31 formaldehyde measured indoors were higher (arithmetic mean 8 |ig/m3, geometric mean 4
32 Hg/rn3, range <5.0-72 |ig/m3). One difference between the Swedish study and the study
33 conducted in China is that the Swedish study examined the incidence of asthma over a 4-year
34 period and did not report an association between formaldehyde exposure and the incidence of
35 asthma (OR 1.2 [95% CI: 0.8-1.7]) among the whole study population. However, when the
This document is a draft for review purposes only and does not constitute Agency policy.
5-11 DRAFT—DO NOT CITE OR QUOTE
-------
1 investigators stratified based on history of atopy, they reported that among children without a
2 history of atopy, a new diagnosis of asthma was significantly more likely at higher
3 concentrations of formaldehyde (OR 1.7 per 10 |ig/m3 [95% CI: 1.1-2.6]) and at higher total
4 concentrations of mold (OR = 4.7 per 10-fold increased in total molds [95% CI: 1.2-18.4] in the
5 classroom air. The finding in increase health effects due to formaldehyde and mold exposures
6 did not appear to control for the other exposure and no information on the potential correlation
7 between the two exposures was provided. In order to evaluate the potential for confounding of
8 the reported formaldehyde association by the reported mold association, the magnitude of effects
9 must be compared on an appropriate scale since the magnitude of an odds ratio depends on the
10 magnitude of the change in exposure level that is expected to produce increased risk. After
11 standardizing the units to the reported geometric standard deviation (GSD), the results for
12 formaldehyde is OR1 = 1.13 per GSD increase in formaldehyde concentration and the results for
13 mold is OR2=1.02 per GSD increase in mold exposure (based on a 10-fold increase from the
14 mean mold exposure) or alternatively, OR3 = 1.06 per GSD increase in mold exposure (based on
15 a 10-fold increase from the minimum mold exposure). As it appears that the magnitude of the
16 formaldehyde effect is stronger than that of the mold effect (following standardization of
17 exposure increment), it can be concluded that the reported formaldehyde effect could not have
18 been due to uncontrolled confounding by mold.
19 The results of studies measuring residential exposure to formaldehyde and asthma are
20 varied, with some demonstrating an association and others finding no relationship. A recent
21 study (Matsunaga et al., 2008) found no association between 24-hour formaldehyde and
22 prevalence of asthma when pregnant women with an exposure to >47 ppb were compared to
23 those with exposure to <18 ppb. However, they reported an increased risk of atopic eczema.
24 This study did not assess the risk of incident asthma. A study utilizing self-reported asthma
25 prevalence as an outcome also found no association with levels of formaldehyde (mean
26 25.9 |ig/m3, range 2.0-66.8 |ig/m3) (Palczynski et al., 1999), although they noted the incidence
27 of allergic diseases was greatest in the highest formaldehyde exposure group but that the groups
28 were too small for statistical evaluation.
29 A study performed by Tuthill (1984) measured formaldehyde exposure for children
30 grades K through 6 by using a combination of proxy variables. Overall, there was no
31 association, but some individual variables showed an increased risk. For example, the reported
32 risk ratio for having new construction or remodeling performed in the house in the past 4 months
1 OR per GSD of formaldehyde = xp[ln(OR per ng/m3)/10 ug/m3 * 2.3 ug/3] = exp[ln(l.7)710*2.3] = 1.13
'3-1
z OR per GSD of mold = xp[ln(OR per 10-fold increase)/ (9*Geo. Mean)*1.6 ug/"] = exp[ln(4.7)/162*1.6] = 1.02
3 OR per GSD of mold = xp[ln(OR per 10-fold increase)/ (9*Minimum)*1.6 ug/3] = exp[ln(4.7)/45*1.6] = 1.06
This document is a draft for review purposes only and does not constitute Agency policy.
5-12 DRAFT—DO NOT CITE OR QUOTE
-------
1 was 2.5 (95% CI: 1.7-3.9). The risk ratio for having new or upholstered furniture in the house
2 (within the past 4 months) was 2.2 (95% CI: 1.2-3.9).
3 The study by Delfino et al. (2003) assessed whether the ambient formaldehyde
4 concentration measured at a central monitoring site was associated with asthma symptoms. The
5 study examined 22 10-15 year olds with at least 1 year of physician-diagnosed asthma and living
6 in a nonsmoking household. The mean levels of formaldehyde were measured to be 7.21 ppb
7 (range 4.27-14.02 ppb). There was a positive association between asthma symptom scores
8 (comparing children who report symptoms interfering with their daily activities versus those
9 with no symptoms or symptoms not great enough to affect their daily activities) and high current
10 levels of formaldehyde (OR 1.90 [95% CI: 1.13-3.19]).
11 Three studies (Tavernier et al., 2006; Gee et al., 2005; Garrett et al., 1999 a,b) were
12 performed by matching children with and without asthma and comparing the levels of
13 formaldehyde in their homes. Gee et al. (2005) reported median formaldehyde levels of
14 0.03 ppm in living rooms and 0.04 ppm in bedrooms. Analyses were limited to univariate
15 comparisons of formaldehyde levels for cases of existing asthma and controls without asthma.
16 The concentrations did not differ in a statistically significant manner. The study by Gee et al
17 (2005) was followed up with a more sophisticated analysis of the same children in the same
18 home. Tavernier et al. (2006) reiterated the earlier finding by Gee et al. (2005) that
19 formaldehyde was not found to be associated with existing asthma. Tavernier et al. (2006) did
20 not report the measured levels of formaldehyde but gave the OR for the highest tertile of
21 exposure compared with the lowest tertile of exposure as 0.99 (95% CI: 0.39-2.50).
22 Garrett et al. (1999 a,b) reported on the risk of allergy and asthma-like respiratory
23 symptoms due to formaldehyde exposure in a cross-sectional survey of households with children
24 with (n = 53) or without (n = 88) doctor-diagnosed asthma. Formaldehyde exposure was
25 characterized by 4 seasonal in-home sampling events across the year for bedrooms and 4-day
26 passive samples collected in living rooms, kitchens and outdoors. Statistically significant linear
27 trends for increased risk of having asthma were seen with increasing formaldehyde levels
28 (p < 0.02); however, the ORs for the association did not remain statistically significant after
29 controlling for parental allergy and asthma (exact ORs and 95% CIs not given). Garrett et al
30 (1999 a,b) also evaluated the prevalence and severity of allergic sensitization to 12 common
31 allergens and reported increased prevalence with increasing formaldehyde concentration in the
32 home. The respiratory symptom score was also increased and demonstrated a significant effect
33 for formaldehyde in a multiple regression after adjusting for multiple risk factors and
34 interactions. For the atopy and respiratory symptom endpoints, severity/prevalence was
35 increased in the medium (20-50 |ig/m3) and high (>50 |ig/m3) exposure groups relative to the
This document is a draft for review purposes only and does not constitute Agency policy.
5-13 DRAFT—DO NOT CITE OR QUOTE
-------
1 low (<20 |ig/m3) exposure group, based on the highest of four seasonal 4-day formaldehyde
2 measurements in the home. The associations between formaldehyde concentrations and severity
3 of allergic sensitization are clearly shown and further substantiated with multivariate regression
4 controlling for potential confounders which showed that the unadjusted effect estimate was not
5 confounded. In logistic regressions, both the prevalence and severity of allergic sensitization to
6 12 common allergens increased with increasing formaldehyde concentration in the home. The
7 crude association for atopy with an increase in formaldehyde concentration per 10 |ig/m3 was
8 OR = 1.34 which increased only slightly when adjusted for parental asthma and gender to and
9 odds ratio of 1.42 per 10 |ig/m3 (95% CI: 0.99-2.04). Passive smoking, the presence of pets,
10 indoor nitrogen dioxide concentrations, airborne fungal spores and house-dust-mite allergens did
11 not influence the effect estimates and were unlikely to be confounders. Additionally, a
12 calculated respiratory symptom score was increased and demonstrated a significant relationship
13 to increased formaldehyde concentration in a multiple linear regression after adjusting for
14 multiple risk factors and interactions. For each of these endpoints, severity/prevalence was
15 increased in the medium (20-50 |ig/m3) and high (>50 |ig/m3) exposure groups relative to the
16 low (<20 |ig/m3) exposure group, based on the highest of four seasonal 4-day formaldehyde
17 measurements in the home.
18 Residential formaldehyde exposure was associated with an increased risk of asthma in a
19 population-based case-control study of 192 children aged 6 months to 3 years (Rumchev et al.,
20 2002). The study, which comprises 88 cases of children discharged from the emergency
21 department of a children's hospital in Perth, Australia, with a primary diagnosis of asthma and
22 104 controls, provides a positive exposure-response relationship. Seasonal in-home
23 formaldehyde measurements taken in the living room and subject's bedroom were used to assess
24 exposure (8-hour passive sampler). The odds ratios (ORs) for risk of asthma by formaldehyde
25 exposure level category were adjusted for numerous risk factors both familial and environmental
26 including, familial history of asthma, age, sex, smoking, presence of pets, and attributes of the
27 home. Of these, age, allergic sensitization to common allergens, and family history of allergy
28 were independent risk factors for asthma (ORs of 1.09, 2.57, and 2.66, respectively). Categorical
29 analysis of the data indicates the ORs for asthma were increased in the two highest formaldehyde
30 exposure groups, reaching statistical significance for household exposures >60 |ig/m3 (48 ppb)
31 (OR of 1.39). Analysis of the data with formaldehyde as a continuous variable indicated there
32 was a statistically significant increase in the risk of asthma (3 % increase in risk per every
33 10 ug/m3 increase in formaldehyde level. All analyses controlled for other indoor air pollutants,
34 allergen levels, relative humidity, and indoor temperature as well as other risk factors.
This document is a draft for review purposes only and does not constitute Agency policy.
5-14 DRAFT—DO NOT CITE OR QUOTE
-------
1 A study of 202 households (mean formaldehyde level of 26 ppb) found that among
2 children aged 6-15 years old and exposed to environmental tobacco smoke, the prevalence of
3 asthma was 45.5% for those with measured levels of formaldehyde in the kitchen >60 ppb. The
4 prevalence of asthma dropped to 15.1% for levels <40 ppb and 0% for 41-60 ppb. No trend in
5 asthma prevalence was seen for children who were not exposed to environmental tobacco smoke
6 (Krzyzanowski et al., 1990).
7 Finally, a study by Norback et al. (1995) reported mean levels of formaldehyde were
8 29 |ig/m3 (range <5-l 10 |ig/m3) in the bedrooms of individuals experiencing nocturnal
9 breathlessness compared with formaldehyde levels of 17 |ig/m3 (<5-60 |ig/m3) among those
10 without nocturnal breathlessness. The OR for this association was 12.5 (95% CI: 2.0-77.9) and
11 the effect was substantially stronger in magnitude than the associations observed for toluene,
12 terpenes and volatile organic compounds which makes confounding by those coexposures
13 unlikely.
14
15 5.1.1.4.2. Supporting animal studies.
16 Several animal studies report increased airway resistance and BC due to inhalation
17 exposures to formaldehyde (Nielsen et al., 1999; Swiecichowski et al., 1993; Biagini et al., 1989;
18 Amdur, 1960). Changes in pulmonary resistance were observed as early as 10 minutes after
19 exposure (Biagini et al., 1989), and reported effect levels ranged from 0.3-13 ppm. Other
20 pulmonary effects were reported in conjunction with BHR, such as increased tracheal reactivity
21 and decreased pulmonary elasticity (Swiecichowski et al., 1993; Amdur, 1960). Although BHR
22 is a common result of Type I hypersensitivity reaction to an allergen, the observation of BHR
23 alone is not sufficient to demonstrate that an agent induces Type 1 hypersensitivity.
24 BHR may be directly induced both pharmacologically and neurogenically (Joos, 2003;
25 Cain, 2001; Meggs, 1995). There is little evidence that formaldehyde itself is an allergen
26 recognized by the immune system, especially via inhalation (Lee et al., 1984). Although
27 formaldehyde exposure has been reported to alter cytokine levels and immunoglobulins in some
28 experimental systems, these immunomodulatory effects do not support a type 1 hypersensitivity.
29 IgE was unchanged (Fujimaki et al., 2004a; Lee et al., 1984), and cytokine profiles were not
30 consistent with the Th-2 cytokines expected in IgE mediated hypersensitivity (Fujimaki et al.,
31 2004a; Ohtsuka et al., 2003).
32 Formaldehyde-induced dermal sensitization show parallel results. The physical signs of
33 irritation and sensitization are consistently shown (e.g., rashes, edema). Some involvement of
34 the immune response has been demonstrated with positive LLNA assays, indicating proliferation
35 of lymphocytes in lymph nodes draining the affected area (Hilton et al., 1998; Arts et al., 1997).
This document is a draft for review purposes only and does not constitute Agency policy.
5-15 DRAFT—DO NOT CITE OR QUOTE
-------
1 Increased expression of Th-2 cytokines in the lymph nodes of mice given dermal applications of
2 formaldehyde does indicate an immune component to the observed sensitization. However, the
3 response does not seem to be mediated by IgE (Arts et al., 1997; Lee et al., 1984).
4 Ito et al. (1996) reported that a tachykinin NKi receptor, but not the histamine HI or
5 bradykinin B2 receptors, is involved in formaldehyde-induced vascular permeability.
6 Neuropeptides NGF and substance P were affected in BAL and stimulated splenocytes from
7 formaldehyde-exposed mice, with greater effects seen in OVA-immunized mice. Tachykinins
8 (e.g., substance P and neurokinin A) are produced by nerve cells and can directly stimulate
9 bronchoconstriction (Van Schoor et al., 2000). Substance P is also a mediator of neurogenic
10 inflammation. Therefore, although formaldehyde may induce some of the symptoms of
11 type 1 hypersensitivity, these symptoms are more likely neurogenic than immunogenic in origin.
12 In contrast, formaldehyde enhances immunogenic hypersensitivity of known allergens
13 (Sadakane et al., 2002; Riedel et al., 1996; Tarkowski and Gorski, 1995). This potentiation
14 varied based on sensitization protocols (respiratory tract versus systemic, frequency and timing
15 of immunization, allergen, etc.) and formaldehyde exposure regimens (concentration, continuous
16 versus intermittent exposures). Taken as a whole, the results support the finding that
17 formaldehyde exposure can aggravate a type 1 hypersensitivity response (see Table 4-53).
18 The mechanism underlying this response has not been elucidated. Formaldehyde-
19 induced IgE production has been reported in some studies (Vandenplas et al., 2004; Wantke
20 et al., 1996a). Other studies suggest that this effect does not appear to be immunogenic in nature
21 (Fujimaki et al., 2004a; Lee et al., 1984). Although formaldehyde exposure has been reported to
22 alter cytokine levels and immunoglobulins in some experimental systems (Fujimaki et al., 2004a;
23 Ohtsuka et al., 2003), these immunomodulatory effects do not support immunogenically
24 mediated type 1 hypersensitivity.
25 These decrements may be mediated via neurogenic potentiation (Sadakane et al., 2002;
26 Riedel et al., 1996; Tarkowski and Gorski, 1995). Tarkowski and Gorski (1995) suggest that
27 formaldehyde may increase permeability of respiratory epithelium and destruction of
28 immunologic barriers. Tachykinin NKI receptor and various neuropeptides (NGF and substance
29 P) have been implicated in formaldehyde-induced sensitization and lend weight of evidence to a
30 neurogenic MOA (Van Schoor et al., 2000; Ito et al. 1996).
31
32 5.1.1.5. Immune Function
33 Although there are some indications of formaldehyde-induced immunomodulation in
34 laboratory animal studies (Jakab, 1992; Morgan et al., 1986a, b, c; Leach et al., 1983) and
35 reports of increased upper respiratory tract infections in formaldehyde-exposed workers
This document is a draft for review purposes only and does not constitute Agency policy.
5-16 DRAFT—DO NOT CITE OR QUOTE
-------
1 (Lyapina et al., 2004; Krzyzanowski et al., 1990; Holness and Nethercott, 1989), the overall
2 database for toxic effects on immune function and competence is very limited. A study of
3 workers using carbamide-formaldehyde glue indicates decreased neutrophil respiratory burst
4 activity (NRBA) (Lyapina et al., 2004). NRBA was reduced in workers with URT inflammation
5 and long-lasting respiratory tract infections, compared with healthy controls, and in
6 formaldehyde-exposed workers with slight or no respiratory infections. The authors
7 hypothesized that the decreased NRBA in symptomatic workers may be an indication of
8 formaldehyde effects in a susceptible population. Since the workers have increased respiratory
9 tract infections as compared with controls, a formaldehyde-specific effect cannot be excluded.
10 These indications of a functional deficit of the immune system are considered adverse and
11 appropriate for consideration as a critical effect. Although this was a small study (n = 29), the
12 exposed workers had increased chronic URT infections and decreased resistance to infections
13 compared with a control population. Additionally, duration of employment was negatively
14 correlated with both erythrocyte count and hematocrit. Measured formaldehyde concentrations
15 for a work shift were 870 ± 390 |ig/m3 (722 ± 324 ppb). This average work-shift concentration
16 is considered to be the LOAEL for increased respiratory tract inflammation and decreased
17 resistance to infections in a worker population.
18
19 5.1.1.6. Neurological and Behavioral Toxicity
20 Studies evaluating the effects of formaldehyde on nervous system structure or function
21 are described in detail in Sections 4.1.1.6 and 4.2.6 and summarized in Section 4.4.8 and
22 Table 4-58. Taken together, the animal and human data support the conclusion that
23 formaldehyde exposure results in neurological and behavioral toxicity. Observed health effects
24 include impaired memory and learning, developmental effects seen as both structural changes in
25 the brain and behavioral changes, and a potential for increased mortality from amyotrophic
26 lateral sclerosis (ALS). Although studies appropriate for RfC derivation do not exist for each
27 potential neurological and behavioral health effect, several studies are available that provide
28 information that needs to be considered when selecting the formaldehyde RfC.
29 Seven of the available neurotoxicity studies were considered as candidates for RfC
30 development (listed in Table 5-1). All seven studies provided reliable documentation of
This document is a draft for review purposes only and does not constitute Agency policy.
5-17 DRAFT—DO NOT CITE OR QUOTE
-------
a 5 KO
2 ~- S
Table 5-1. Candidate points of departure (fODs) including duration adjustments for nervous system toxicity in
key human and anima\ studies
Reference
Species
PODa
Type
ppb
Exposure scenario
Hours/day
Days/week
Duration
POD duration adjustments'"
POD x Hours/day x Days/week = ppb
Ratio0
Effect
Developmental neuropathology effects
Sarsilmaz et
al. (2001)
Asian et al.
(2006)
Rat
Rat
LOAEL
LOAEL
6,000
6,000
6
6
5
5
30 days
30 days
6,000 x 6/24 x 5/7 = 1,070
6,000 x 6/24 x 5/7 = 1,070
5.6
5.6
Volume and cell
number change in
brain regions
following neonatal
exposure
Volume and cell
number change in
brain regions
following neonatal
exposure
Human neurobehavioral outcomes
Bach et al.
(1990 f
Human
NOAEL
170
5.5
1
1 day
170d x x =170
1
Changes in short-
term memory and
ability to
concentrate. Single
5. 5 -hour exposure
Psychomotor effects
Senichenkova
(1991)
Rat
LOAEL
400
4
7
GD 1-19
400 x 4/24 x 7/7 = 67
6
Changes in open
field motor activity
(exploratory
activity and
habituation in
offspring following
in utero exposure
§
§
oo
O
o
-------
a 5 KO
o S. S3
*•
Table 5-1. foints of departure (fOD) for nervous system toxicity in key human and animal studies (continued)
Reference
Species
PODa
Type
ppb
Exposure scenario
Hours/day
Days/week
Duration
POD duration adjustments'"
POD x Hours/day x Days/week = ppb
Ratio0
Effect
Cognitive effects
Malek et al.
(2003c)
Pitten et al.
(2000)f
Malek et al.
(2003a)
Rat
Rat
Rat
LOAEL
LOAEL
LOAEL
130e
2,600
100e
2
0.17
2
1
7
7
1 day
90 days
10 days
130 x 2/4e x =65
2,600f
100 x 2/4e x 7/7 = 50
2
..
2
Concentration-
dependent
decreases in
activity by a variety
of measures
following a single
exposure
Impaired memory
in a spatial maze.
Magnitude of effect
increased with
continued exposure
through 12 weeks
Impaired learning
in a water maze.
Short-term (10 day)
exposure with
testing conducted 2
hours following
daily exposure.
§
VO
al mg/m3 = 0.813 ppm. All identified PODs were based on statistically significant findings at the study LOAELs. Full study details are provided in Section
4.1.1.6 (Bach et al., 1990) or 4.2.1.6 and Table 4-57 (all other studies).
bBoth actual levels of experimental exposures, and duration adjusted PODs are shown.
°POD unadjusted dose/duration-adjusted dose.
dTesting was conducted during or following exposure, duration was not adjusted.
Testing was conducted 2 hours postexposure; duration was adjusted to 4 hours to include the entire period between start of exposure and testing.
fDue to the uncertainty in continuous exposure adjustments and the unusually short (10 minutes) exposure in this study, no adjustment to continuous exposure is
presented.exposure, study design, and evaluation
O
O
-------
1 procedures, and all demonstrated robust findings of changes in nervous system structure or
2 function following formaldehyde exposure. All but one of the candidate studies present
3 information at multiple exposure levels to provide an understanding of the exposure response
4 relationship. One selected study (Senichenkova, 1991) provided less robust information, with
5 evaluation at only a single exposure level, but was considered useful as supporting the findings
6 of two other studies (Sarsilmaz et al., 2007; Asian et al., 2006) regarding neurological sequelae
7 of developmental exposure. All of the selected studies using experimental animals were
8 conducted in rats, although several studies in mice demonstrated dose-related neurotoxic effects
9 following formaldehyde exposure. These studies in mice were not considered for RfC
10 development because of the possibility that results might be confounded by reflex bradypnea at
11 the doses tested in each study; selected behavioral studies in rats were not similarly confounded
12 by reflex bradypnea because the effect occurs in rats only at doses above those at which the
13 effects of concern were seen (see Section 4.2.6 for details).
14 In order to improve transparency and facilitate comparison of health effect levels across
15 study types and health effects, Table 5-1 summarizes the PODs and exposure scenarios for each
16 selected study and describes the effects on which the selected POD is based. Dose conversions
17 used to adjust from actual experimental exposure concentrations to continuous exposure
18 concentrations are detailed. It should be noted that available studies providing dose-response
19 information regarding the effects of formaldehyde exposure on the nervous system were all of
20 short duration, and thus information regarding the relationship between formaldehyde toxicity
21 and exposure duration (i.e., whether toxicity increases with longer exposures at a given exposure
22 level, or is more related to the maximum exposure concentration) is limited. However, the
23 rodent study by Pitten et al. (2000) and the epidemiology study by Weisskopf et al. (2009)
24 provide strong support for an association between increasing neurotoxicity and increasing
25 duration of exposure.
26 Although chronic human studies are preferred for RfC derivation, no adequate human
27 study of chronic duration is available (see Section 4.1.1.6 for detailed discussion of available
28 human studies). The available human studies were sufficiently strong to raise concern regarding
29 formaldehyde effects on the nervous system; however, most did not provide sufficient exposure
30 information to permit derivation of a POD for use in quantitative dose-response assessment.
31 Available epidemiologic studies (most notably Weisskopf et al. [2009] and Kilburn et al. [1987,
32 1985]) provided limited exposure information. Weisskopf et al. (2009) reported a non-
33 statistically significant increase in the rate ratio for ALS for ever being exposed to formaldehyde
34 with RR=1.34 (95% CI: 0.93-1.92) among 987,229 people followed by an American Cancer
35 Society study, but no information regarding exposure concentrations was available. However,
This document is a draft for review purposes only and does not constitute Agency policy.
5-20 DRAFT—DO NOT CITE OR QUOTE
-------
1 when the cohort was restricted to people without missing data on duration of exposure to
2 formaldehyde, a statistically significant association was demonstrated as RR=2.47 (95% CI:
3 1.59-3.86, pO.OOOl). Weisskopf et al. (2009) also demonstrated statistically significant
4 increased exposure-response for risk of mortality from ALS associated with increased duration
5 of formaldehyde exposure (p = 0.0004). Interpretation of the findings of Kilburn et al. (1987,
6 1985) is complicated by concomitant exposure of many subjects to other solvents. Although the
7 chamber study by Lang et al. (2008) included a concentration-response assessment of changes in
8 reaction time, as previously discussed, the effects detected were difficult to interpret and the
9 study was not considered useful for RfC derivation.
10 One acute human study, Bach et al. (1990), which evaluated changes in cognitive
11 function following a single formaldehyde exposure, was considered for evaluation of a cRfC as
12 the chamber exposures were well defined and effects at multiple levels of exposure were
13 reported. In that study, concentration-related changes in short-term memory and ability to
14 concentrate were seen during a single 5.5-hour exposure at a range of levels (32, 170, 390, and
15 890 ppb). The study was designed as a comparison of effects of short-term formaldehyde
16 exposure in previously occupationally exposed individuals with effects in controls without
17 previous occupational exposure. Because occupational exposure levels were not assessed,
18 exposure measurements from the previously exposed workers are not appropriate for use in RfC
19 derivation. The authors reported a statistically significant exposure-response relationship for
20 three related cognitive measures (number of additions completed, number of errors, and reaction
21 time) in the 'addition test' assessment indicating a deficit in performance. Complete data were
22 not presented, but graphical presentations in the article indicated that the effect was seen at all
23 doses tested, with an apparent NOAEL of 170 ppb (see Figure 5-1).
24 No BMD modeling could be performed on these data because the graphical
25 representation could not be accurately digitized. The statistical analysis indicated no interaction
26 between formaldehyde effect and previous occupational exposure (i.e., the magnitude and
27 direction of the effect were similar in previously exposed and previously unexposed subjects)
28 and separate data were not presented for the two groups; thus, the LOAEL represents effects in
29 the combined study groups. Overall, the published paper lacks detail and it is difficult to
30 evaluate some aspects of the reported findings, in particular where magnitude and direction of
31 effect are not provided. Finally, the authors noted that controls and the high-exposure group
32 were not well matched on two key parameters (age and education level), adding uncertainty to
33 the reported exposure-response relationship (at the high dose). Although this study was
34 considered valuable in documenting neurological effects in humans following exposure to
35 relatively low concentrations of formaldehyde, the above concerns limit its utility for
This document is a draft for review purposes only and does not constitute Agency policy.
5-21 DRAFT—DO NOT CITE OR QUOTE
-------
1 quantitative human health risk assessment. Therefore, this study is not considered of sufficient
2 quality for RfC derivation.
3 In the absence of adequate human data, controlled studies in laboratory animals are
4 considered (see Section 4.2.1.6 for detailed discussion of available animal studies). There are no
3
cr
_. g,
°%
3 Q-
l i
fr§
<" en
tu
o.
0
300
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
O.S 1.0 -
Formaldehyde concentration
Figure 5-1. Change in number of additions made in 10 minutes following
formaldehyde exposure at 0.04, 0.21, 0.48, or 1.1 mg/m3 (32,170, 390, or 890
ppb).
Note: Vertical bars are the standard errors of the means, dashed line shows the
95% CI.
Source: Bach et al. (1990).
chronic studies and only one subchronic animal study evaluating neurological and behavioral
effects of formaldehyde exposure. Pitten et al. (2000) demonstrated impaired retention of a
previously learned task in rats exposed at concentrations of 2,600 or 4,600 ppb, 10 minutes per
day, 7 days/week, for 90 days (statistically significant, p<0.05). In this study, the magnitude of
the impairment increased over time, even though testing was performed 22 hours after exposure,
indicating that repeated formaldehyde exposure led to a worsening of effect. The study design,
test methods, and reporting of the results are all of adequate quality for both hazard assessment
and quantitative risk assessment. However, the short duration (10 minutes) of the repeated daily
exposures is a severe limitation to establishing a chronic RfC based on this study, due to
uncertainties in extrapolating from 10 minutes to a 24-hour exposure (see Table 5-1). Because
This document is a draft for review purposes only and does not constitute Agency policy.
5-22 DRAFT—DO NOT CITE OR QUOTE
-------
1 this study as designed indicates an accumulation of effect with repeated exposure, it is useful in
2 documenting the existence of a duration component to the exposure-response relationship. It
3 follows that concentration alone, without an adjustment for duration of exposure, would be
4 inadequate as an exposure metric; however inadequate information is available to inform the
5 appropriate magnitude of the duration effect. Therefore, although Pitten et al. (2000) is a
6 well-conducted study, the data are of limited utility for RfC derivation.
7 Finally, there are several well-documented acute and subacute animal studies that provide
8 exposure-response information for neurological and behavioral endpoints relevant for RfC
9 derivation. Several laboratory animal studies that evaluate neurological effects following in
10 utero or neonatal exposure address potentially susceptible life stages. Sarsilmaz et al. (2007) and
11 Asian et al. (2006) observed changes in brain structure (cell number and/or volume changes in
12 specific brain regions) following 30 days of exposure to neonatal rats (p<0.001). A related
13 finding by Senichenkova (1991) demonstrated changes in behavior (open field motor activity,
14 including habituation) in young rats following in utero exposure (p<0.05). Effects of concern
15 were seen at all doses in these studies, resulting in PODs of 67 ppb following in utero exposure
16 and 1,070 ppb following early postnatal exposure, based on LOAEL values adjusted for
17 continuous exposure (see Table 5-1). These studies support the possibility of
18 neurodevelopmental effects attributable to in utero or early postnatal formaldehyde exposure, at
19 levels similar to or below those causing other types of effects.
20 The other three studies in Table 5-1 evaluate behavioral changes in rats following
21 exposure to formaldehyde. Malek et al. (2003c) found concentration-related changes in motor
22 activity following a single 2-hour exposure at concentrations from 130-5,180 ppb (with testing
23 2 hours following cessation of exposure; p<0.005)). In a second study, Malek et al. (2003a)
24 found concentrated-related changes in performance on a learning task at similar exposure levels
25 (100-5,400 ppb) when 2-hour exposures were repeated for 10 consecutive days (p<0.05);
26 performance was evaluated 2 hours after cessation of exposure, and concentration-related
27 learning deficits were seen at all exposure levels (see Table 5-2 and Figure 5-2).
28 Although other studies evaluating neurobehavioral effects were available in the
29 formaldehyde database (see Chapter 4), these studies by Malek et al. (2003a, c) were considered
30 to be the most robust, documenting effects at relatively low exposure levels. Both studies also
31 included evaluation at multiple concentrations and showed concentration-related increases in
32 effect. In the Malek et al. (2003a) study with repeated exposures, it is unclear whether or not the
33 measured effect primarily reflects the most recent exposure or cumulative exposure; therefore,
34 the adjustment for continuous exposure was made over the final exposure period and the
35 two hours following exposure (4 hours total), as was done for the single-exposure study (Malek
This document is a draft for review purposes only and does not constitute Agency policy.
5-23 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
Table 5-2. Effects of formaldehyde exposure on completion of the labyrinth test by
male and female LEW.1K rats
Male rats
Control
O.lppma
0.5 ppm
5.4 ppm
Female rats
Control
0.1 ppm
0.5 ppm
5. 4 ppm
Swimming time (sec)
Dayl
105
100
97
105
Day 6
12.2
12.9
16.7C
25.7C
Day 10
6.33
6.07
7.60b
10.9C
Swimming time (sec)
Dayl
103
96
97
98
Day 6
12.5
12.3
14.6C
23. 5C
Day 10
6.47
7.53
7.60b
9.73C
Error rate (mean)
Dayl
7.4
7.7
7.6
7.7
Day 6
0.5
5.0C
4.4C
5.0C
Day 10
0.0
3.2C
1.8C
2.8C
Error rate (mean)
Dayl
7.9
7.1
8.0
7.9
Day 6
0
5.2C
4.6C
5.2C
Day 10
0.0
3.0C
2.2C
2.6C
aRats were exposed to formaldehyde for 2 hours/day, for 10 consecutive days.
bDifferent from control, p < 0.05.
"Different from control, p < 0.005.
Source: Malek et al. (2003a).
This document is a draft for review purposes only and does not constitute Agency policy.
5-24 DRAFT—DO NOT CITE OR QUOTE
-------
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
1 23456789 10
Day of labyrinth test
Figure 5-2. Effects of formaldehyde exposure on the error rate of female
LEW.1K rats performing the water labyrinth learning test
Source: Drawn from data reported in Malek et al. (2003a).
et al., 2003c). After appropriate duration adjustments, PODs for these studies range from 50 to
67 ppb (based on LOAELs), and the types of effects seen provide support for the Bach et al.
(1990) study that detected cognitive impairments in humans following a single exposure (with a
NOAEL of 170 ppb).
Summary of neurological and behavioral effects. In summary, the available studies for
formaldehyde and nervous system outcomes have demonstrated that the nervous system is a
sensitive target following inhalation of formaldehyde. In experimental animals, changes in
nervous system function were seen following acute and subchronic exposures; studies evaluating
neurological changes following chronic exposure were unavailable. Available human studies
that evaluated nervous system effects following inhalation exposure were found to have many
study-specific uncertainties and, thus, were not suitable to serve as the primary basis for a
chronic RfC. The Weisskopf et al. (2009) study of ALS, in particular, suggests that humans may
be at risk for severe neurological effects from formaldehyde exposure; however, this study
This document is a draft for review purposes only and does not constitute Agency policy.
5-25 DRAFT—DO NOT CITE OR QUOTE
-------
1 lacked the exposure concentration information necessary to derive an RfC. Neurological
2 findings from the rodent inhalation (acute and subchronic) studies that were judged to be
3 adequate for dose-response assessment identified unadjusted LOAELs ranging from 100 to
4 6,000 ppb, with LOAELs adjusted for continuous exposure in the range of 50 to 1,070 ppb. Use
5 of these PODs in risk assessment would require addressing uncertainties regarding animal-to-
6 human extrapolation, short study durations, and extrapolation from LOAELs.
7 Among the adequate studies, EPA considered Malek et al. (2003a) to be the most
8 appropriate for calculation of a cRfC for neurological and behavioral toxicity, based on the
9 exposure level at which effects were seen (100 ppb), the type of effect (impaired learning),
10 which is relevant to humans, and the use of a repeated-exposure paradigm (2 hours/day over a
11 period of 10 days), which addresses different exposure durations. This choice is supported by
12 similar effects seen in other studies (Lu et al., 2008; Pitten et al., 2000; Bach et al., 1990) and by
13 other neurologic effects seen at similar exposure levels (Malek et al., 2003c; Senichenkova,
14 1991; Sheveleva, 1971).
15
16 5.1.1.7. Developmental and Reproductive Toxicity
17 As described in Sections 4.1 and 4.2, both human epidemiologic data (see
18 Section 4.1.1.7) and experimental animal studies (see Section 4.2.7 and Tables 4-70 and 4-73)
19 demonstrate an association between formaldehyde inhalation exposure and adverse
20 developmental and reproductive effects, where adversity is characterized as per EPA risk
21 assessment guidelines (U.S. EPA, 199la, available at
22 http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=23162; 1996, available at
23 http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2838). Adverse outcomes were observed
24 across the various manifestations of developmental toxicity, including fetal death, structural
25 alterations (including congenital malformations), growth retardation, and functional
26 development. Additionally, in spite of the lack of a comprehensive database of studies for the
27 evaluation of the overall effects of formaldehyde on the reproductive system and its function, the
28 available evidence demonstrates toxicity to the male reproductive system in multiple animal
29 studies, as well as effects on the female reproductive system in both rodents and epidemiologic
30 studies, where an association with impaired fertility and increased spontaneous abortions were
31 noted.
32 Potential principal studies for specific adverse outcomes are presented and evaluated
33 below including reproductive effects (male and female), fetal death, growth retardation, and
34 structural alterations. The only available evidence for functional alterations is based on
35 developmental neurotoxicity studies which are presented and evaluated in Section 5.1.1.6.
This document is a draft for review purposes only and does not constitute Agency policy.
5-26 DRAFT—DO NOT CITE OR QUOTE
-------
1 Among the animal studies with developmental and reproductive effects after inhalation
2 exposure (presented in Tables 4-70 ad 4-73), 12 endpoints from nine studies were selected for
3 candidate PODs (see Table 5-3). The criteria for inclusion are that the studies provided reliable
4 documentation of exposure, study design, and positive findings of developmental or reproductive
5 structural, functional or precursor effects. Six of the studies evaluated effects after multiple dose
6 levels, providing dose-response information. The other three studies, with a control and a single
7 FA dose, were included as candidate PODs because effects were observed for endpoints that
8 were either not assessed or observed in the other six studies (e.g., cryptorchidism). Table 5-3
9 summarizes animal studies deemed suitable for deriving quantitative dose-response information
10 for reproductive and developmental outcomes and their corresponding PODs, adjusted for
11 continuous exposure. Calculations that were used in dose conversions and exposure duration
12 adjustments for the POD values are included. In general, repeated daily exposures of laboratory
13 animals are adjusted from a partial day to a 24-hour exposure and then weighted for the number
14 of days per week the exposures occurred. No chronic animal studies evaluating these endpoints
15 were available, so only subchronic and acute studies are considered.
16 The human epidemiologic data on developmental and reproductive outcomes are
17 discussed in Section 5.1.1.7.1. below. Exposure duration adjustments to the only suitable human
18 study (Taskinen et al., 1999) are more complex due to uncertainties in the exposure data and the
19 potential for nonoccupational exposures. For this discussion the reported 8-hour TWA
20 exposures will be used for the Taskinen et al. (1999) study. Further duration adjustments to this
21 study are discussed in Section 5.1.2.2.5 for cRfC derivation.
22
23 5.1.1.7.1. Spontaneous abortion and fetal death.
24 Increased risk of spontaneous abortion following maternal occupational formaldehyde
25 exposure was reported in a number of epidemiologic studies (Taskinen et al., 1999, 1994; John et
26 al., 1994; Seitz and Baron, 1990; Axelsson et al., 1984). The studies did not appear to be overtly
27 influenced by common principle biases found in epidemiologic studies. Considered together, the
28 studies are consistent with an adverse effect of formaldehyde exposure on pregnancy loss, where
29 adversity is characterized as per EPA risk assessment guidelines (U.S. EPA, 1991a, 1996). Of
30 these studies, Taskinen et al. (1999) had the higher quality quantitative exposure data reporting
31 reduced fecundity and spontaneous abortion in the exposed workers. Taskinen et al. (1999) is an
32 occupational study with a well-considered study design, including measurements of exposure
33 and outcomes, and relatively high study power. The study population consisted of 602 female
34 workers in Finland who had at least one successful childbirth and first employment in the
This document is a draft for review purposes only and does not constitute Agency policy.
5-27 DRAFT—DO NOT CITE OR QUOTE
-------
a 5 KO
o S. S3
*•
Table 5-3. Candidate fODs including duration adjustments for developmental and reproductive toxicity in key
animal studies
Reference
Species
POD
Type
ppba
Exposure scenario
Hours/
day
Days/
week
Duration
POD duration adjustments
Adjusted
POD Hours/ Days/ POD
(ppb) x day x week = (ppb)
Ratiob
Effect; comments
Spontaneous abortion and fetal death
Kitaev et al.
(1984)
Sheveleva(1971)
Rat
Rat
LOAEL
LOAEL
400
400
4
4
5
7
6 months
premating
CDs 1-19
400 x 4/24 x 5/7 =50 8
400 x 4/24 x 7/7 = 70
5.7
Increased (>threefold) embryo
degeneration on gestational
days 2-3 after 4 months
maternal premating treatment
Increased (50%)
preimplantation loss8
Structural alterations0
Senichenkova
(1991)
Senichenkova
and Chetobar
(1996)
Rat
Rat
LOAEL
LOAEL
400
400
4
4
7
7
CDs 1-19
CDs 1-19
400 x 4/24 x 7/7 = 70
400 x 4/24 x 7/7 = 70
5.7
5.7
Increased (13%) litter
incidence of internal organ
anomalies, including 20%
increase in undescended
testes; 9% decreased fetal
incidence of hyoid
ossification8
Increased (21%) fetal and
litter incidences of
cryptorchidism and increased
(6%) fetal incidences of total
anomalies8
Growth retardation
Saillenfait et al.
(1989)
Rat
BMCL
1,300
6
7
CDs 6-20
1,300 x 6/24 x 5/7 = 325
4
Decreased male fetal body
weights8 (BMR = 5%)
§
to
oo
fe
H
o
o
-------
a 5
o S.
*•* TO
rs i
KO
S3
Table 5-3. Candidate fODs including duration adjustments for developmental and reproductive toxicity in key
animal studies (continued)
Reference
Species
POD
Type
ppba
Exposure scenario
Hours/
day
Days/
week
Duration
POD duration adjustments
Adjusted
POD Hours/ Days/ POD
(ppb) x day x week = (ppb)
Ratiob
Effect; comments
Functional development d
Male reproductive toxicity
Ozen et al.
(2002)
Ozen et al.
(2005)
Sarsilmaz et al.
(1999)
Zhou et al.
(2006)
Rat
Rat
Rat
Rat
LOAEL
LOAEL
LOAEL
LOAEL
10,000
5,000
10,000
8,050
8
8
8
12
5
5
7
7
4 or 13
weeks
91 days
4 weeks
2 weeks
10,000 x 8/24 x 5/7 = 2,380
5,000 x 8/24 x 5/7 = 1,190
10,000 x 8/24 x 7/7 = 2,380
8,050 x 12/24 x 7/7 = 4,025
4.2
4.2
4.2
2
Decreased testis weight at
4 weeks (2%) and 13 weeks
/oo/\
(8/0)
Decreased (40%) serum
testosterone levels at 91 days
Decreased (5%) Leydig cell
numbers at 4 weeks
Decreased (-25%) testis
weight; alteration of
epididymal sperm [decreased
(38%) count, decreased (19%)
motility, and increased
(>3-fold) abnormal
morphology] at 2 weeks
Female reproductive toxicity
Kitaev et al.
(1984)
Rat
NOAEL
400
4
5
4 months
premating
400 x 4/24 x 5/7 = 50
8
Increased (-66%) follicle-
stimulating hormone at
4 months
§
to
VO
fe
H
O
o
al mg/m3 = 0.813 ppm. All identified PODs were based on statistically significant findings at the study LOAELs. The study details are provided in Section
4.2.1.7. and Tables 4-70 and 4-73. For Saillenfait et al. (1989), the BMCL was calculated (see "effect; comments" column above for details).
bPOD unadjusted dose/duration-adjusted dose.
°Neuropathological alterations following exposures during postnatal development (from the studies by Asian et al. [2006] and Sarsilmaz et al. [2007]) are
addressed in the neurobehavioral toxicity Section 4.2.6 and Table 5-2.
functional developmental endpoints (from the study by Senichenkova [1991]) are addressed in the neurobehavioral toxicity Section 4.2.6 and Table 5-2.
GDs = Gestation days
-------
1 wood-working industry beginning at least 6 months prior to the studied pregnancy. Mean daily
2 formaldehyde inhalation exposures during the time-to-pregnancy period were estimated for each
3 worker, based on task-level exposure measurements and work history.
4 Exposure was reported as a daily exposure index representing the average daily exposure
5 for the time-to-pregnancy period, and three exposure classes were defined as low, medium and
6 high with equivalent mean work-shift TWA exposure of 18, 76 and 219 ppb, respectively.
7 Fecundity density ratio (FDR) was significantly reduced in the high exposure group compared to
8 the referent group (FDR=0.64, 95% CI: 0.43-0.92, p = 0.02) indicating that it took longer for the
9 highly exposed women to become pregnant compared to women who were unexposed. The
10 investigators stratified the 39 women in the high exposure group by glove use. While
11 stratification by glove use reduced the statistical power of each comparison, the magnitude of the
12 effect in each strata was not markedly effected and the 95% confidence intervals of the
13 unstratified and stratified results all overlapped. Figure 5-3 shows the study results stratified by
14 glove use in women in the high-exposure group. While the adverse effect of high exposure to
15 formaldehyde was somewhat more pronounced among the women who did not wear gloves,
16 possibly suggesting that a component of dermal exposure might contribute to the effect, it is
17 unclear what, if any, dermal exposure is expected based on the nature of the work. Regardless,
18 there remains uncertainty as to whether effects are solely due to inhalation exposure. Taskinen
19 et al. (1999) also reported the risk of spontaneous abortions was statistically significantly
20 increased with reported ORs of 3.2 (95% CI: 1.2-8.3), 1.8 (95% CI: 0.8-4) and 2.4 (95% CI:
21 1.2-4.8) for the high, medium and low exposure groups, respectively. The finding of increased
22 risk of spontaneous abortion is consistent with the finding of delayed conception as measured by
23 the fecundity density ratio.
24 In some available rodent studies (Kitaev et al., 1984; Sheveleva, 1971), evidence of
25 increased embryo degeneration in early gestation or of preimplantation loss (findings that are
26 generally comparable to spontaneous abortion in humans) was observed. In the Kitaev et al.
27 (1984) study, early implantation losses resulted following treatment of dams prior to mating.
28 This may support a possible contribution of prepregnancy exposures to the spontaneous
29 abortions observed in Taskinen et al. (1999). Quantification of the findings by Kitaev et al.
30 (1984) and Sheveleva (1971) resulted in adjusted PODs of 50 and 70 ppb, respectively, based
31 upon study LOAELs (see Table 5-3).
32
33
This document is a draft for review purposes only and does not constitute Agency policy.
5-30 DRAFT—DO NOT CITE OR QUOTE
-------
1
1.4 -
1.2 -
1
0
a? 0.8
o>
E O-6
Q
LL.
0.4
0.2
0
(
»
<
i
All Women (N=39) No Gloves (N= 17) Gloves (N= 22)
2 Figure 5-3. Fecundity density ratio among women exposed to formaldehyde
3 in the high exposure index category with 8-hour time-weighted average
4 formaldehyde exposure concentration of 219 ppb (Taskinen et al., 1999)
5
6
7 5.1.1.7.2. Structural alterations.
8 Studies of occupational exposures to formaldehyde examined the incidence of congenital
9 malformations, but exposure and outcome data were not fully characterized and therefore could
10 not be carried forward to RfC development. Animal studies (Senichenkova and Chetobar, 1996;
11 Senichenkova, 1991) reported increases in internal organ anomalies; the most frequently
12 observed structural anomaly was a delay in fetal testis descent (at times characterized as
13 cryptorchidism in the study reports). For both studies, which exposed rats to formaldehyde for
14 4 hours/day during gestation, adjusted PODs based upon LOAELs were 70 ppb (see Table 5-3).
15 These studies included only one treatment level, precluding the ability to establish a dose-
16 response relationship, and the observed outcomes were not noted in other developmental toxicity
17 studies with similar exposure scenarios, thus limiting the strength of the studies for use in RfC
18 derivation.
19
This document is a draft for review purposes only and does not constitute Agency policy.
5.31 DRAFT—DO NOT CITE OR QUOTE
-------
1 5.1.1.7.3. Growth retardation.
2 Decreased fetal weight was observed in a number of animal studies that exposed pregnant
3 rats to formaldehyde during gestation. Of these, based on adequacy of dose-response
4 information, Saillenfait et al. (1989) was considered appropriate for consideration for RfC
5 development. In this study, rats were administered formaldehyde 6 hours/day on gestational
6 days (GDs) 6-20. Decreased male fetal body weight (BW) was modeled with a BMR of 5%
7 mean change, a BMCL was established, and, as shown in Table 5-3, the resulting duration-
8 adjusted POD of 325 ppb was derived. The relevance of this finding to human exposures was
9 qualitatively supported by a population-based study by Grazuleviciene et al. (1998) that reported
10 an association between atmospheric formaldehyde exposure and low birth weight; although a
11 dose-response relationship could not be adequately quantified from the information provided.
12
13 5.1.1.7.4. Male reproductive toxicity.
14 Evidence of adverse effects on male reproductive system endpoints following inhalation
15 exposure to formaldehyde was observed in a number of animal studies, where adversity is
16 characterized as per EPA risk assessment guidelines (U.S. EPA, 1991a, 1996). The effects
17 include decreased testes weight, changes in Ley dig cell quantity and quality, degeneration of
18 seminiferous tubules, decreased testosterone levels, alterations in biomarkers of toxicity in the
19 testes, and alterations in sperm count, morphology, and/or motility (Golalipour et al., 2007; Xing
20 et al., 2007; Zhou et al., 2006; Ozen et al., 2005, 2002; Sarsilmaz et al., 1999; Guseva, 1972).
21 Several of these studies included inhalation exposure of rats to formaldehyde 8 hours/day,
22 5 days/week for 4 and/or 13 weeks (Ozen et al., 2005, 2002; Sarsilmaz et al., 1999) and included
23 exposure-response information that was considered adequate for RfC derivation. In a study by
24 Ozen et al. (2002), increased severity of statistically significant testes weight decreases was
25 related to both dose and duration of treatment. Similarly, in the study by Golalipour et al.
26 (2007), seminiferous tubular diameter and epithelial height were reduced in rats following 18
27 weeks of formaldehyde inhalation exposure, with the severity of outcome positively correlated to
28 the number of hours/week that the animals were exposed. Sarsilmaz et al. (1999) noted dose
29 dependent decreases in Ley dig cell quantity after 4 weeks of treatment, while decreased testis
30 weight and atrophy of seminiferous tubules were observed by Zhou et al. (2006) after only
31 2 weeks of treatment. The reported outcomes in these independent studies illustrate a
32 biologically consistent toxicological profile of treatment-related male reproductive toxicity.
33 PODs, adjusted for continuous exposure, ranged from 1,190 to 4,025 ppb, where the lowest POD
34 was associated with the longest exposure period and vice verse (see Table 5-3).
35 5.1.1.7.5. Female reproductive toxicity.
This document is a draft for review purposes only and does not constitute Agency policy.
5-32 DRAFT—DO NOT CITE OR QUOTE
-------
1 Evidence of decreased fecundability was observed in the study by Taskinen et al. (1999),
2 which was described above for spontaneous abortions. Delays in the time to conception that
3 characterized this outcome, as well as increases in the incidence of endometriosis, were
4 statistically significantly associated with occupational exposures to formaldehyde. As these
5 effects were observed in the high exposure group, the unadjusted NOAEL for each of these
6 effects is 76 ppb (8 hour-TWA) based on the next lowest exposure group. Uncertainties
7 included lack of information human variability, as well as on the extrapolation of data from
8 studies of short duration to risk estimates for chronic exposures. As discussed above for
9 spontaneous abortions, the use of these data for cRfC derivation could result in values that would
10 likely be an underestimation of risk because they assume that all the risk was from inhalation
11 exposure and ignore the apparent contribution of dermal exposure (i.e., the dermal-exposure-
12 adjusted candidate inhalation RfCs might be higher). For decreased fecundability, a POD can
13 also be identified based on the data from only the women who wore gloves. The fecundability
14 density ratio (FDR) for the women in the highest exposure group was statistically significantly
15 reduced at FDR=0.64 (95% confidence interval [CI] 0.43-0.92). Evidence of spontaneous
16 abortions in the same study, as described above, may also be indicative of female reproductive
17 toxicity.
18 In animal studies, assessment of the female reproductive system was quite limited. An
19 increase in the mean follicle-stimulating hormone (FSH) levels in rats, observed at the highest
20 exposure level tested in Kitaev et al. (1984) was found to be sufficient to derive a duration-
21 adjusted POD of 50 ppb (see Table 5-3).
22
23 5.1.1.7.6. Summary of developmental and reproductive toxicity studies suitable for RfC
24 development.
25 A review of the developmental and reproductive toxicity studies in humans and animals
26 that would be suitable for cRfC development demonstrated that the developing organism and the
27 reproductive system are targets for toxicity following formaldehyde exposure by inhalation. In
28 the animal studies, effects during early development were observed following maternal
29 premating or gestational exposures at duration-adjusted PODs ranging from 50-325 ppb. The
30 minimal data available on female reproductive toxicity demonstrated an adjusted POD of 50 ppb
31 with subchronic (4-month) premating exposure, while more extensive evaluation of male
32 reproductive outcomes identified adjusted PODs of 1,190-4,025 for testicular and sperm
33 abnormalities after exposures of from 2 weeks to 3 months in duration. The animal studies
34 demonstrate the broad range of adverse outcomes to the reproductive system and the developing
35 organism following inhalation exposure to formaldehyde and highlight concerns regarding the
This document is a draft for review purposes only and does not constitute Agency policy.
5.33 DRAFT—DO NOT CITE OR QUOTE
-------
1 inadequacy of the database for the assessment of these outcomes (as described in Chapter 4).
2 These data also support the human relevance of female reproductive and/or embryonic and fetal
3 developmental effects, since some outcomes were similarly observed in both human and animal
4 studies.
5 The animal study data were not selected for RfC derivation, since a high-quality
6 epidemiology study (Taskinen et al., 1999) was available for the purpose of deriving a chronic
7 RfC. This study, a well-designed population-based case-control study of women who were
8 occupationally exposed to formaldehyde, included a well-defined study population which was
9 adequately selected to allow for meaningful comparisons of health effects among individuals
10 with different levels of exposure to formaldehyde. Potential confounding factors such a
11 selection bias and inaccurate self-reporting were not considered to have had a significant
12 influence on the study findings. The increased risk of spontaneous abortion observed in
13 Taskinen et al. (1999), and perhaps the observed decrease in fecundity, is internally consistent
14 and coherent with other reports of increased risk of pregnancy loss associated with exposure to
15 formaldehyde (John et al., 1994; Taskinen et al., 1994; Seitz and Baron, 1990; Axelsson et al.,
16 1984). It is also supported by similar adverse outcomes observed in the animal data (Kitaev et
17 al., 1984; Sheveleva, 1971).
18
19 5.1.2. Summary of Critical Effects and Candidate RfCs
20 5.1.2.1. Selection of Studies for Candidate RfC Derivation
21 The above reviews of data from both human and animal studies identified health effects
22 associated with formaldehyde exposure. Detailed information on these findings is given in
23 Chapter 4 (see Sections 4.1 and 4.2), and a qualitative summary of the noncancer hazard
24 identification is provided in Section 4.4 for each of the identified health effect categories:
25 sensory irritation, upper respiratory tract pathology, respiratory effects, increased atopic
26 response, immune function, reproductive and developmental toxicity, and neurobehavioral
27 toxicity. In this chapter, results for each health effect category are reviewed and studies are
28 identified which are adequate to inform the exposure-response relationship for health effects
29 from inhalation exposure (see Section 5.1.1). Although the database of published studies that are
30 currently available does not provide adequate quantitative data to derive cRfCs for all
31 qualitatively identified endpoints, at least one adequate study was identified for each of the
32 health effect categories discussed above. For all but one of the categories, at least one study was
33 available that provided epidemiologic (human) data, based on occupational or residential
34 exposures, which was judged adequate to provide a quantitative basis for a cRfC.
This document is a draft for review purposes only and does not constitute Agency policy.
5.34 DRAFT—DO NOT CITE OR QUOTE
-------
1 In order to select the principal study or studies most appropriate for use as the basis of the
2 RfC for formaldehyde, the relative merits of these studies were evaluated with respect to study
3 quality, characteristics of the study population, the quality and frequency of exposure
4 measurements, and the exposure levels at which effects are observed. The ideal RfC would be
5 derived from a reported exposure level without an appreciable risk of deleterious effects in
6 humans, including sensitive populations, with little uncertainty. Additionally, where possible,
7 the RfC should be derived with consideration of all of the identified health effects. The several
8 factors that were collectively taken into consideration for these studies (in no particular order)
9 included the following:
10
11 • Were studies of laboratory animals or humans?
12 - Human studies were generally preferred over laboratory animal studies for similar
13 health effects, when both were of good quality, given the uncertainties in interspecies
14 extrapolation.
15 • What was the study size?
16 - Larger studies were generally preferred over smaller studies because they can give
17 more precise estimates of response levels associated with specific exposure levels.
18 • Among the epidemiologic (human) studies, were exposures from an occupational setting
19 or from a residential setting?
20 - Studies of health effects from residential exposures were generally preferred over
21 studies of health effects from occupational exposures because residential exposures
22 tend to have a smaller range of variability and are less prone to large intermittent
23 exposure peaks.
24 - Residential exposures are more representative of the exposures of the general
25 population.
26 • Among the epidemiologic (human) studies, were children among the study population in
27 which health effects were observed?
28 - Studies of health effects that assessed the effect of formaldehyde on children's health,
29 representing a potentially more susceptible life-stage for some effects, were given
30 some preference because they provide formaldehyde-specific data relevant to the
31 components of the RfC derivation that address potentially sensitive life-stages and
32 populations.
33 • Relative to the other studies under consideration for RfC development, how accurately
34 were formaldehyde concentrations measured?
35 - Studies based on relatively more accurately measured formaldehyde concentrations
36 were generally preferred over studies that estimated exposures.
This document is a draft for review purposes only and does not constitute Agency policy.
5.35 DRAFT—DO NOT CITE OR QUOTE
-------
1 • Studies that reported effects at relatively lower formaldehyde concentrations, potentially
2 indicative of more sensitive endpoints, were generally preferred.
3
4 Taking all the factors into consideration collectively, the individual studies are presented
5 in Table 5-4.
6 For sensory irritation, four studies are identified with adequate exposure information for
7 RfC derivation, and all are observational studies of humans (see Table 5-4). Of these, 3 studies
8 were conducted in residential populations, including children and the elderly (Liu et al.,1991;
9 Ritchie and Lehnen, 1987; Hanrahan et al., 1984). Each of these studies includes in-home
10 formaldehyde measurements for each participant. Liu et al. (1991) provide the best exposure
11 measurements, with 7-day in-home passive air samples collected in two seasons. The
12 occupational study by Holmstrom and Wilhelmsson (1988) provides evidence of sensory
13 irritation in workers; however, only the mean and range of exposures for all workers is given.
14 Furthermore, occupational exposures can include high peak exposures. The residential studies
15 are preferred for development of candidate RfC. Although there are differences in study size and
16 the quality of exposure measurements between the three residential studies, their results are
17 mutually supportive, defining similar effect levels in similar populations, and the use of the
18 three residential studies was considered to provide adequate consideration of the sensory
19 irritation endpoint. Therefore, all 3 studies are selected (Liu et al.,1991; Ritchie and Lehnen,
20 1987; Hanrahan et al., 1984) and will be evaluated together in the following section.
21 Histological changes in the upper respiratory tract are well documented in animal studies
22 and have been observed in several worker studies (see Section 4.4). Although the study of resin
23 production workers (Holmstrom and Wilhelmsson, 1988; Holmstrom et al., 1989) provides the
24 best documentation of effect level for this health category in humans, it is not carried through for
25 development of a candidate RfC. As with the sensory irritation endpoint reported in these
26 studies, exposure is described for the worker cohort by a simple mean, with a range of exposures
27 given for all workers. Therefore, these data do not provide an exposure-response relationship
28 and the POD would be the mean exposure level of all workers, regardless of effect. This is less
29 exact than other available studies which provide exposure-response relationships. Additionally,
30 animal studies provide a broad database which supports sensory irritation as a more sensitive
31 endpoint than histological changes in the nasal mucosa.
32 Reduced pulmonary function is associated with formaldehyde exposure in several human
33 studies (students and workers). The best single study demonstrating decreased pulmonary
This document is a draft for review purposes only and does not constitute Agency policy.
5-36 DRAFT—DO NOT CITE OR QUOTE
-------
a 5
o S.
*•* TO
rs i
KO
S3
Table 5-4. Summary of candidate studies for formaldehyde RfC development by health endpoint category
Health
endpoint
category
Sensory
Irritation
Upper
Respiratory
Tract Pathology
Sensitization:
Asthma and
atopy
Study
Liu etal. (1991)
Ritchie and
Lehnen (1987)
Hanrahan et al.
(1984)
Holmstrom and
Wilhelmsson
(1988)
Holmstrom and
Wilhelmsson
(1988);
Holmstrom et al.
(1989)
Garrett et al.
(1999 a,b)
Species
Human
Human
Human
Human
Human
Human
Setting
Residential
Residential
Residential
Occupational
Occupational
Residential
Children
Yes
Yes
Yes
(teenagers)
No
No
Yes
Study
size
1,394
2,007
61
106
132
68 with
pathology
148
Formaldehyde
measurements
Two locations at one
time period (winter or
summer); 7-day
passive monitors
Two locations at one
time period;
30-minute sample
Two locations at one
time period;
60-minute sample
Several measurements
at factory
workstations taken
over 7 years
Several measurements
at factory
workstations taken
over 7 years
Four locations over up
to four time periods;
4-day passive
monitors
Specific
endpoints
Eye irritation
Eye, nose, and
throat sensory
irritation
10% increased
prevalence of
burning eyes
Eye irritation
Loss of ciliated
epithelium;
goblet cell
hyperplasia;
squamous cell
metaplasia
Increased
allergy;
increased
asthma-like
symptoms
Observed
effects"
(ppb)
95
200
130
210
240
28
POD (ppb)
LOAEL=95
NOAEL=50
BMCL10=70
NOAEL=70
LOAEL=240
LOAEL=28
§
fe
H
o
o
-------
a 5 KO
o S. S3
*•
Table 5-4. Summary of candidate studies for formaldehyde RfC development by health endpoint category
(continued)
Health
endpoint
category
Pulmonary
Function
Neurological
Reproductive
and
Developmental
effects
Immune
Function
Study
Krzyzanowski et
al. (1990)
Malek et al.
(2003a)
Taskinen et al.
(1999) (FDR)
Taskinen et al.
(1999) (SAB)
Lyapina et al.
(2004)
Species
Human
Rat
Human
Human
Human
Setting
Residential
Laboratory
Occupational
Occupational
Occupational
Children
Yes
—
No
No
No
Study
size
208
120
602
602
29
Formaldehyde
measurements
Four locations over
two time periods
(opposite seasons);
7-day passive
monitors
Intentional exposures
at specific levels
Actual and surrogate
measurements
estimated by
occupational hygienist
Actual and surrogate
measurements
estimated by
occupational hygienist
Average shift
concentrations based
on measures 8-hour
exposures
Specific
endpoints
10% Reduction
in PEFR
Impaired
learning
Decreased
fecundity
density ratio
(FDR)
Increased risk of
spontaneous
abortion (SAB)
Increased
respiratory tract
infections,
decreased
neutrophil
respiratory burst
activity
Observed
effectsa
(ppb)
27
100
226b
26b
722
POD (ppb)
BMCL10=17
LOAEL=100
NOAEL=86
LOAEL=26
LOAEL=722
§
oo
aThis is the lowest level of exposure at which adverse effects were observed, the LOAEL, in effect, or the cut-off point for adversity for BMCLs.
5 bSee Section 5.1.2.6.2 for methods to adjust exposure levels from Taskinen et al. (1999).
O
O
-------
1 function is the moderate residential study by Krzyzanowski et al. (1990). The study was
2 specifically designed to include homes with children between the ages of 5-15. Results
3 presented for children (n = 208) provide an exposure-response relationship for reduced PEFR.
4 Data quality is considered high for this study, both in terms of the in-home exposure
5 measurements (7-day passive monitors, two time periods) and the contemporaneous in-home
6 measurement of pulmonary function. Sources of potential confounding or bias were considered
7 by the study authors and adequately taken into account in the study. Therefore, this study is
8 retained for derivation of a candidate RfC.
9 Several studies report increased asthma and/or allergic sensitization in children
10 associated with increased formaldehyde exposure in school or homes (see Section 5.1.4). Of
11 these, two studies are further evaluated here (Garrett et al., 1999 a,b; Rumchev et al., 2002). The
12 study by Rumchev et al. (2002) is a case-control study of asthma incidence in children, and the
13 study by Garrett et al. (1999 a,b) is designed to study several related health effects (asthma,
14 sensitization and respiratory symptoms) in asthmatic and nonasthmatic children. Both studies
15 measure in-home formaldehyde levels with multi-day passive samples. Survey data and health
16 outcome data are considered of high quality in each study. Additionally, sources of potential
17 confounding or bias were considered by the study authors and adequately taken into account in
18 the study. Therefore, both studies are retained for derivation of a candidate RfCs. Although
19 several studies of school children support these findings, the residential studies were considered
20 more appropriate for RfC derivation because individual in-home formaldehyde levels were
21 associated with the health outcome data.
22 Multiple lines of evidence support the occurrence of neurotoxicity following exposure to
23 formaldehyde, however, none of the available studies in humans were considered to be of
24 adequate quality for derivation of a point of departure for use in quantitative assessment. Of the
25 available neurotoxicity studies, Malek et al. (2003a), in which impaired learning was seen in rats
26 following exposure at 100 ppb, was selected as a potential candidate for RfC development (see
27 Section 5.1.6). A NOAEL was not identified for this effect. In view of the other studies
28 available in the formaldehyde database (including multiple human studies of potentially sensitive
29 populations), and considering the uncertainty in extrapolating from the exposure conditions in
30 the Malek et al. (2003a) study (two hour exposures, repeated on ten consecutive days) to a
31 chronic exposure scenario, this study was not carried forward for derivation of a candidate RfC.
32 It is important to note that the resulting RfC may therefore not fully consider the documented
33 neurotoxic effects of formaldehyde.
34 Of the various reproductive and developmental effects associated with formaldehyde
35 exposure, reduced fecundity and increased risk of spontaneous abortions are primarily studied in
This document is a draft for review purposes only and does not constitute Agency policy.
5.39 DRAFT—DO NOT CITE OR QUOTE
-------
1 humans (see Section 5.1.7). Of the available epidemiology studies, only one study provides
2 individual exposure estimates of adequate quality to support RfC development (Taskinen et al.,
3 1999). Exposure-response relationships for decreased fecundability density ratio and increased
4 risk of spontaneous abortions are seen with increased categories of worker exposures. Several
5 potential confounding exposures are evaluated in the study, and the association of decreased
6 fecundability density ratio observed in the study is most convincingly associated with increased
7 formaldehyde exposure (Taskinen et al., 1999). Potential sources of bias were also adequately
8 addressed in the study. This is considered a high quality study and is retained for cRfC
9 derivation.
10 Although Lyapina et al. (2004) have documented decreased neutrophil respiratory burst
11 activity in exposed workers, the overall weight of evidence for deficit in immune function due to
12 formaldehyde exposure is weak. There is a trend for increased respiratory tract infections in
13 formaldehyde-exposed individuals, but it is a direct result of impaired immune function or,
14 perhaps, increased infection due to direct effects on the protective barriers of the nasal mucosa.
15 Animal studies do not support a finding of a deficit in immune function with formaldehyde
16 exposure. The study by Lyapina et al. (2004) is a small study, and the findings of decreased
17 neutrophil respiratory burst activity were in those individuals with more upper respiratory tract
18 infections, so there is some question of causality. The data evaluation does not provide an
19 exposure-response relationship, but, rather, exposure for the cohort is expressed as a mean
20 exposure of 722 ppb. Although the potential for impairment of immune function is an important
21 health effect, the overall evidence for this effect and this specific study are relatively weak
22 compared to other data available to support RfC derivation for formaldehyde. Therefore, this
23 study is not carried further in the quantitative analysis.
24 In summary, the best studies evaluated herein for the derivation of an RfC for
25 formaldehyde exposure and the related health effects are: 1) Sensory irritation (Liu et al.,1991;
26 Ritchie and Lehnen, 1987; Hanrahan et al., 1984); 2) reduced pulmonary function
27 (Krzyzanowski et al., 1990); 3) sensitization (atopy and asthma) (Garrett et al., 1999 a,b and
28 Rumchev et al., 2002); and 4) reduced fecundity and increased spontaneous abortion (Taskinen
29 et al., 1999). It is recognized that not all identified health effects are represented in these studies.
30
31 5.1.2.2. Derivation of Candidate RfCsfrom Key Studies
32 5.1.2.2.1. Candidate RfC derivation for Krzyzanowski et al. (1990) (Pulmonary function).
33 The study by Krzyzanowski et al. (1990) is a high quality epidemiology (human) study of
34 health effects in a random sample of residents and their families. The study was specifically
35 designed to include only households that had children 5-15 years of age, a sensitive life-stage for
This document is a draft for review purposes only and does not constitute Agency policy.
5-40 DRAFT—DO NOT CITE OR QUOTE
-------
1 respiratory effects. The study was of moderate size, when the effects in children were analyzed
2 separately from adults, with the final analysis based on 208 children—a cohort large enough to
3 show statistically significant results. The formaldehyde monitors were prepared by the
4 Lawrence Berkeley Laboratories and were considered to be precise and highly reliable. The
5 7-day passive formaldehyde monitors generally provide the lowest limit of formaldehyde
6 detection. The investigators specifically tested an a priori hypothesis and conclusively
7 demonstrated to a high level of statistical significance that increased residential formaldehyde
8 exposures were associated with decreased pulmonary function as measured by peak expiratory
9 flow rate (PEFR) in children. This effect was clearly shown at relatively low concentrations of
10 formaldehyde as the mean concentration in the homes was 26 ppb with more than 83% of homes
11 having measured concentration less than 40 ppb. This study also reported specific regression
12 modeling results that allowed EPA to calculate the point of departure for RfC development using
13 a BMCL as the point of departure.
14 The effects of formaldehyde exposure on pulmonary function represent a sensitive
15 endpoint with a reported 10% reduction in PEFR at 27 ppb. Among children with physician-
16 diagnosed asthma, the observed effects of increased formaldehyde exposure on decreased PEFR
17 were more pronounced—a clear indication of variability in response. The American Thoracic
18 Society (ATS, 2000) considers decreased pulmonary function an adverse health effect, even
19 when it is transient and subclinical. "Assuming that the relationship between the risk factor and
20 the disease is causal, the committee considered that such a shift in the risk factor distribution,
21 and hence the risk profile of the exposed population, should be considered adverse, even in the
22 absence of the immediate occurrence of frank illness" (ATS, 2000). The ATS (2000) stated that
23 individuals in an exposed population experiencing a shift in the distribution of pulmonary
24 function were at potential risk from another agent due to the reduction in their reserve capacity to
25 address additional insults. In the study by Krzyzanowski et al. (1990), the investigators
26 demonstrated statistically significant interaction between formaldehyde exposures, smoking, and
27 chronic cough. That is, a formaldehyde concentration that caused decreased pulmonary function
28 at residential levels also caused chronic cough in the presence of environmental tobacco
29 exposures. Higher prevalence rates of physician-diagnosed asthma and chronic bronchitis were
30 also shown at higher concentrations of formaldehyde (60-140 ppb), an effect that was
31 exacerbated by environmental tobacco exposures.
32 Figure 5-4 illustrates the reductions in peak expiratory flow rate (PEFR) in children
33 (<15 years of age) in relation to indoor residential formaldehyde concentrations estimated by a
34 random effects model based on 3,021 observations in 208 subjects. Formaldehyde levels in the
This document is a draft for review purposes only and does not constitute Agency policy.
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
QC
LJ_
LU
360
340
320
300
280
260
240
220
200
Bedtime
Morning, Non-Asthmatics
Morning, Asthmatics
0
20
40 60
HCHO(ppb)
80
100
Figure 5-4. Estimated reduction in peak expiratory flow rate (PEFR) in
children in relation to indoor residential formaldehyde concentrations.
Source: Krzyzanowski et al. (1990).
home were significantly related to reductions in PEFR in children both at bedtime and in the
morning (p < 0.05). PEFR measurements in the morning versus at bedtime were significantly
different (p < 0.05). Formaldehyde-related reductions in PEFR were greater in the morning in
asthmatic children than in nonasthmatic children (p < 0.05).
Candidate RfC derivation based on Krzvzanowski et al. (1990):
Critical effect: Based on this study, which specifically included a susceptible population,
the critical effect is reduction in PEFR in children. PEFR was the most sensitive measure
of disease or impaired lung function reported in this population, with decreases in lung
function reported in children who lived in homes with average measured formaldehyde
concentrations as low as 30 ppb (Krzyzanowski et al. (1990). Children were more
sensitive to formaldehyde-associated decreases in PEFR than adults, so the cRfC derived
focused on the results in the 208 children.
This document is a draft for review purposes only and does not constitute Agency policy.
5-42 DRAFT—DO NOT CITE OR QUOTE
-------
1 Point of departure: A BMR of 10% reduction in PEFR was selected as a cut-off point
2 for adversity, based on rationales articulated by the ATS (2000)4. Using this BMR and
3 the model coefficient in Table 5 of Krzyzanowski et al. (1990), a BMCLio of 17 ppb
4 (BMCio = 27 ppb) was derived for all children.5 Although the authors noted that
5 asthmatic children were more sensitive, the necessary data were not provided in the
6 report to calculate a BMCL for asthmatic children alone. Thus, 17 ppb, the BMCL based
7 on all children in the study, was used as the POD.
8
9 Application of study-specific Uncertainty Factors (VFs):
10 Interspecies UF = 1: No interspecies adjustment is needed, as this is a human study.
11 LOAEL-to-NOAEL UF = 1: Because a BMCL was used for the POD and the BMR of
12 10% reduction in PEFR was considered to be a cut point for adversity, no
13 LOAEL-to-NOAEL UF was needed (UFL =1).
14 Subchronic-to-chronic UF = 1: The study addresses ongoing residential exposure to
15 formaldehyde. Although information on the duration of exposure for each
16 participant is not provided, the residential nature of the study suggests a longer
17 term exposure than the duration of the study. It was judged that a population-
18 based study of residential exposures is sufficient to derive a chronic RfC without
19 adjusting for a subchronic observation period — at least for adults and older
20 children, and the children in this study were mostly older children (e.g., older than
21 7 years).
The ATS (2000) recommended that "a small, transient loss of lung function, by itself, should not automatically be
designated as adverse" and cited EPA's 1989 review of ozone, which offered a graded classification of lung function
changes in persons with asthma as "mild," "moderate," or "severe" for reductions of less than 10, 10-20, and more
than 20%, respectively (U.S. EPA, 1989). ATS (2000) concluded that, in evaluating the adverse health effects of air
pollution at the level of population health (compared to individual risk), "[a]ssuming that the relationship between
the risk factor and the disease is causal, the committee considered that such a shift in the risk factor distribution, and
hence the risk profile of the exposed population, should be considered adverse." This was specifically considered by
ATS (2000) even when "[e]xposure to air pollution could shift the distribution towards lower levels without bringing
any individual child to a level that is associated with clinically relevant consequences." A moderate adverse effect at
functional decrements of 10-20% was considered the best indicator of adverse effects in the study population. This
criterion had been similarly applied in EPA's Air Quality Criteria for Ozone and Related Photochemical Oxidants
(U.S. EPA, 2006d) for pulmonary function.
5 According to the regression model in Table 5 in Krzyzanowski et al. (1990), the coefficient ± standard error for
formaldehyde (in ppb) is -1.28 ± 0.46 and the background PEFR is 349.6 L/minute. Thus, a 10% reduction in PEFR
is -3 5 L/minute and the 95% (one-sided) upper bound on the slope for PEFR as a function of formaldehyde exposure
is -1.28 - (1.645 x 0.46), or -2.04 L/minute-ppb. Dividing 35 L/minute by 2.04 L/minute-ppb yields 17 ppb as the
BMCL.
This document is a draft for review purposes only and does not constitute Agency policy.
5.43 DRAFT—DO NOT CITE OR QUOTE
-------
1 Human variability UF = 3: The study was designed to include homes with children, and
2 a POD can be established based on reduced PEFR in children, who were more
3 sensitive to the health effects than the adults in the study. Therefore, the POD
4 represents data for a sensitive life stage, an aspect of human (intraindividual)
5 variability. With respect to the human (interindividual) variability UF, although
6 environmental tobacco smoke and socioeconomic status did not affect the
7 formaldehyde results in children, asthmatic children were more sensitive to the
8 effects of formaldehyde exposure on PEFR; thus, asthmatic children represent a
9 population with increased susceptibility for this effect. The prevalence rate for
10 physician-diagnosed asthma in the children was 15.8% in this study, which is
1 1 higher than the national prevalence of about 5.9% for ages 5 to 17 years.6 Thus
12 the BMCL based on all children may be influenced by a higher prevalence of
13 susceptible children for the critical effect. The authors do report that the PEFR
14 was reduced to a greater degree in asthmatic children (as shown in Figure 5-4),
15 and a lower BMC of 17 ppb can be calculated in this subgroup versus a BMC of
16 27 ppb for all children. However, the published regression statistics do not
17 provide sufficient detail to calculate a BMCL specific for asthmatic children. In
18 addition, other potentially sensitive populations (for example, elderly individuals
19 or individuals with respiratory diseases) may not be adequately represented in the
20 study. Therefore, an UF for human variability of 3 is applied to address the
21 observed increased sensitivity of asthmatic children in lieu of a calculated BMCL
22
23 RfC=7 - - ,= , ^ = 5.6PPb (5-1)
(UFAxUFLxUFsxUFH) (ixlxlxS)
24
25 UFA = 1 (interspecies UF)
26 UFL= 1 (LOAEL-to-NOAEL UF)
27 UFS = 1 (subchronic-to-chronic UF)
28 UFH = 3 (human variability UF)
29
The national prevalence rate of asthma in children ages 5-17 is according to the Centers for Disease Control and
Prevention (CDC) (MMWR 49(40):908-911, 2000). Although the Krzyzanowski et al. (1990) study was conducted
in the late 1980s, prevalence data from the National Health Interview Survey for 1997 were used for comparison
because that is the earliest year for which data are available after a 1997 redesign of the survey. Previously, the
survey asthma question was not specific for physician-diagnosed asthma, so the redesigned results were considered
to be more comparable to the physician-diagnosed asthma definition in the Krzyzanowski et al. (1990) study.
This document is a draft for review purposes only and does not constitute Agency policy.
5.44 DRAFT—DO NOT CITE OR QUOTE
-------
1 specific to asthmatic children and to ensure adequate protection for other
2 potentially sensitive populations.
3
4 5.1.2.2.2. Candidate RfC derivation for Rumchev et al (2002) (Asthma).
5 Residential formaldehyde exposure was associated with an increased risk of asthma in a
6 population-based case-control study of 192 children aged 6 months to 3 years (Rumchev et al.,
7 2002). While it is acknowledged that accurately diagnosing asthma in young children is
8 difficult, as the diagnosing physician was unaware of the formaldehyde level in the children's
9 home, any diagnostic error would be unrelated to formaldehyde concentrations and would not
10 induce a spurious association. It is noted that the endpoint is physician-diagnosed asthma. The
11 study, which comprises 88 cases of children discharged from the emergency department of a
12 children's hospital in Perth, Australia, with a primary diagnosis of asthma and 104 controls,
13 provides a positive exposure-response relationship adequate for RfC derivation. Seasonal in-
14 home formaldehyde measurements taken in the living room and subject's bedroom were used to
15 assess exposure (8-hour passive sampler). The ORs for risk of asthma by formaldehyde
16 exposure level category were adjusted for numerous risk factors, both familial and
17 environmental, including familial history of asthma, age, sex, socioeconomic status, smoking,
18 presence of pets, air conditioning, humidifier, and gas appliances. Of these, age, allergic
19 sensitization to common allergens, and family history of allergy were independent risk factors
20 for asthma (OR = 1.09, 2.57, and 2.66, respectively). Odds ratios were further adjusted for the
21 effects of the measured indoor air pollutants (see Rumchev et al., 2004), indoor allergen levels of
22 dust mites, relative humidity, and indoor temperature. Categorical analysis of the data indicates
23 that the ORs for asthma were increased in the two highest formaldehyde exposure groups,
24 reaching statistical significance for household exposures > 60 |ig/m3 (48 ppb) (OR = 1.39) (see
25 Figure 5-5). Analysis of the data with formaldehyde as a continuous variable provides a
26 statistically significant increase in the risk of asthma (3% increase in risk per every 10 |ig/m3
27 increase in formaldehyde level.)
28
29
This document is a draft for review purposes only and does not constitute Agency policy.
5.45 DRAFT—DO NOT CITE OR QUOTE
-------
1.6-
O
2
O
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
14-
1.2-
10-
08-
06J
10-29 30-49 50-59
Formaldehyde p.gm-3
60+
Figure 5-5. Odds ratios for physician-diagnosed asthma in children
associated with in-home formaldehyde levels in air.
Source: Rumchev et al. (2002).
Candidate RfC derivation based on Rumchev et al. (2002):
Critical effect: Diagnosis of childhood asthma (case-control study).
Point of departure: ANOAEL of 33 ppb (40 |ig/m3; midpoint of the 30-49 |ig/m3
category) was selected because the OR for asthma in the next highest exposure category
was considered to be part of an exposure-related trend of increasing asthma risk and,
therefore, biologically significant.
Application of Study-Specific Uncertainty Factors (VFs):
Interspecies UF = 1: No interspecies adjustment is needed as this is a human study.
LOAEL-to-NOAEL UF = 1: No LOAEL-to-NOAEL UF was needed because the POD
wasaNOAEL(UFL=l).
Subchronic to chronic UF = 3: The study addresses ongoing residential exposure to
formaldehyde. Although information on the duration of exposure for each
participant is not provided, the residential nature of the study suggests a longer
term exposure than the duration of the study. Study participants were 3 years or
younger, therefore the duration of exposure could not meet the expected
definition for a chronic study of one-tenth the lifespan. However, asthma often
develops during childhood, indicating a less-than chronic duration of exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
5-46 DRAFT—DO NOT CITE OR QUOTE
-------
1 Since asthma may develop throughout childhood it is unclear whether a study of
2 children under 3 years of age would be of adequate duration for this
3 developmental window. Therefore, an uncertainty factor of 3 was applied as a
4 subchronic to chronic adjustment.
5 Human variability UF = 1 or 3: As a case-control study, all new cases of childhood
6 asthma which met the study criteria were eligible for inclusion and the cases
7 likely included children predisposed to asthma. Individuals with a family history
8 of asthma and/or genetic markers for genes believed to predispose individuals to
9 asthma would represent a susceptible population. Therefore, the cases in this
10 study address children as a susceptible population for first diagnosis of asthma.
11 Additionally, there was an association of a familial history of asthma with the
12 diagnosis of children's asthma in this cohort (OR = 2.66). Not all sources of
13 human variability which may contribute to a diagnosis of asthma are known, and
14 there are likely additional sources of interindividual variability among children
15 and among individuals with a family history of asthma, thus it is unlikely that all
16 sources of human variability were adequately represented in the study population.
17
18 The two alternatives are described below and cRfCs are derived for each alternative.
19
Alternative A: Rumchev et al. (2002)
Human variability UF = 3:
To account for potentially susceptible individuals beyond those represented in the study
population, an uncertainty factor of 3 for human variability is applied.
„„ NOAEL 33ppb
RfC = -, r = 7 — r = 3.3
(UFAxUFLxUFsxUFH) (1x1x3x3)
UFA = 1 (interspecies UF)
UFL = 1 (LOAEL-to-NOAEL UF)
UFs = 3 (subchronic-to-chronic UF)
UFH = 3 (human variability UF)
20
21
22
This document is a draft for review purposes only and does not constitute Agency policy.
5.47 DRAFT—DO NOT CITE OR QUOTE
-------
Alternative B: Rumchev et al. (2002)
Human variability UF = 1:
EPA's Technical Report of the RfD and RfC Processes Technical Report (US EPA, 2002a)
indicates that UFH of 1 has been applied in cases where there are data "very specific
about the particular vulnerability of infants and children within certain age ranges to an
agent." Asthma and allergic sensitization to common allergens develop during
childhood and young adulthood defining a developmental window during which
individuals are most susceptible to the development of asthma. Since this study
includes only children up to 3 years of age, the UF for subchronic exposure is applied
above acknowledging that this study does not cover the susceptible developmental
window. No additional adjustment is applied for inter-individual variability among
children. It is acknowledged that additional sources of human variability are possible—
but it is believed that childhood is a key developmental window for initial diagnosis of
asthma. The technical report acknowledges that applying a UFH of 1 may be
appropriate where "even within these populations it is possible that some variability
still exists.
„„ NOAEL 33ppb
RfC = 7 r = 7 — r = 11
(UFAxUFLxUFsxUFH) (ixlxSxl)
UFA = 1 (interspecies UF)
UFL = 1 (LOAEL-to-NOAEL UF)
UFS = 3 (subchronic-to-chronic UF)
UFH = 1 (human variability UF)
1
2
3 5.1.2.2.3. Candidate RfC derivation for Garrett et al. (1999 a,b) (Asthma, respiratory
4 symptoms, atopy and severity of allergic sensitization).
5 Garrett et al. (1999 a,b) reported on the risk of allergy and asthma-like respiratory
6 symptoms due to formaldehyde exposure in a cross-sectional survey of households with children
7 7-14 years old with (n = 53) or without (n = 95) doctor-diagnosed asthma. Formaldehyde
8 exposure was characterized by four seasonal in-home sampling events using 4-day passive
9 samples collected in bedrooms, living rooms, kitchens, and outdoors. In logistic regressions,
10 both the prevalence and severity of allergic sensitization to 12 common allergens increased with
11 increasing formaldehyde concentration in the home. Additionally, a calculated respiratory
12 symptom score was increased and demonstrated a significant relationship with increased
13 formaldehyde concentration in a multiple linear regression after adjusting for multiple risk
14 factors and interactions. For each of these endpoints, severity/incidence was increased in the
15 medium (20-50 |ig/m3) and high (>50 |ig/m3) exposure groups relative to the low (<20 |ig/m3)
This document is a draft for review purposes only and does not constitute Agency policy.
5-48 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure group, based on the highest of four seasonal 4-day formaldehyde measurements in the
2 home (see Figures 5-6 and 5-7).
3 The findings of Garrett et al. (1999 a,b) are supported by the observation of an increased
4 bronchial responsiveness to mite allergen in a chamber study of 19 sensitized adult asthmatics
5 exposed to formaldehyde at a concentration of 100 |ig/m3 for 30 minutes (Casset et al., 2006).
6 Additionally, inhalation exposures to formaldehyde have been shown to increase an animal's
7 response to other common allergens via inhalation (Fujimaki et al., 2004b; Sadakane et al., 2002;
8 Riedel et al., 1996; Tarkowski and Gorski, 1995).
9
10 Candidate RfC derivation for increased allergic sensitization from Garrett et al. (1999 a,b):
11 Critical effects: Allergic sensitization—Increase in allergic sensitization (proportion of
12 atopic children). Severity of allergic sensitization measured both as number of positive
13 skin tests to common allergens and the recorded allergen wheal ratio for those tests.
14 Asthma—increase in proportion of asthmatic children. Respiratory
15 symptoms—Increased respiratory symptom score.
16
17 Point of departure: For all critical effects, categorical analyses are presented that show an
18 increase in the midexposure group (16-40 ppb) and high exposure group (>40 ppb) relative to
19 the low-exposure group (<16 ppb) (see Figures 5-6 and 5-7). However, it is unknown if the
20 findings in the low-exposure group are comparable to the responses that would be observed in an
21 unexposed population. Therefore, the low-exposure group cannot be considered a NOAEL but
22 rather serves as a referent group for the two other exposure groups. Thus, the LOAEL is based
23 on health effects observed in the midexposure group (16-40 ppb) for all three critical effects. As
24 neither the mean or median exposure levels are provided for the exposure categories used to
25 analyze the health effects data, the midpoint of the exposure category is selected for the LOAEL:
26 28 ppb.
27 Application of study-specific Uncertainty Factors fUFs):
28 Interspecies UF = 1: No interspecies adjustment is needed as this is a human study.
29 LOAEL-to-NOAEL UF = 3: As discussed, the midexposure group is selected as the
30 LOAEL since the low-exposure group is the referent group; there is no true
31 unexposed control. It is unclear whether or not a full LOAEL to NOAEL
32 uncertainty factor is warranted for these data. The authors did provide evidence
33 for increased atopy for every increase of 16 ppb of exposure with borderline
34 statistical significance when adjusted for several potential confounders (OR = 1.4;
This document is a draft for review purposes only and does not constitute Agency policy.
5.49 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
cnn/
•— /^o/v
^p ^O /o
0s
"*— ' /no/
0) QC;O/
— ^n50
Exposure group, formaldehyde level
(ug/m3)
40
2 35
8 30
a
£25
1 20
S 15
e
8 I0
0.5
00
•••
1
MMN
1
—
••i
WiP
i
i
i
< 20 ng/m3 20-50 pg/m3 > 50 jig/ro3
Higliest recorded formaldehyde level
Figure 5-6. Prevalence of asthma and respiratory symptom scores in
children associated with in-home formaldehyde levels. Trend analysis
indicates statistical significance in these increases {percent asthmatic
children, unadjusted (^=0.03) and respiratory symptom score (p=0.03)}.
Source: Garrett et al. (1999a); Garrett et al., 1999b (errata).
10
11
12
13
14
15
16
17
18
19
20
pnoA -7co/
_N /Uu/o
xO
SL* fino/.
c
SI 50% -
2
'5.
rf
1Uu/o
33%
b4%
•
<20 20-50 >50
Exposure group, formaldehyde level
(ug/m3)
Positive SPTs / allergen wheal ratio
0 — M W *>• l/>
0 p p p p 0
* Positive skin prick tests
Allergen whcal ratio
I
T-J
1 3
1 -
I
< 20 HS/ro1 20-50 jjg/nv1 > 50 (le/m1
Highest recorded formaldehyde level
Figure 5-7. Prevalence and severity of allergic sensitization in children
associated with in-home formaldehyde levels. Trend analysis indicates
statistical significance in these increases {percent atopic children (p = 0.002),
positive skin prick tests (p = 0.001) and severity as allergen wheal ratio
(p = 0.004)}.
Note: Skin prick tests included 12 environmental allergens (cat, dog, grass
[two types], house dust, dust mite [two strains] and fungi [five strains]).
Source: Garrett et al. (1999a); Garrett et al., 1999b (errata).
This document is a draft for review purposes only and does not constitute Agency policy.
5-50 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
95% CI: 0.98-2.00). An UF of 3 adjusts the LOAEL to a similar range and is
consistent with this alternative presentation of the data.
Subchronic to chronic UF = 1: The study addresses ongoing residential exposure to
formaldehyde. Although information on the duration of exposure for each
participant is not provided, the residential nature of the study suggests a longer
term exposure than the duration of the study. It is judged that a population-based
study of residential exposures is sufficient for derivation of a chronic RfC without
adjusting for a subchronic observation period.
Human variability UF = 1 or 3: This study was designed to assess allergic sensitization,
asthma prevalence and respiratory symptoms in children with relation to in-home
formaldehyde levels. The recruitment of participants was designed to include
households (50%) with asthmatic children, resulting in 43 households with at
least one asthmatic child and 37 without asthmatic children for a total of
148 children (35% asthmatic). Parental allergy and asthma were also assessed
and included as adjustment variables in the data evaluation. Therefore the study
population includes individuals reflecting several key aspects of human
variability for asthma and allergic sensitization (age, familial history of disease),
and addresses the links between allergic sensitization and asthma. Both asthma
and allergic sensitization are risk factors for increased respiratory symptoms.
The two alternatives are described below and cRfCs derived for each alternative
Alternative A: Garrett et al. (1999)
Human variability UF = 3: It is unclear whether the effect levels in the study truly
reflect the effect levels in sensitive populations, since study findings controlled for
both asthma and family history. Therefore, a value of three was used for the
human variability UF.
RfC =
LOAEL
_ 28 ppb
(UFAxUFLxUFsxUFH)~ (ix 3x1x3)
= 2.8 ppb
= 1 (interspecies UF)
UFL= 3 (LOAEL-to-NOAEL UF)
UFS = 1 (subchronic-to-chronic UF)
UFH = 3 (human variability UF)
23
This document is a draft for review purposes only and does not constitute Agency policy.
5.51 DRAFT—DO NOT CITE OR QUOTE
-------
Alternative B: Garrett et al. (1999)
Human variability UF = 1:
Individuals with a family history of asthma and/or genetic markers for genes are
believed to be predisposed to asthma and this would define a susceptible population within
children. In this study parental disease status is a marker for potential genetic susceptibility.
Although exposure-response relationships are not provided for individuals with a familial
history of disease, analyses provided suggest the results reflect responses from these
individuals. Among children with parental allergy, allergic children were exposed to higher
formaldehyde levels than non-allergic children (p = 0.02), relating higher formaldehyde
exposure to sensitization even among those with a likely genetic susceptibility. As shown in
Figure 5-8, formaldehyde levels are related to increased asthma incidence with a significant
linear trend (p = 0.02), yet this relationship loses significance when controlling for parental
allergy and asthma, suggesting the measured response on which the POD is based is driven by
children with a potential for genetic susceptibility.
An EPA Technical Report of the RfD and RfC Processes (US EPA, 2002a) indicates
that a UFn of 1 can be applied in cases where data are "very specific about the particular
vulnerability of infants and children within certain age ranges to an agent." Asthma and
allergic sensitization to common allergens develop during childhood and young adulthood.
Therefore no additional adjustment is applied for human variability. The technical report
acknowledges that "even within these populations it is possible that some variability still
exists", but that a UFH of 1 is still applied.
LOAEL 28ppb
RfC = 7 r = 7 ——r = 9.3 ppb
(UFAxUFLxUFsxUFH) (ixSxlxl)
UFA = 1 (interspecies UF)
UFL= 3 (LOAEL-to-NOAEL UF)
UFS = 1 (subchronic-to-chronic UF)
UFH = 1 (human variability UF)
1
2
3 5.1.2.2.4. Candidate RfC derivation for Ritchie and Lehnen, 1987; Hanrahan et al. ,1984
4 and Liu et al., 1991 (Sensory irritation).
5 There are three studies that report sensory irritation in humans from chronic exposures in
6 a residential environment and provide sufficient exposure data to support quantitative assessment
7 (Liu et al., 1991; Ritchie and Lehnen, 1987; Hanrahan et al., 1984). Each study reports site-
8 specific exposure measurements and presents some metric of individual exposure. These
9 residential studies employ in-home measurements for each study participant, either as average
10 exposure level (Ritchie and Lehnen, 1987; Hanrahan et al., 1984) or as calculated cumulative
11 exposure based on the time in the home (Liu et al., 1991). Eye irritation is reported at similar
12 levels of residential formaldehyde exposure in the three studies (see Figures 5-8 and 5-9). Each
This document is a draft for review purposes only and does not constitute Agency policy.
5-52 DRAFT—DO NOT CITE OR QUOTE
-------
1 study provides an exposure-response relationship for prevalence of sensory irritation in relation
2 to in-home formaldehyde exposure based on individual level data.
3 4 .5 6 7 B 9
FORnflLOEHYDE CONCENTRBTION IN PPn
= 100 100-300 >300
Formaldehyde Concentration (ppb)
4
5
6
Panel A: Regression of prevalence of eye irritation
versus indoor formaldehyde concentration (ppm) in
mobile homes (30—60 minute air sample in each
home).
Note: Dashed lines show upper and lower 95th percentile
confidence intervals on model results. Model based on
reported eye irritation from individuals in 42 mobile homes.
Source: Hanrahanet al. (1984).
Panel B: Prevalence of eye irritation in groups
defined by in-home formaldehyde exposure (30-
60 minute air sample in each mobile home).
Note: Eye irritation rate is given by smoking status:
active smokers (n = 143), passive exposure to smoke
(n = 133)andnonsmokers(w = 180).
Data source: Ritchie and Lehnen (1987).
Figure 5-8. Positive exposure-response relationships reported for in-home
formaldehyde exposures and sensory irritation (eye irritation).
This document is a draft for review purposes only and does not constitute Agency policy.
5.53 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
ZO/o
g
i/>
fl) ono/
> zu/o
12
Formaldehyde Exposure (ppm-hr per week)
Figure 5-9. Positive exposure-response relationships reported for in-home
formaldehyde exposures and sensory irritation (burning eyes).
Note: Cumulative formaldehyde exposure was estimated for each participant from
measured in-home formaldehyde levels (7-day passive air sample) and reported
hours spent in the home. Prevalence rates are given for both summer (n = 1,388)
and winter (n = 1,093) survey periods.
Data source: Liu et al. (1991).
Ritchie and Lehnen (1987) examined formaldehyde-associated effects on eye, nose, and
throat irritation in a large residential study with 2,007 participants from 841 homes. Based on
in-home measurements of formaldehyde concentration, participants were categorized into
three exposure groups: low (<100 ppb), mid (100-300 ppb) and high (>300 ppb) (average of
two 30-60 minute air samples per home). Ritchie and Lehnen (1987) observed clear exposure-
response relationships in the percentage of residential occupants reporting eye, nose, and throat
irritation. For example, in nonsmoking mobile home residents, incidence scores for eye irritation
were 1,-18% and 86%, and for nose/throat irritation were 5-17% and 78%, respectively, for the
three exposure groups. The exposure-response relationships were similar regardless of type of
home, mobile (n = 851) or conventional (n = 1,156). Although smoking status was also a
predictor of irritation, in-home formaldehyde concentrations were a stronger predictor of health
This document is a draft for review purposes only and does not constitute Agency policy.
5.54 DRAFT—DO NOT CITE OR QUOTE
-------
1 effects. The study included children and the elderly and results were consistent across age
2 groups. Children <7 years of age were only included in the eye irritation analyses because of
3 concerns about the quality of parental reporting for nose and throat effects in young children.
4 The selection criteria for participants indicate that more sensitive individuals may have been
5 over-represented in the study population.7 All study participants were self-selected, with a
6 physician's approval, perhaps resulting in a higher proportion of individuals experiencing
7 various irritant and upper respiratory tract symptoms, which may represent a sensitive population
8 for eye, nose, or throat irritation.
9 Hanrahan et al. (1984) reported an exposure-response relationship for burning eyes and
10 eye irritation in a study of 61 teenage and adult residents of mobile homes. As in the Ritchie and
11 Lehnen (1987) study, in-home formaldehyde measurements were obtained for all participants
12 and measured formaldehyde levels were used to characterize average in-home exposures
13 (30-60 minute air sample). Eye irritation was associated with in-home formaldehyde exposures
14 (p< 0.05) (both as "burning eyes" and "eye irritation"), and the authors provided a graphical
15 representation of the best-fitting regression model for exposures between 100 and 800 ppb.
16 From inspection of this graph, the prevalence of eye irritation predicted at 100 ppb is
17 approximately 4% with an upper bound of 18% (95th percentile CI) (see Figure 5-8, Panel A).
18 Because the limit of detection for formaldehyde in indoor air was 100 ppb, data or model results
19 are not provided below 100 ppb.
20 The third residential study is a random-sample study of over 1,000 mobile home residents
21 (1,394 in the summer; 1,096 in the winter) that included both young children and the elderly (Liu
22 et al., 1991). Cumulative weekly exposures were based on in-home formaldehyde sampling and
23 a participant survey of time spent at home. Air sampling was conducted for a 7-day period using
24 a passive sampler in each home (summer and winter). The resulting estimates of cumulative
25 exposure assumed no formaldehyde exposure outside of the home. Cumulative formaldehyde
26 exposure was a significant predictor of numerous irritant symptoms in a multivariate linear
27 logistic regression, including "burning eyes" (p < 0.05). The prevalence of eye irritation
28 increased with increasing cumulative exposure in a categorical analysis of participants
29 20-64 years old for both summer and winter exposure estimates (see Figure 5-9). Eye irritation
30 was above 10% in the lowest exposure group (0-7.0 ppm-hours/week) and increased to 17.1%
31 and 21.4% in the mid- and high-exposure group, respectively, for the summer survey time;
Participants in this study were self-selected residents who were concerned about possible formaldehyde exposure
and had obtained a written request from a physician to have the Minnesota Department of Health test their homes as
part of a free program; thus, people with symptoms may be overrepresented in this study compared with the general
population. This potential overrepresentation does not necessarily imply a selection bias because it is unlikely that it
was associated with the measured formaldehyde exposure levels in participants' homes.
This document is a draft for review purposes only and does not constitute Agency policy.
5.55 DRAFT—DO NOT CITE OR QUOTE
-------
1 winter rates were slightly lower but showed a similar increase with increasing cumulative
2 exposure.
3 Taken together, these three studies report increased eye irritation from residential
4 exposures that are below the BMCLs calculated from acute exposures in the laboratory. Each
5 study has the strength of having individual in-home exposure measurements and demonstrates a
6 positive exposure-response relationship for sensory irritation within a range of residential
7 formaldehyde exposures (both conventional and mobile homes). Potentially confounding factors
8 (such as allergens and some other in-home exposures) have been taken into account and
9 statistical analyses of the data include relevant covariates (e.g., age, sex, smoking status). As
10 such, these studies provide a basis for development of a cRfC for sensory irritation.
11 Additionally, the study populations have been drawn from the general population, including
12 children and the elderly, and have not been limited to those healthy enough for full-time
13 employment (as is often the case in occupational cohorts).
14 All three studies support a finding of increased eye irritation for exposures above 100 ppb
15 (see Figures 5-8 and 5-9). However, the shape of the exposure-response curve below 100 ppb,
16 or an indication of a no-effect level, is less clear. Two of the studies indicate 1-4% eye irritation
17 in residents where formaldehyde exposures were measured at 100 ppb or less (Ritchie and
18 Lehnen, 1987; Hanrahan et al., 1984). Thus, there is uncertainty in considering 100 ppb as a no-
19 effect level for increased eye irritation for these studies. When modeled, the 95% CIs around the
20 point estimate of 4% eye irritation were 1-18% eye irritation, illustrating the range of response
21 rates at 100 ppb that are consistent with the observed data (Hanrahan et al., 1984). Additionally,
22 the presentation of results by exposure category in Ritchie and Lehnen (1987) is inexact and has
23 individuals with exposures at the low end of the categorical range being grouped with those at
24 higher exposures in the range, obscuring any exposure-response relationship within the
25 categorical range. For these reasons, a POD for RfC derivation from either of these studies
26 should reflect these uncertainties. Therefore, for the NOAEL representing the category of
27 individuals with <100 ppb, in which 1-2 % eye irritation was observed, the upper end of this
28 exposure category is not used, but rather the midpoint, 50 ppb (Ritchie and Lehnen, 1987).
29 Although Hanrahan et al. (1984) provided no model results below 100 ppb, an extrapolation of
This document is a draft for review purposes only and does not constitute Agency policy.
5-56 DRAFT—DO NOT CITE OR QUOTE
-------
1 the graphical results (see Figure 5-8, Panel A) provides an estimated BMCLio of 70 ppb . No
2 additional duration adjustments were made from the in-home exposure measurements to
3 continuous exposure because neither time away from the home, nor potential exposures outside
4 of the home, were characterized in either study.
5 Of the three studies, only Liu et al. (1991) provides exposure measurements below
6 100 ppb, with a reported detection limit of 10 ppb formaldehyde for the in-home air monitoring.
7 Additionally, air samples were collected using a 7-day passive sampler which is more
8 representative of average residential exposures than a one-time, 30-60 minute, air sample.
9 Therefore, the data collected by Liu et al. (1991) are more suited to understanding the exposure-
10 response relationship for eye irritation of exposures below 100 ppb. In addition to controlling
11 for age, gender, and smoking status, Liu et al. (1991) controlled for the presence of chronic
12 respiratory disease when assessing the effects of formaldehyde on symptoms of sensory
13 irritation. Finally, this study provides results for both summer and winter survey periods,
14 addressing seasonal variation in both formaldehyde levels and sensory irritation. The use of the
15 cumulative exposure metric considers not only the concentration of formaldehyde but also the
16 number of hours during the week each participant spent in their residence. Linear logistic
17 regression indicates that cumulative formaldehyde exposure was a statistically significant
18 predictor of burning eyes for both winter and summer survey periods. However, no BMCL can
19 be calculated because no regression coefficients were provided in the report. Data were
20 provided for the categorical analysis illustrating a positive exposure-response relationship
21 (redrawn in Figure 5-9). Based on the categorical results, the midexposure group
22 (7-12 ppm-hours/week) demonstrated an increased response compared with the low-exposed
23 group. Since the prevalence rate in the low-exposed group was above 10% for burning eyes, this
24 exposure group does not represent a NOAEL, but rather serves as a referent for the midexposure
25 group. Therefore, the POD is derived from the midpoint of 7-12 ppm-hours/week,
26 9.5 ppm-hours/week. Using a conversion factor applied by the authors, the cumulative exposure
27 of this midexposure group
Figure 1 of Hanrahan et al. (1984) shows predicted values and 95% confidence intervals (CIs) for the percent
prevalence of a burning-eyes response for formaldehyde concentrations >100 ppb (See Panel A in Figure 5-9 above).
A short extension of the upper 95% CI to the concentration associated with 13% prevalence (i.e., a 10% increased
prevalence above an assumed background response rate of 3%; this assumed background rate was chosen to be
conservatively high to err on the side of not underestimating the actual value, given that the value was approximated
from a visual extension of the upper 95% CI curve) suggests a BMCL of approximately 70 ppb for 10% increased
prevalence. The actual value is unknown but is clearly below 100 ppb, which is the minimum exposure
concentration depicted in the figure.
This document is a draft for review purposes only and does not constitute Agency policy.
5.57 DRAFT—DO NOT CITE OR QUOTE
-------
1 corresponds to a continuous home exposure of 70-120 ppb for an individual who spends 60% of
2 the week in the home, with a midpoint of 95 ppb.
3
4 Candidate RfC derivation for sensory irritation:
5 Critical effect: Prevalence of sensory irritation (eye irritation, burning eyes).
6
7 Point of departure: Each of the studies discussed above has different strengths and
8 weaknesses for the determination of a POD for sensory irritation. Nevertheless, the
9 effect levels and PODs derived from each study are in relatively close agreement with
10 less than a twofold span from lowest to highest. Therefore each POD is carried through
11 to calculate a cRfC:
12
13 NOAEL = 50 ppb (Ritchie and Lehnen, 1987)
14 BMCLio = 70 ppb (Hanrahan et al., 1984)
15 LOAEL = 95 ppb (Liu et al., 1991)
16
17 Application of Uncertainty Factors fUFs)
18 Interspecies UF = 1: No interspecies adjustment is needed as this is a human study.
19 LOAEL-to-NOAEL UF: An uncertainty factor of 1 is applied to the NOAEL and
20 BMCLio established as PODs from Ritchie and Lehnen, (1987) Hanrahan et al.
21 (1984) studies. An uncertainty factor of three is applied to the LOAEL of 95 ppb
22 based on the Liu et al. (1991) study, as the prevalence rates for this exposure level
23 are below 20% for an effect that is of relatively low severity. In addition, the
24 LOAEL is not significantly above the NOAEL and BMCLio from the other
25 studies that evaluated the same endpoint.
26
27 Subchronic to chronic UF = 1: These studies address ongoing residential exposure to
28 formaldehyde. Although information on the duration of exposure for each
29 participant is not provided, the residential nature of the study suggests a longer
30 term exposure than the duration of the study. It is judged that a population-based
31 study of residential exposures is sufficient for derivation of a chronic RfC without
32 adjusting for a subchronic observation period.
33
34 Human variability UF = 1 or 3: All three studies were population-based and included
35 children, the elderly and both sexes. Sample sizes for two of the studies were
This document is a draft for review purposes only and does not constitute Agency policy.
5-58 DRAFT—DO NOT CITE OR QUOTE
-------
1 very large (1,394 for Liu et al. [1991]; 2,007 for Ritchie and Lehnen [1987]),
2 increasing the likelihood that sensitive populations were included. Analysis of
3 the data controlled for sex, smoking status, and age group.
4
5 The two alternatives are described below and cRfCs derived for each alternative
6
9
10
Alternative A
Sensory irritation studies:
Human variability UF = 3: For all studies, the analysis was based on prevalence rates,
decreasing the likelihood that effects on sensitive individuals would be lost due to
response averaging. For Ritchie and Lehnen (1987), the prevalence rate in the
<100 ppb exposure group (represented by a NOAEL of 50 ppb, the midpoint) was
1-4%. For Hanrahan et al. (1984), the POD is a BMCL corresponding to a
10% response rate. Given these prevalence rates and the fact that the sensory
irritation effects assessed are considered minimally adverse, a human variability
UF of 3 was considered adequate for this endpoint.
Ritchie and Lehnen (1987):
NOAEL 5Qppb 1a ,
RfC = 7 - r = 7 - — — r = 17 ppb
(UFAxUFLxUFsxUFH) (ixlxlxS)
Hanrahan et al. (1984):
23
7 _ _
(UFAxUFLxUFsxUFH) (ixlxlxS)
Liu etal. (1991):
D_ LOAEL 95 ppb
RfC = 7 - r = 7 - — — r = 9.5 ppb
(UFAxUFLxUFsxUFH) (ix 3x1x3)
This document is a draft for review purposes only and does not constitute Agency policy.
5.59 DRAFT—DO NOT CITE OR QUOTE
-------
Alternative B
Sensory irritation studies
Human variability UF = 1: Two studies included a broad age range allowing some
assessment of human variability due to life stage. Ritchie and Lehnen (1987)
evaluated the influence of age on sensory irritation in the following age groups
65 years.
An age effect for eye irritation was not evident in these data and pooled data are
presented for this endpoint. Liu et al. (1991) report that greater eye irritation was
reported in participants of 20-64 years than in those younger than 20 or older than
65 years. The elderly population (>65 years) was well-represented in this study
(39% of participants in the summer and 34% in the winter). The modeled results on
which the BMCLio is based for Hanrahan et al. (1984) are normalized to 48 years of
age (the mean age of respondents), which is consistent with the age group considered
the most responsive in the Liu et al. (1999) study. Therefore the PODs derived from
these studies do account somewhat for human variability across the life stage.
The critical effects of sensory irritation (eye, nose, and throat irritation) are considered
minimally adverse health effects. The nominal response rates for eye irritation of
1-4% for in-home exposures below 100 ppb from which the PODs were derived
suggest that the PODs are below significant response levels. Additionally, as the data
are reported as prevalence rates, there is no masking of effect from sensitive
individuals (as may occur when benchmark responses are average values of biometric
parameters).
Finally, sensory irritation is a POE effect. Therefore, sources of human variability such as
absorption, distribution, and metabolism of a compound are unlikely to influence
incidence rates for this endpoint. There may be human variability in the sensitivity of
the trigeminal nerve to formaldehyde binding and stimulation.
Taken together, these studies address many potential sources of human variability.
Therefore, it is judged that further adjustment to address human variability is not
warranted for the minimally adverse health effect of sensory irritation. Thus a UFH of
1 is applied to all three studies. It is acknowledged that there is the potential for
sources of variability not captured in these studies.
Ritchie and Lehnen (1987):
JHT N°AEL 5°PPb
-------
1 5.1.2.2.5. Candidate RfC derivation for Taskinen et al. (1999) (Fecundity density ratio).
2 On review of the candidate developmental and reproductive toxicity studies in humans
3 and animals (presented in Section 5.1.3.2.7), the Taskinen et al. (1999) epidemiology study was
4 considered to be the strongest for the purpose of deriving a chronic RfC. This study was a well-
5 designed population-based case-control study of women who were occupationally exposed to
6 formaldehyde. The study population was well defined and adequately selected to allow for
7 meaningful comparisons of health effects among individuals with different levels of exposure to
8 formaldehyde. Potential selection bias and the self-reporting of spontaneous abortion are not
9 considered to have had a significant influence on the study findings. Additionally, the decreased
10 FDR and increased risk of spontaneous abortion observed in Taskinen et al. (1999) are internally
11 consistent and coherent with other reports of increased risk of pregnancy loss associated with
12 exposure to formaldehyde (John et al., 1994; Taskinen et al., 1994; Seitz and Baron, 1990;
13 Axelsson et al., 1984) and is supported by animal data (Kitaev et al., 1984; Sheveleva, 1971).
14 The Taskinen et al. (1999) study allows the consideration of three potential critical
15 effects: endometriosis, increased spontaneous abortion, and decreased fecundity density ratio
16 (FDR). However, there is little independent support for the finding of increased risk of
17 endometriosis and the ORs for organic solvent exposure within this study (OR = 14.7; 95% CI:
18 3.1-70) were much greater than for formaldehyde (OR = 4.5, 95% CI: 1.0-20), indicating a
19 reasonable potential for confounding of the formaldehyde association. The finding of increased
20 risk of spontaneous abortions is supported by independent findings in other formaldehyde-
21 exposed cohorts (John et al., 1994; Taskinen et al., 1994; Seitz and Baron, 1990; Axelsson et al.,
22 1984). As this study was designed to examine the effect of workplace formaldehyde exposures
23 on FDR, the study design and data collection best support this finding. The exposure estimates
24 were assessed to represent what the researchers considered the relevant exposures for evaluating
25 risk factors that might influence time-to-pregnancy. Although data on miscarriages were
26 collected to control the time-to-pregnancy findings for potential confounding from
27 formaldehyde-related spontaneous abortions, it is less certain that the exposure measurements
28 coincide with the defined spontaneous abortion cases. Spontaneous abortions were only
29 included in calculations of exposure-specific ORs if a participant indicated that she was
30 employed at the same location when she had the spontaneous abortion and when the time-to-
31 pregnancy exposure assessment was done. The analysis showed that there were statistically
32 significantly increased risks of spontaneous abortion in the lowest exposure group. While this
33 finding was consistent with other studies showing adverse reproductive effects of formaldehyde
34 and appears to be causal, the Taskinen et al. (1999) spontaneous abortion results did not clearly
35 control for all the potential confounders that were controlled for in the FDR analyses (i.e.,
This document is a draft for review purposes only and does not constitute Agency policy.
5-61 DRAFT—DO NOT CITE OR QUOTE
-------
1 organic solvents and phenols). While the other coexposures were not associated with FDR and
2 therefore not confounders, endometriosis was strongly associated with organic solvents.
3 Therefore, for these endpoints, the study design and strength of results best support the use of
4 decreased FDR in formaldehyde-exposed women as the critical effect for this study.
5 It is preferable that the critical effect for a specific study be the most sensitive of the
6 effects which is well supported by the study. As spontaneous abortions are significantly
7 increased in the low-exposure group and the response in the midexposure group is considered a
8 no-effect level for decreased FDR, there is uncertainty that an RfC based on the FDR NOAEL
9 would be protective for the more sensitive effect. Although the finding of increased risk of
10 spontaneous abortion is qualitatively convincing, there is more uncertainty in the applicability of
11 the exposure assessment for quantitative risk assessment. Additionally, there is greater
12 uncertainty in the use of the exposure adjustments for the low-exposure group on which the
13 LOAEL for spontaneous abortion is based because the exposure adjustments account for more of
14 the work time in the low-exposure group than the medium and high exposure groups (see
15 Table 5-5).
16 There are several sources of uncertainty in the exposure estimates for use in RfC
17 derivation. As discussed above, the average exposure estimate for the low exposure group
18 includes a greater proportion of nonassessed background exposures. This is evidenced in part by
19 the reported average exposure being below background levels for these workers, even with
20 exposure measurements as high as 300 ppb. The unaccounted for nontask exposures may
21 represent time during the day spent in the work facility, or time in a different job or work
22 environment. Additionally, task-level exposure measurements were available for only 27% of
23 women in the low exposure group, versus 38% and 69% of women in the medium and high
24 exposure groups, indicating less certainty in exposure classification for the low exposure group.
25
26 Duration adjustment for candidate study points of departure. Normally, exposures from
27 occupational studies are adjusted to account for the daily breathing volume appropriate to an
28 environmental (versus occupational) setting and for exposure every day of the year (U.S. EPA,
29 1993). However, with formaldehyde, there is potential for exposure outside of work from in-
30 home and environmental sources of formaldehyde (Chapter 2). A contemporaneous study of
31 formaldehyde exposures in Finland reports average exposure of 21.4 ppb (measured over
32 48 hours with a personal monitor) (Jurvelin et al., 2001). Furthermore, both the mean exposure
33 (18 ppb 8hr TWA) and lowest reported exposure (10 ppb 8hr TWA) of the 'low exposed'
34 category are below the reported average ambient exposures for Finland (21.4 ppb). Thus, it is
This document is a draft for review purposes only and does not constitute Agency policy.
5-62 DRAFT—DO NOT CITE OR QUOTE
-------
2
3
4
5
6
7
8
9
10
1 1
12
13
14
Table 5-5. Adjustment for nonoccupational exposures to formaldehyde.
Panel A: Proportion of workshift corresponding to the exposure group mean task-
level formaldehyde exposure (ppb) and the exposure group daily exposure index
(8 hour-TWA).
Exposure
group («)
Low (119)
Medium (77)
High (39)
Reported mean
exposure
(ppb, 8 hr-TWA)
Mean
18
76
219
Range
1-39
40-129
130-630
Measured task-
level exposures
(PPb)
Mean
70
140
330
Range
10-300
50-400
150-1,000
Estimate of time during
workday for formaldehyde
related tasks assuming mean
exposure levels.
%of
worktimea
26%
54%
66%
Hours per
8 Hr workshift
2
4.3
5.3
"Calculated as mean exposure (ppb 8 hour-TWA) divided by mean task-level exposures for the exposure group.
Panel B: Recalculation of daily exposure index (8 hour-TWA) where
background formaldehyde exposure is estimated for worktime spent on tasks
considered unrelated to occupational use of formaldehyde.
Exposure
group («)
Low (119)
Medium (77)
High (39)
Estimate of formaldehyde
exposure during formaldehyde-
related work tasks
Mean task
level
exposure
(PPb)
70
140
330
% of worktime in
formaldehyde task
26%
54%
66%
Estimate of formaldehyde
exposure from background
levels during the workshift
Background
formaldehyde
(PPb)
21.4
21.4
21.4
% of time in
nonformaldeh
yde-related
task
74%
46%
34%
Alternative
daily
exposure
index
(ppb, 8 Hr-
TWA)
34
86
226
15
16
17
18
19
20
21
likely that exposure estimates for study participants include time during the workday when
women reported no formaldehyde exposure and a zero exposure was assessed for a
nonformaldehyde related task. Additionally, participants may have qualified for the study based
on employment date but may not have been working with formaldehyde during the entire time-
to-pregnancy period. In both cases, the investigators in Taskinen et al. (1999) appear to have
This document is a draft for review purposes only and does not constitute Agency policy.
5-63 DRAFT—DO NOT CITE OR QUOTE
-------
1 assumed that, while the women were away from their "exposed" workplace, their exposure to
2 formaldehyde was zero, not accounting for background occupational exposures and ambient
3 levels of formaldehyde. This explains why both the mean exposure as well as lower end of
4 workshift exposures for women in the low exposure group were reported at and below expected
5 ambient levels. The women in the low exposure category had task-level workplace exposures of
6 up to 300 ppb in addition to experiencing some work time at background exposure levels.
7 Compared to women who only experienced background exposure levels, those in the low
8 exposure category were at significantly higher risk of spontaneous abortion.
9 The reported data do not provide information to correct for background formaldehyde
10 exposure during the workday for each participant. However, the published mean exposure
11 values may be used to provide some idea of the impact of including background exposures on
12 the study PODs. Comparison of the values listed in Table 4 of Taskinen et al. (1999) allows for
13 the estimation of the percentage of work time spent performing tasks involving formaldehyde
14 exposure (see Table 5-5, Panel A). For the women in the low exposure category, this percentage
15 is 26% (mean of measured workplace exposures of 70 ppb times 26% equals the mean of the
16 TWA exposure of 18 ppb). Using the same method, the women in the "medium" and "high"
17 exposure category were performing tasks involving formaldehyde exposure approximately 54%
18 and 66% of their work time, respectively. Assuming that the women spent the remainder of their
19 work time at the background concentration of 21.4 ppb (Jurvelin et al., 2001), a more appropriate
20 estimate of the women's 8-hour TWA formaldehyde exposures would be 34 ppb for the low
21 category, 86 ppb for the medium category, and 226 ppb for the high category (see Table 5-5,
22 Panel B).
23
24 Candidate RfC derivation for Taskinen et al. (1999):
25 Critical effect: Decreased FDR.
26
27 Point of departure: For decreased FDR, the midexposure level is considered a NOAEL.
28 The mean exposure as an 8-hour TWA for the workday is reported as 76 ppb. EPA has
29 adjusted this POD to account for potential background formaldehyde exposures during
30 the workshift (see Table 5-5) resulting in an adjusted POD of 86 ppb. No further
31 duration adjustment is made to this POD to account for background levels of
32 formaldehyde exposure outside of the workplace.
33
This document is a draft for review purposes only and does not constitute Agency policy.
5-64 DRAFT—DO NOT CITE OR QUOTE
-------
1 Application of study-specific Uncertainty Factors fUFs):
2 Interspecies UF = 1: No interspecies adjustment is needed as this is a human study.
3 LOAEL-to-NOAEL UF = 1: Selection of an NOAEL as the POD.
4 Subchronic to chronic UF = 1: The study design represents a study population with a
5 range of exposure durations, including chronic exposures. By drawing the study
6 population from full-time employees and members of the wood-working union,
7 there is an expectation that the study population reflects the demographic of that
8 group as a whole. Although specific summary information is not published for
9 this study group (e.g., average length of employment), the lack of this reporting in
10 itself does not seem to justify an UF for subchronic-to-chronic exposure given the
11 overall study design. As a study adequate for assessing reproductive effects in a
12 chronically exposed cohort, no further adjustment was considered needed.
13 Human variability UF = 10: The study population included women employed in the
14 wood-working industry who were healthy enough to be gainfully employed.
15 Additionally, study inclusion criteria ensured that all study participants had at
16 least one pregnancy resulting in a live birth during the study period (1985-1995).
17 Therefore, these women were reproductively successful. The authors judged that
18 selective participation did not influence potential confounders such as irregular
19 menstruation or earlier miscarriages, which could impact the time to pregnancy
20 results. Susceptible populations were not addressed and, in fact, the women in the
21 study may be considered healthier than the general population in terms of
22 reproductive health. Therefore, an uncertainty factor of 10 for human variability
23 was applied.
24
NOAEL 86 ppb 0 ^ ,
25 RfC = 7 T = T ——r = 8.6 ppb (5-2)
(UFAxUFLxUFsxUFH) (ixlxlxio)
26
27 UFA = 1 (interspecies UF)
28 UFL= 1 (LOAEL-to-NOAEL UF)
29 UFS = 1 (subchronic to chronic UF)
30 UFH = 10 (human variability UF)
31
This document is a draft for review purposes only and does not constitute Agency policy.
5-65 DRAFT—DO NOT CITE OR QUOTE
-------
1 5.1.2.3. Evaluation of the Study-Specific Candidate RfCs
2 Seven studies were selected as key studies for consideration in RfC derivation (see
3 Section 5.1.2, Table 5-4). Candidate RfCs from these studies address various health effects
4 including: sensory irritation, respiratory effects, asthma, increased allergic sensitization, and
5 decreased fecundity (see Table 5-6).
6 Three of the seven studies address sensory irritation of the eye, nose, and throat (Liu
7 et al., 1991; Ritchie and Lehnen, 1987; Hanrahan et al., 1984). The PODs for sensory irritation
8 range from 50 to 95 ppb for a health effect that is considered minimally adverse.
9 Two alternatives are presented for the human variability uncertainty factor in RfC derivation
10 based on these SI studies. Alternative A (UFH = 3) results in cRfCs from 9.5 to 23 ppb.
11 Alternative B (UFH = 1) results in cRfCs from 32 to 70 ppb.
12 A cRfC of 9 ppb is derived for decreased FDR in an occupational study of women in the
13 wood-working industry (Taskinen et al., 1999). This endpoint is supported by four other
14 epidemiologic studies and is considered a potential health concern for occupationally exposed
15 women (John et al., 1994; Taskinen et al., 1994; Seitz and Baron, 1990; Axelsson et al., 1984).
16 However, there is some uncertainty regarding the influence of peak exposures in the work place
17 on the apparent exposure-response relationship based on average workday exposures calculated
18 for study participants. It is unknown if the observed decreased FDR can be attributed to the
19 average exposures from which the cRfC is derived or if it is a result of the measured exposures
20 (as high as 1,000 ppb). If this were the case the cRfC of 9 ppb, based on the average time-
21 weighted exposures, would be protective for decreased fecundity.
22 Three studies identify adverse health effects in residential populations including children:
23 increased incidence of asthma, decreased pulmonary function, increase in respiratory symptoms,
24 and increased allergic sensitization (Rumchev et al., 2002; Garrett et al., 1999 a,b; Krzyzanowski
25 et al., 1999). Asthma, allergic sensitization, pulmonary function, and symptoms of respiratory
26 disease are not only clinically related, but etiologically related, and it is reasonable that they are
27 considered together from a public health perspective. These health effects are observed below
28 the exposure levels that result in sensory irritation and the resulting cRfCs are correspondingly
29 lower, in a range between 2.8 and 11 ppb, depending on the study, endpoint considered, and the
30 application of alternative uncertainty factors for human variability (see Table 5-6).
These three studies of related health effects: asthma, allergic sensitization, pulmonary
function, and symptoms of respiratory disease in children from in-home exposure to
formaldehyde (Rumchev et al., 2002; Garrett et al., 1999 a,b; Krzyzanowski et al., 1999) were
chosen as the basis for the derivation of the RfC. These cocritical studies are mutually
This document is a draft for review purposes only and does not constitute Agency policy.
5-66 DRAFT—DO NOT CITE OR QUOTE
-------
a 5
o S.
*•* TO
rs i
§
KO
S3
Table 5-6. Summary of reference concentration (RfC) derivation from critical study and supporting studies
Endpoint
Study
Study
size
Homes
Children
POD (ppb)
Application of
study-specific UF
UFL
UFS
UFH
cRfC (ppb)
Respiratory effects/asthma and sensitization
Reduction of PEFR in
children (10%)
Asthma incidence
Increased asthma;
allergic sensitization
Krzyzanowski
etal. (1990)
Rumchev et al.
(2002)
Garrett et al.
(1999 a,b)
208
192
148
Yes
Yes
Yes
Yes
Yes
Yes
BMCL10 = 17
NOAEL = 33
LOAEL = 28
1
1
3
1
3
1
3
5.6
Alternative A
3
3.3
Alternative B
1
11
Alternative A
3
2.8
Alternative B
1
9.3
Sensory Irritation
Eye irritation,
burning eyes
Ritchie and
Lehnen(1987)
Hanrahan et al.
(1984)
2,007
61
Yes
Yes
Yes
Some
teenagers
NOAEL = 50
BMCL10 = 70
1
1
1
1
Alternative A
3
17
Alternative B
1
50
Alternative A
o
J
23
Alternative B
1
70
fe
H
o
o
-------
a 5
o S.
*•* TO
rs i
KO
S3
Table 5-6. Summary of reference concentration (RfC) derivation from critical study and supporting studies
(continued)
Endpoint
Eye irritation,
burning eyes
(continued)
Study
Liu etal. (1991)
Study
size
1,394
Homes
Yes
Children
Yes
POD (ppb)
LOAEL = 95
Application of
study-specific UF
UFL
3
UFS
1
UFH
cRfC (ppb)
Alternative A
3
9.5
Alternative B
I
32
Reproductive/Developmental
Decreased
fecundability density
ratio (FDR)
Taskinenetal.,
1999
602
No
No
NOAEL= 86
1
1
10
8.6
§
oo
fe
H
Notes: 1: The final RfC will be rounded to one significant digit per EPA policy. Since the Candidate RfC is an interim calculation, two-significant digits are retained as common
practice in mathematics {i.e., one significant digit more that the final result, to avoid rounding errors compounding across multiple mathematical manipulations}.
o
o
-------
1 supportive and provide similar cRfCs. Therefore, the RfC is taken as the mean of the cRfCs of
2 the cRfCs of the three cocritical studies. For two of these studies (Rumchev et al., 2002; Garrett
3 et al., 1999 a,b), EPA is providing alternatives for the application of the UF addressing human
4 variability. These alternatives result in a threefold difference in cRfCs for each study when
5 considering the critical effects of childhood asthma and allergic sensitization (see Table 5-6).
6 Alternative A, described above for each study, acknowledges that evaluation of these effects in
7 children does address some aspects of human variability, but there remains the potential for
8 additional interindividual variability within the studied population, thus a UF of 3 is warranted.
9 Alternative B, described above for each study, also acknowledges that these studies address
10 human variability and susceptible populations. However in alternative B it is judged that since
11 children are a sensitive lifestage for these effects (asthma and atopy), and are likely the most
12 sensitive population, an UF of 1 may be applied. It is acknowledged that some degree of
13 interindividual variability may remain.
14
Alternative A: Application of a UF of 3 for human variability
Co-critical studies: Rumchev et al. (2002); Krzyzanowski et al. (1999); Garrett et al.
(1999)
Critical endpoints: Asthma, allergic sensitization, pulmonary function, and symptoms of
respiratory disease in children.
Candidate RfCs:
cRfC = 5.6 ppb—decreased PEFR (Krzyzanowski et al., 1999)
cRfC = 3.3 ppb—increased physician-diagnosed asthma (Rumchev et al., 2002)
cRfC = 2.8 ppb—increased asthma, atopy and respiratory symptoms (Garrett et al., 1999)
_-„ 5.6 ppb + 3.3 ppb + 2.8 ppb \\.lppb . ,
RfC: RfC = — — ^— = ^— = 4 ppb
15
16
This document is a draft for review purposes only and does not constitute Agency policy.
5-69 DRAFT—DO NOT CITE OR QUOTE
-------
Alternative B: Application of a UF of 1 for human variability
n = 3 remains for Krzyzanowski et al., 1999)
Co-critical studies: Rumchev et al. (2002); Krzyzanowski et al. (1999); Garrett et al.
(1999)
Critical endpoints: Asthma, allergic sensitization, pulmonary function, and symptoms of
respiratory disease in children.
Candidate RfCs:
cRfC = 5.6 ppb— decreased PEFR (Krzyzanowski et al., 1999)
cRfC =11 ppb — increased physician diagnosed asthma (Rumchev et al., 2002)
cRfC = 9.3 ppb — increased asthma, atopy and respiratory symptoms (Garrett et al., 1999)
. . = .
1
2
3 5.1.3. Database Uncertainties in the RfC Derivation
4 The database of available laboratory animal studies, human clinical and epidemiological
5 studies, and supporting mechanistic information for formaldehyde is substantial. Many of the
6 health effects are well studied in animals and humans, especially those endpoints related to
7 sensory irritation and respiratory effects at the POE, such as respiratory tract pathology, asthma
8 and reduced pulmonary function. This is reflected in the number and high quality of human
9 studies presented in Table 5-4 and supporting data summarized in Chapter 4.
10 The data also indicate effects in other health effect categories, specifically neurotoxic
11 effects, reproductive toxicity, and developmental effects (see Section 5.1.2). These are areas
12 where additional research are needed to reduce uncertainty and better characterize the potential
13 for health effects and the concentrations at which they might occur in humans.
14 The existing toxicological study database strongly supports the potential for
15 formaldehyde to cause both reproductive and developmental toxicity (see Chapter 4; Tables 4-69
16 and 4-72). There is, however, no assessment of these endpoints from a satisfactory
17 two-generation toxicity study to fully evaluate the effect of formaldehyde exposure on
18 reproductive and developmental endpoints. Data are adequate to derive a cRfC of 9 ppb for
19 decreased fecundability density ratio (FDR) from a human occupational study (Taskinen et al.,
20 1999). This study also reports an increase in spontaneous abortions, although there is
21 uncertainty on the exposure levels of concern for this endpoint; spontaneous abortions may also
22 contribute to the decreased FDR on which one of the cRfCs is based. The greatest uncertainty in
This document is a draft for review purposes only and does not constitute Agency policy.
5-70 DRAFT—DO NOT CITE OR QUOTE
-------
1 the cRfC for decreased FDR is the use of a time-weighted exposure metric which does not
2 address possible contributions of peak exposure levels to the observed health effect. As such, it
3 is possible that this cRfC is lower than is needed for protection against decreased FDR. The
4 cRfC for decreased FDR does suggest that the RfC derived from the better studied respiratory
5 effects would be protective of that reproductive/developmental endpoint, but there remain
6 uncertainties as to the full range of potential reproductive and developmental effects. No data
7 exist to sufficiently inform the exposure-response relationship for other reproductive and
8 developmental endpoints as they relate to RfC derivation (see Section 5.1.2.6). For example,
9 male reproductive effects and structural and behavioral developmental effects (including
10 postnatal development) are not addressed by a study of decreased FDR. This is a database
11 deficiency. A survey of the currently available data indicates observed effect levels of
12 5,000-10,000 ppb for male reproductive endpoints and 400 ppb and above for growth
13 retardation and structural anomalies in animal studies. However, these studies employed only
14 one treatment level, precluding the ability to establish a dose-response relationship, thus limiting
15 the strength of the studies for use in RfC derivation.
16 Similarly, there is evidence that formaldehyde can cause neurotoxic effects. There is a
17 deficit of studies with appropriate exposure scenarios to support derivation of an RfC reflecting
18 the potential for observed neurotoxicity due to formaldehyde exposure. None of the available
19 human studies that evaluated neurological effects were adequate for use in quantitative risk
20 assessment, although they did identify neurological effects of concern, including changes in
21 memory and concentration (e.g., Bach et al. [1990]; Kilburn et al. [1987, 1985]) and increased
22 risk of mortality from amyotrophic lateral sclerosis (ALS) with increasing duration of exposure
23 to formaldehyde (Weisskopf et al., 2009). The human and animal data indicate the potential for
24 serious neurological and behavioral effects from short-term formaldehyde exposure (see
25 Section 5.1.2.6). Limited studies in humans, as well as controlled studies in established animal
26 models, confirm the neurotoxic effects of formaldehyde at exposure levels of 100-170 ppb
27 (Malek et al., 2003a, c; Bach et al., 1990) (see Table 5-1). For example, an adverse effect level
28 of 100 ppb for impaired learning is reported for short-term exposures (2 hours/day for 10 days)
29 in rats (Malek et al., 2003a). For this effect, appropriate duration adjustment for extrapolation of
30 a 2-hour repeated exposure over a limited number of days is uncertain. Given the nature of these
31 health effects, and the potential for children to be exposed in the home to levels as high as
32 100 ppb (the level at which effects were seen in animals following a single exposure), this is a
33 significant data gap. Studies are inadequate to determine whether exposure to levels of
34 formaldehyde at or below those that impact children's respiratory health and sensitization will
35 cause neurotoxicity in humans, including endpoints such as impaired learning and memory.
This document is a draft for review purposes only and does not constitute Agency policy.
5-71 DRAFT—DO NOT CITE OR QUOTE
-------
Approaches to the application of a database uncertainty factor:
Options EPA is considering include:
(1) Provide an RfC derived from studies of respiratory and allergenic responses and protective
of sensory irritation effects with a database uncertainty factor of one given significant data on
formaldehyde, but noting that further research reproductive, developmental and neurotoxic
effects would be valuable.
(2) Provide an RfC with a database uncertainty factor of one, with this RfC explicitly
identified as being protective of the well-studied effects.
(3) Apply a database UF of 3 to the RfC derived from studies of respiratory and allergenic
responses to reflect the potential that reproductive, developmental, or neurotoxic effects might
occur at lower doses:
(3) Provide both an RfC identified as protective of the better-studied effects and an RfC with a
database uncertainty factor of 3 incorporated to account for limits to the data on reproductive,
developmental and neurotoxic effects.
2
3
4 It is unclear what uncertainty factors are appropriate to account for human variability and
5 deficiencies in the overall database. For this reason, several alternatives have been presented.
6
7 5.1.4. Uncertainties in the RfC Derivation
8 By design, the RfC is an estimate of an exposure level at which it is unlikely there would
9 be deleterious effects to the human population (including sensitive subgroups) during a lifetime
10 of exposure. Although the RfC is derived from the best available studies, there are a number of
11 uncertainties that underlie the RfC. Some of these uncertainties are addressed quantitatively by
12 applying UFs on a study-specific a basis for RfCs based on animal studies, less-than-chronic
13 exposures, use of a LOAEL as the POD, and to address human variability for the relevant
14 endpoint (see Section 5.1.3). This section elaborates on some of the sources of uncertainty in the
15 final RfC.
16 As the RfC is derived from human studies, the majority in a residential setting, study
17 aspects that are often a great source of uncertainty are of no concern (e.g., use of animal studies,
18 study of a worker population). The uncertainties discussed below apply specifically to the
19 database of formaldehyde studies and the process to derive the RfC.
20
This document is a draft for review purposes only and does not constitute Agency policy.
5-72 DRAFT—DO NOT CITE OR QUOTE
-------
1 5.1.4.1. Point of Departure
2 Most of the studies considered for RfC derivation did not provide enough data to support
3 BMD modeling, which is generally the preferred approach for obtaining a POD for a given
4 dataset (the preference for a POD based on BMD modeling does not, as a general rule, apply to
5 comparisons across datasets). Rather, the PODs for most studies were LOAELs or NOAELs,
6 which have a number of shortcomings relative to a POD obtained from BMD modeling (i.e., a
7 BMCL or BMDL):
8
9 • LOAELs and NOAELs are a reflection of the particular exposure/dose levels used in a
10 study, contributing some inaccuracy to the POD determination.
11 • LOAELs and NOAELs are often determined based on statistical significance and, thus,
12 reflect the number of study subjects or test animals. Studies are typically dissimilar in
13 detection ability and statistical power, with smaller studies tending to identify higher
14 exposure levels as NOAELs compared with larger but otherwise similarly designed
15 studies.
16 • Different LOAELs and NOAELs represent different response rates, so direct qualitative
17 and quantitative comparisons are not possible.
18
19 PODs identified from BMD models overcome some of the deficiencies associated with
20 LOAELs and NOAELs. Benchmark models were used for two inhalation data sets, Hanrahan
21 et al. (1984) and Krzyzanowski et al. (1990).
22 It should also be noted, however, that even for BMCLs/BMDLs there is often
23 uncertainty, in particular for continuous responses, about what response level to select as the
24 BMR, i.e., where to define the cut-off point between a level of change that is not adverse and one
25 that is adverse. In addition, BMD models currently in use are purely mathematical models and
26 are not intended to accurately reflect the biology of the effect being modeled.
27 Another source of uncertainty in the POD is the adjustment for continuous exposure.
28 RfCs are meant to apply to continuous (24 hour/day) exposures. Exposure patterns in human
29 and laboratory animal inhalation studies are typically not continuous and assumptions must be
30 made in converting reported exposure levels to equivalent continuous exposures. Similarly,
31 there are uncertainties about potential dose rate effects, in particular the effect of peak exposures
32 in occupational studies.
33
34 5.1.4.2. Extrapolation from Laboratory Animal Data to Humans
35 Because the inhalation database for formaldehyde contains many human studies for a
36 variety of health effects, it was not necessary to rely on animal data for the endpoints from which
This document is a draft for review purposes only and does not constitute Agency policy.
5.73 DRAFT—DO NOT CITE OR QUOTE
-------
1 to derive the RfC. Thus, unlike for most RfCs, this is not a source of uncertainty in the RfC for
2 formaldehyde.
3
4 5.1.4.3. Human Variation
5 Heterogeneity among humans is another uncertainty associated with extending results
6 observed in a limited human study population or laboratory animal experiment to a larger, more
7 diverse human population.
8 For three of the studies used to derive the RfC, a value of 3 was used for the human
9 variability UF (rather than the default value of 10) because the studies had an apparent over-
10 representation of populations expected to have increased susceptibility (see Section 5.5.3.1):
11
12 • The residential study by Ritchie and Lehnen (1987) evaluated eye, nose, and throat
13 irritation in a large number of subjects, including children and the elderly. As a result of
14 the study's participation criteria, individuals with greater sensitivity were potentially
15 over-repre sented.
16 • Thirty percent of the subjects in the residential study by Krzyzanowski et al. (1990) were
17 children who are more sensitive to formaldehyde-associated decreases in PEFR than
18 adults. The cRfC determination for this study focused on the results in the children,
19 among whom asthmatics were over-represented (roughly three times) compared with the
20 national average of 9.4% in 2008 (Bloom et al., 2009).
21 • Garrett et al. (1999 a,b) conducted a cross-sectional survey of allergy and asthma-like
22 symptoms in children with or without a doctor's diagnosis of asthma. The study was
23 designed to include a high proportion of asthmatic children, a sensitive population for the
24 effects being studied.
25
26 EPA notes, however, that, while a human variability UF of 3 rather than 10 was used to
27 account for certain special attributes of these studies/effects, there is still uncertainty about how
28 much of the overall population heterogeneity is actually reflected even in these relatively diverse
29 residential studies.
30
31 5.1.4.4. Subchronic-to-Chronic Extrapolation
32 RfCs are intended to apply to chronic lifetime exposures. If a study is subchronic
33 (typically less than 10% of a lifetime), a UF for subchronic-to-chronic extrapolation is generally
34 applied to the cRfC for that study. For the key human residential and occupational studies used
35 to derive the RfC in this assessment, the average durations of exposure in the households or
36 workplaces under study are unknown. In this assessment, these studies were considered chronic
37 in nature and no subchronic-to-chronic UF was applied. However, there is uncertainty about
This document is a draft for review purposes only and does not constitute Agency policy.
5.74 DRAFT—DO NOT CITE OR QUOTE
-------
1 whether or not the responses observed fully reflected the potential effects of chronic exposure,
2 especially in children, where, for example, impacts on the developing respiratory and immune
3 systems could be predisposing the children to further adverse effects later in life.
4
5 5.1.5. Previous Inhalation Assessment
6 There is no previous EPA RfC assessment for formaldehyde with which to compare and
7 contrast the RfC developed in this assessment.
8
9 5.2. QUANTITATIVE CANCER ASSESSMENT BASED ON THE NATIONAL
10 CANCER INSTITUTE COHORT STUDY
11 For quantitative assessment of cancer risk, it is generally preferable to use good-quality
12 epidemiologic data, when available, over laboratory animal data. The follow-up studies by
13 Hauptmann et al. (2004) and Beane Freeman et al. (2009) of the large National Cancer Institute
14 (NCI) retrospective cohort mortality study of U.S. workers involved in the production or use of
15 formaldehyde, with quantitative exposure estimates for the individual workers, present an
16 opportunity to perform quantitative cancer risk assessments of nasopharyngeal cancer (NPC) and
17 lymphohematopoietic cancers (Hodgkin lymphoma and leukemia) based on human data.
18 Although other upper respiratory tract cancers were also identified as being causally associated
19 with formaldehyde exposure in the weight-of-evidence analysis in Section 4.5, NPC was the only
20 upper respiratory tract cancer with exposure-response data adequate for the derivation of unit
21 risk estimates in the Hauptmann et al. (2004) follow-up study of solid tumors. Similarly, the
22 weight-of evidence analysis in Section 4.5 concluded that there were causal relationships
23 between formaldehyde exposure and all lymphohematopoietic cancers as a group, leukemias as a
24 group and myeloid leukemia (see Section 4.1.2.2.1.4). Overall the epidemiologic evidence was
25 considered supportive of a causal association between formaldehyde exposure and both Hodgkin
26 lymphoma and multiple myeloma (see Section 4.1.2.2.1.4 and Section 4.5). However, from the
27 Beane Freeman et al. (2009) follow-up study of lymphohematopoietic malignancies, only all
28 leukemias combined and Hodgkin lymphoma were judged to have exposure-response data
29 adequate for the derivation of unit risk estimates (see Section 5.2.3.1 below).
30
This document is a draft for review purposes only and does not constitute Agency policy.
5.75 DRAFT—DO NOT CITE OR QUOTE
-------
1 5.2.1. Choice of Epidemiology Study
2 Several follow-up studies of formaldehyde exposure in industrial workers with some
3 exposure-response information have recently become available. These studies are discussed in
4 more detail in chapter 4 and the appendix (Human Health) and are reviewed only briefly here.
5 Hauptmann et al. (2004) and Beane Freeman et al. (2009) presented follow-ups of the NCI study
6 (originally described by Blair et al. [1986]) of workers at 10 U.S. plants producing or using
7 formaldehyde. Marsh et al. (2007, 2002) focused on pharyngeal cancer and, in particular, NPC
8 mortality in sequential follow-up analyses of the Marsh et al. (1996) cohort study, which
9 examined 1 of the 10 plants studied by NCI. Pinkerton et al. (2004) presented a follow-up of the
10 National Institute for Occupational Safety and Health (NIOSH) study of workers exposed to
11 formaldehyde in three U.S. garment plants (originally described by Stayner et al. [1988]).
12 Coggon et al. (2003) presented an extended follow-up of a study of workers in six British
13 factories where formaldehyde was produced or used (originally described by Acheson et al.
14 [1984] and previously followed up by Gardner et al. [1993]). In addition, Hauptmann et al.
15 (2009) recently conducted a case-control study of lymphohematopoietic and brain cancers, with
16 exposure-response analyses, nested in the cohorts of "professional" workers (funeral industry
17 workers, in this case) studied by Hayes et al. (1990) and Walrath and Fraumeni (1983, 1984).
18 The analyses presented here are based on the NPC (Hauptmann et al., 2004) and
19 lymphohematopoietic cancer (Beane Freeman et al., 2009) results from the NCI follow-up
20 studies. The NCI cohort study is the largest of the three independent industrial worker studies
21 and is the only one with sufficient individual exposure data for exposure-response modeling. In
22 addition, the NCI study is the only one of the three studies that used internal comparisons rather
23 than standardized mortality ratios (SMRs), thus minimizing the impact of the healthy worker
24 effect, which can attenuate observed effect estimates. The NCI cohort consists of 25,619
25 workers (88% male) employed in any of the 10 plants prior to 1966. A follow-up through 1994
26 presented exposure-response analyses for nine NPC deaths as well as analyses of deaths from
27 other solid cancers based on 865,708 person-years of follow-up (Hauptmann et al., 2004). The
28 most recent follow-up based on 998,106 person-years of observation (through 2004) analyzed
29 319 deaths attributed to lymphohematopoietic malignancy from a total of 13,951 deaths (Beane
30 Freeman et al., 2009). The results for solid cancers from this recent follow-up had not yet been
31 published at the time of this draft assessment. A detailed exposure assessment was conducted
32 for each worker in the NCI cohort, based on exposure estimates for different jobs held and tasks
33 performed (Stewart et al., 1986). Exposure estimates were made using several different
34 metrics—peak exposure, average intensity, cumulative exposure, and duration of exposure.
35 Respirator use and exposures to formaldehyde-containing particulates and other chemicals were
This document is a draft for review purposes only and does not constitute Agency policy.
5-76 DRAFT—DO NOT CITE OR QUOTE
-------
1 also considered. For the NPCs, significant trends were observed for the cumulative and peak
2 exposure metrics (Hauptmann et al., 2004). For the lymphohematopoietic cancers, significant
3 trends were observed primarily for all lymphohematopoietic cancers and for Hodgkin lymphoma
4 with the peak exposure metric (Beane Freeman et al., 2009).
5 The NIOSH follow-up study (Pinkerton et al., 2004) analyzed mortality data (2,206
6 deaths; 59 from lymphatic and hematopoietic cancers) from their cohort of 11,098 workers
7 (82% female). Leukemia and aleukemia were elevated for workers with >10 years of exposure
8 and for workers with >20 years since first exposure. However, since no historical exposure level
9 data were available for this cohort, individual worker exposures could not be estimated and
10 exposure-response modeling was not conducted. The British cohort updated by Coggon et al.
11 (2003) consisted of 14,014 male workers, and the follow-up included 5,185 deaths (83 from
12 lymphohematopoietic cancers). In this cohort, lung cancer mortality was statistically
13 significantly increased, especially in workers in the high-exposure category; however, actual
14 exposure estimates were not available for exposure-response modeling (worker exposures were
15 categorized as nil/background, low, moderate, or high, depending on the job considered to have
16 had the highest exposure). Lymphohematopoietic cancers were not elevated in the British
17 cohort, although, as discussed above, the results were based on external comparisons against
18 national mortality statistics. Neither the NIOSH nor the British study reported increased risks of
19 NPC, although only 1 case (0.96) was expected in the NIOSH cohort (Pinkerton et al., 2003) and
20 only 2.0 cases were expected in the British cohort (Coggon et al., 2003). 95% confidence
21 intervals for the relative risk of NPC from these studies (0.07-3.55 and 0.00-3.00, respectively)
22 were estimated by Bosetti et al. (2008) and are not inconsistent with the NPC findings of
23 Hauptmann et al. (2004).
24 In the Hauptmann et al. (2009) nested case-control study, exposures were estimated for
25 each case and control using multiple exposure metrics. Because of limitations in the exposure
26 assessment, however, this study, while useful for hazard assessment, was not used by EPA to
27 derive quantitative risk estimates. Of primary concern, the worker histories were obtained from
28 surrogate responders (next of kin and coworkers). While this approach can produce good quality
29 results for general metrics such as ever embalming or years of embalming, which yielded
30 statistically significant associations (for ever embalming) and trends (for years in jobs with
31 embalming) for lymphohematopoietic cancer of nonlymphoid origin and, in particular, myeloid
32 leukemia, validity declines for more specific variables such as number and duration of
33 embalmings per calendar time period and frequency of spills per calendar time period. These
34 latter variables are needed in the exposure model used to estimate exposures for metrics such as
35 cumulative exposure. Where information on a particular variable was obtained from multiple
This document is a draft for review purposes only and does not constitute Agency policy.
5.77 DRAFT—DO NOT CITE OR QUOTE
-------
1 respondents, Hauptmann et al. (2009) reported a substantial amount of discordance for variables
2 such as number of any embalmings and number of autopsied embalmings. Furthermore,
3 considerable amounts of data were missing. For example, Hauptmann et al. (2009) reported that
4 all but 16 of 44 cases of lymphohematopoietic cancer of nonlymphoid origin had 30% or more of
5 their work history missing. Moreover, although of lesser concern in and of itself, even where
6 good retrospective data on the model variables may have been available, there was additional
7 uncertainty in the estimates resulting from the exposure model. In a validation phase, using real
8 measurements from independent embalmings, the final exposure model, after modifications to
9 correct for initial overestimation, explained 74% of the variability.
10 In addition to limitations in the exposure assessment used in the Hauptmann et al. (2009)
11 study, there were substantial uncertainties about the quantitative precision of the exposure-
12 response relationship observed for myeloid leukemia, which was the one cancer type examined
13 in this study for which there was consistent evidence of an association with formaldehyde
14 exposure. Of the 34 myeloid leukemia cases in the Hauptmann et al. (2009) study, there was
15 only one unexposed case for all of the exposure metrics. Thus, the relative risk estimates (odds
16 ratios) derived in comparison to the referent group are unstable. To address this problem,
17 Hauptmann et al. (2009) created an alternate referent group comprised of workers with
18 <500 lifetime embalmings. As might be expected, since this alternate referent group is no longer
19 an unexposed referent group, odds ratios for the various levels of the different exposure metrics
20 declined considerably (e.g., for the cumulative exposure metric, odds ratios based on the
21 unexposed referent group were 4-5 times higher than those based on the <500-embalmings
22 referent group), although they remained increased relative to the referent group. Thus, although
23 the results of the Hauptmann et al. (2009) study were supportive of the hazard assessment, the
24 overall uncertainty in the quantitative exposure-response data, particularly in the exposure
25 assessment, from the study was considered prohibitive for the development of quantitative
26 cancer risk estimates.
27
28 5.2.2. Nasopharyngeal Cancer
29 5.2.2.1. Exposure-Response Modeling of the National Cancer Institute Cohort
30 A detailed exposure assessment was conducted for the NCI cohort, and quantitative
31 exposure estimates were generated for each worker (Stewart et al., 1986). Formaldehyde
32 exposure estimates, including 8-hour time-weighted average (TWA) exposures and level and
33 frequency of peak exposures, were derived for each job, work area, and calendar year
34 combination. A peak was defined as a short-duration exposure (typically <15 minutes) above
35 the TWA. Cumulative exposures (in ppm x years) were estimated by multiplying the time a
This document is a draft for review purposes only and does not constitute Agency policy.
5-78 DRAFT—DO NOT CITE OR QUOTE
-------
1 worker spent in a specific job by the TWA exposure for that job and summing over all the jobs
2 held by the worker. Duration was the total time spent in jobs with formaldehyde exposure, and
3 average intensity was the ratio of cumulative exposure to duration. Formaldehyde exposures
4 after 1980 were not taken into account in the follow-up study, but this was considered to have a
5 minimal impact on the results (see Section 5.2.2.4).
6 The results of NCI's internal analyses for NPC, using the peak exposure, average
7 intensity, cumulative exposure, and duration of exposure metrics, are presented in Table 5-7.
8 The relative risks (rate ratios) (RRs) were estimated using log-linear Poisson regression models
9 stratified by calendar year, age, sex, and race and adjusted for pay category
10 (salary/wage/unknown). The NCI investigators used the low-exposure category as the reference
11 category to "minimize the impact of any unmeasured confounding variables since nonexposed
12 workers may differ from exposed workers with respect to socioeconomic characteristics"
13 (Hauptmann et al., 2004). A 15-year lag interval was used in estimating exposures in order to
14 account from a minimal latency period for the development of solid cancers, including NPCs.
15 As can be seen in Table 5-7, peak exposure is the exposure metric that provides the
16 strongest exposure-response relationship with NPC. However, it is not clear how to extrapolate
17 RR estimates based on these peak exposure estimates to meaningful estimates of lifetime extra
18 risk of cancer from environmental exposures, where the risk is usually considered to be from
19 continuous lifetime exposures to low environmental levels. In addition, peak exposure is a more
20 subjective measure than the other metrics, it is not based on actual measurements, and it is a
21 categorical rather than continuous measure. Furthermore, the "true" exposure metric best
22 describing the biologically relevant delivered dose of formaldehyde is unknown. The
23 cumulative exposure metric provides a good fit to the data (p trend = 0.029 for all person-years),
24 and, since this is generally the preferred metric for quantitative risk assessment for
25 environmental exposure to carcinogens, cumulative exposure was chosen as the exposure metric
26 for the risk estimate calculations for NPC in this assessment.
27 The nonexposed person-years were included in the primary cancer risk analyses
28 presented here in order to be more inclusive of all the exposure-response data. Such data are
29 typically included in exposure-response modeling. Furthermore, the data were stratified by pay
30 category, which should alleviate some concerns about the nonexposed workers having different
31 socioeconomic characteristics. Final results for the exposed person-years only are presented for
32 comparison.
33 As described above, Hauptmann et al. (2004) investigated the relationship between
34 formaldehyde exposure and NPC mortality using log-linear Poisson regression models. They
35 also conducted log-linear trend tests using the general model RR = epx, where |3 represents the
This document is a draft for review purposes only and does not constitute Agency policy.
5.79 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
regression coefficient for exposure and X is exposure as a continuous variable. The trend
models were stratified by calendar year, age, sex, and race and adjusted for pay category.
Dr. Hauptmann
Table 5-7. Relative risk estimates for mortality from nasopharyngeal malignancies
(ICD-8 code 147) by level of formaldehyde exposure for different exposure metrics
Relative risk (number of deaths)
Peak exposure (ppm)
0
1.00d(2)
>0 to <2.0a
-(0)
2.0 to <4.0
-(0)
>4.0
1.83 (7)
Average intensity (ppm)
0
1.00d(2)
>0 to <0.5
-(0)
0.5 to <1.0
0.38(1)
>1.0
1.67(6)
Cumulative exposure (ppm x years)
0
2.40 (2)
>0 to <1.5
1.00(3)
1.5 to <5.5
1.19(1)
>5.5
4.14(3)
Duration of exposure (years)
0
1.77(2)
>0 to <5
1.00(4)
5 to <15
0.83 (1)
>15
4.18(2)
p trendb
0.044
0.126
0.029
0.206
p trend0
<0.001
0.066
0.025
0.147
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
""Reference category for all categories.
bLikelihood ratio test (1 degree of freedom) of zero slope for formaldehyde exposure (continuous variable, except for
peak exposure metric) among all (nonexposed and exposed) person-years.
'Likelihood ratio test (1 degree of freedom) of zero slope for formaldehyde exposure (continuous variable, except for
peak exposure metric) among exposed person-years only.
dReference category due to no cases in the low-exposure category.
Source: Hauptmann et al. (2004).
provided EPA with the |3 estimates (and their standard errors) from the trend tests for NPC and
the cumulative exposure metric for all person-years and for exposed person-years only (personal
communication from Michael Hauptmann, NCI, to Jennifer Jinot, EPA, March 29, 2004). These
estimates are presented in Table 5-8.
This document is a draft for review purposes only and does not constitute Agency policy.
5-80 DRAFT—DO NOT CITE OR QUOTE
-------
1 Table 5-8. Regression coefficients from NCI log-linear trend test models for NPC
2 mortality from cumulative exposure to formaldehyde3
3
Person-years
All
Exposed only
P
(per ppm x year)
0.05183
0.05318
Standard error
(per ppm x year)
0.01915
0.01914
4
5 "Models stratified by calendar year, age, sex, and race and adjusted for pay category; cumulative exposures
6 calculated using a 15-year lag interval.
7
8 Source: Personal communication from Michael Hauptmann to Jenifer Jinot (March 29, 2004).
9
10
11 5.2.2.2.Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer Mortality
12 The regression coefficients presented in Table 5-8 were used to predict the extra risk of
13 NPC mortality from environmental exposure to formaldehyde.
14
15 Extra ri sk = (Rx-Ro)/( 1 -Ro),
16
17 where Rx is the lifetime risk in the exposed population and Ro is the lifetime risk in an
18 unexposed population (i.e., the background risk). Extra risk estimates were calculated using the
19 P regression coefficients and a life-table program that accounts for competing causes of death.9
20 U.S. age-specific 1999 all-cause mortality rates for all race and gender groups combined
21 (National Center for Health Statistics [NCHS], 2002) were used to specify the all-cause
22 background mortality rates in the life-table program. NCHS 1996-2000 age-specific
23 background mortality rates for NPC were provided by Dr. Eisner of NCI's Surveillance,
24 Epidemiology and End Results (SEER) program (personal communication from Milton Eisner,
25 SEER, to Jennifer Jinot, EPA, December 19, 2003). Risks were computed up to age 85 because
26 cause-specific mortality (and incidence) rates for ages above 85 years are less reliable.
27 Conversions between occupational formaldehyde exposures and continuous environmental
28 exposures were made to account for differences in the number of days exposed per year (240
29 versus 365) and in the amount of air inhaled per day (10 versus 20 m3). An adjustment was also
9
This program is an adaptation of the approach that was previously used in BEIRIV, "Health Risks of Radon and
Other Internally Deposited Alpha Emitters." National Academy Press, Washington, DC, 1988, pp. 131-134. The
same methodology was also used more recently in EPA's 1,3-butadiene health risk assessment (U.S. EPA, 2002). A
spreadsheet illustrating the life table used for the extra risk calculation for the derivation of the LECooos for NPC
incidence (see Section 5.2.2.3) is presented in Appendix C.
This document is a draft for review purposes only and does not constitute Agency policy.
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
made for the 15-year lag period. The reported standard errors for the regression coefficients
were used to compute the one-sided 95% upper confidence limits (UCLs) for the extra risks
based on a normal approximation.
Point estimates and one-sided 95% UCLs for the extra risk of NPC mortality associated
with varying levels of continuous exposure to formaldehyde are presented in Table 5-9. The
model predicts extra risk estimates that are fairly linear for exposures below about 0.001 to
0.01 ppm but not for exposures above 0.01 ppm.
Table 5-9. Extra risk estimates for NPC mortality from various levels of continuous
exposure to formaldehyde
Exposure concentration
(ppm)
0.0001
0.001
0.01
0.1
1
10
Extra risk
1.69x 10~7
1.69x 10~6
1.76x 10~5
2.63 x 10~4
6.22 x 10"1
9.82 x 10"1
95% UCL on extra
risk
2.71 x 10~7
2.73 x 10~6
2.90 x 10~5
5.75 x 10~4
9.00 x 10"1
9.85 x 10"1
Consistent with EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
the same data and methodology were also used to estimate the exposure level (effective
concentration [ECX]) and the associated (one-sided) 95% lower confidence limit (LECX)
corresponding to an extra risk of 0.05% (x = 0.0005). Although EPA guidelines emphasize the
use of exposure levels associated with a 10% extra risk level for the POD for low-dose
extrapolation, that would not be appropriate in this instance. A 10% extra risk level is very high
for responses generally observed in epidemiology studies; thus, a 1% extra risk level is typically
used for epidemiologic data to avoid upward extrapolation. For NPC, however, even the
1% level of risk is associated with RR estimates that are substantially higher than those observed
in the epidemiology study. Hence, even a 1% extra risk level would be an upward extrapolation.
Based on the life-table program, the RR estimate for an extra risk of 1% for NPC mortality is
46. Even 0.1% yields an RR estimate on the high end of the observable range of the
epidemiology study (RR = 5.5). A 0.05% extra risk level yields an RR estimate of 3.27, which
better corresponds to the RRs in the range of the data. Thus, 0.05% extra risk was selected for
determination of the POD, and, consistent with EPA's Guidelines for Carcinogen Risk
This document is a draft for review purposes only and does not constitute Agency policy.
5-82 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Assessment (U.S. EPA, 2005a), the LEG value corresponding to that risk level was used as the
POD. While this may appear to be an inordinately low response level, it must be recognized that
NPC has a very low background mortality rate (e.g., lifetime background risk is about 0.00022);
therefore, a 1% extra risk (i.e., 0.01) would be a huge increase relative to the background risk.
This is consistent with the fact that, even with a large cohort followed for a long time, only
nine NPC deaths were observed in the NCI follow-up through 1994.10
Because formaldehyde is a mutagenic carcinogen and the weight of evidence supports the
conclusion that formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic
MOA (see Section 4.5), a linear low-dose extrapolation was performed in accordance with
EPA's carcinogen risk assessment guidelines (U.S. EPA, 2005a). The ECooos, LECooos, and
inhalation unit risk estimates for NPC mortality are presented in Table 5-10.
Table 5-10. ECooos? LECooos? and inhalation unit risk estimates for NPC mortality
from formaldehyde exposure based on the Hauptmann et al. (2004) log-linear trend
analyses for cumulative exposure
Person-years
All
Exposed only
ECooos
(ppm)
0.15
0.15
LECooos
(ppm)
0.093
0.091
Unit risk3
(Ppm"1)
5.4 x 10~3
5.5 x 10~3
17
18
19
20
21
22
23
24
25
26
27
28
29
30
aUnit risk = 0.0005/LEC00o5.
5.2.2.3.Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer Incidence
EPA cancer risk estimates are typically derived to represent a plausible upper bound on
increased risk of cancer incidence, as from experimental animal incidence data. Cancer data
from epidemiology studies are more often mortality data, as is the case in the NCI study. For
cancers with low survival rates, mortality-based estimates are reasonable approximations of
cancer incidence risk. However, for NPC, the survival rate is substantial (51% at 5 years in the
1990s in the United States, according to Lee and Ko [2005]), and incidence-based risks are
preferred because EPA is concerned with cancer occurrence, not just cancer mortality.
Therefore, an additional calculation was done using the same regression coefficients
provided by Dr. Hauptmann (see Table 5-8) but with age-specific NPC incidence rates for
10 Ten NPCs were reported on death certificates and included in NCI's SMR analysis, but one of these cases was
apparently misclassified on the death certificate, so only nine cases were used to estimate the RRs in the internal
comparison analysis, as discussed by Hauptmann et al. (2004).
This document is a draft for review purposes only and does not constitute Agency policy.
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
1996-2000 from SEER in place of the NPC mortality rates in the life-table program. SEER
collects cancer incidence data from a variety of geographical areas in the United States. The
incidence data used here are from SEER 12, a registry covering about 14% of the U.S.
population, which was the most current SEER registry at the time this analysis was done. SEER
1996-2000 age-specific background incidence rates for NPC were provided by Dr. Eisner of
NCI's SEER program (personal communication from Milton Eisner, SEER, to Jennifer Jinot,
EPA, December 18, 2003). The incidence-based calculation relies on the reasonable
assumptions that NPC incidence and mortality have the same exposure-response relationship for
formaldehyde exposure and that the incidence data are for first occurrences of NPC or that
relapses provide a negligible contribution. The calculation also relies on the fact that NPC
incidence rates are small compared with the all-cause mortality rates.
The resulting ECooos, LECooos, and inhalation unit risk estimates for NPC incidence are
presented in Table 5-11. The unit risk estimate for cancer incidence is twofold higher than the
corresponding mortality-based estimate, for all person-years. This sizeable discrepancy can be
attributed to the high survival rates for NPC.
Table 5-11. ECooos, LECooos, and inhalation unit risk estimates for NPC incidence
from formaldehyde exposure based on the Hauptmann et al. (2004) trend analyses
for cumulative exposure
Person-years
All
Exposed only
ECooos
(ppm)
0.074
0.072
LECooos
(ppm)
0.046
0.045
Unit risk"
(Ppm"1)
1.1 x 10~2
1.1 x 10~2
21
22
23
24
25
26
27
28
29
30
31
32
33
"Unit risk = 0.0005/LEC00o5.
The preferred estimate for the inhalation cancer unit risk for NPC is the estimate of
1.1 x 10~2 per ppm derived using incidence rates for the cause-specific background rates, for all
person-years. The results from the exposed person-years are essentially identical.
Because NPC is a rare cancer, with a relatively low number of cases occurring per year in
the United States, a rough calculation was done to assure that the unit risk estimate derived for
NPC incidence is not implausible in comparison to actual case numbers. For example, assuming
an average constant lifetime formaldehyde exposure level of 5 ppb for the U.S. population, the
inhalation unit risk estimate for NPC equates to a lifetime extra risk estimate of 5.5 x 10~5.
Assuming an average lifetime of 75 years (this is not EPA's default average lifetime of 70 years
This document is a draft for review purposes only and does not constitute Agency policy.
5-84 DRAFT—DO NOT CITE OR QUOTE
-------
1 but rather a value more representative of actual demographic data) and a U.S. population of
2 300,000,000, this lifetime extra risk estimate suggests a crude upper-bound estimate of
3 220 incident cases of NPC attributable to formaldehyde exposure per year. Alternatively,
4 assuming an average constant lifetime formaldehyde exposure level of 20 ppb, the calculation
5 suggests a crude upper-bound estimate of 880 incident cases of NPC per year. Both upper bound
6 estimates, using different assumed lifetime exposure levels, are well below the estimated
7 2,100 total incident NPC cases per year calculated from a published NPC incidence rate for the
8 United States of 0.7/100,000 person-years (Lee and Ko, 2005).u
9
10 5.2.,2.4.Sources of'Uncertainty
11 The two major sources of uncertainty in quantitative cancer risk estimates are generally
12 interspecies extrapolation and high-to-low dose extrapolation. The risk estimates derived from
13 the Hauptmann et al. (2004) analyses of the NCI cohort are not subject to interspecies
14 uncertainty since they are based on human data. However, substantial uncertainty remains in the
15 extrapolation from occupational exposures to lower environmental exposures. Although the
16 actual exposure-response relationship at low exposure levels is unknown, the linear low-dose
17 extrapolation that was used is warranted by the strong support for formaldehyde carcinogenicity
18 having a mutagenic MOA (see Section 4.5). The linear low-dose extrapolation from the
19 95% lower bound on the exposure level associated with the extra risk level serving as the
20 benchmark response is generally considered to provide a plausible upper bound on the risk at
21 lower exposure levels. Actual low-dose risks may be lower to an unknown extent.
22 Other sources of uncertainty emanate from the epidemiologic study and its analysis
23 (Hauptmann et al., 2004), including the retrospective estimation of formaldehyde exposures in
24 the cohort, the modeling of the epidemiologic exposure-response data, the appropriate exposure
25 metric for exposure-response analysis, and potential confounding or modifying factors.
26 The same team of investigators (Stewart et al., 1986) conducted a detailed retrospective
27 exposure assessment to estimate the individual worker exposures. Formaldehyde exposures
28 were estimated for specific jobs/tasks based on monitoring data, discussions with workers and
29 plant managers, and assessment by industrial hygienists. Individual worker estimates were
30 derived for a variety of exposure metrics based on work histories. This exposure assessment was
31 a major undertaking, involving over 100 person-months. Hauptmann et al. (2004) suggested that
1: With the application of age-dependent adjustment factors (see Section 5.4.4), the lifetime unit risk estimate for
NPC would increase by a factor of 1.66, and the crude upper-bound estimates of the incident cases per year
attributable to formaldehyde exposure would similarly increase by a factor of 1.66. The resulting adjusted estimates
of 365 and 1460 for 5 ppb and 20 ppb exposure levels, respectively, are still well below the estimated total number
of 2100 incident cases per year in the United States.
This document is a draft for review purposes only and does not constitute Agency policy.
5-85 DRAFT—DO NOT CITE OR QUOTE
-------
1 employment of such a detailed exposure assessment would tend to minimize exposure
2 misclassification for average and cumulative exposure and duration of exposure but that peak
3 exposure estimates could be more susceptible to misclassification because they were not based
4 on actual measurements. In addition, the follow-up study did not take into account exposures
5 after 1980. Hauptmann et al. (2003) stated that any underestimation of (total) exposure resulting
6 from the 1980 cutoff "would be small because only 3.7% of all person-years were contributed by
7 workers who were 65 years or younger and in exposed jobs in 1980" and because exposure
8 levels were believed to have been much lower after 1980 than in earlier years.
9 As discussed in Chapter 4 and the appendix (Human Health), Marsh et al. (1996) also
10 estimated individual worker exposures at 1 of the 10 plants (Wallingford, Connecticut) studied
11 by the NCI team, and 5 of the 9 NPC deaths were from that plant. The Marsh et al. (1996)
12 exposure estimates were about 10-fold lower than those derived by the NCI team for the workers
13 at the Wallingford plant. Marsh et al. (2002) hypothesized that "the NCI used data from several
14 facilities to estimate exposures in a single facility." However, the NCI investigators maintained
15 that they estimated exposures for each plant separately. While the exact reasons for such a large
16 discrepancy are unclear, some differences in the assessment procedures which could have
17 resulted in substantial differences in the estimates are apparent. First, according to Marsh et al.
18 (1996), 91.7% of the white male Wallingford plant workers were specified as being exposed to
19 formaldehyde in the NCI study, while only 83.3% were considered to have been exposed in the
20 Marsh et al. (1996) analysis (it should be noted that these two cohorts of the Wallingford plant
21 are not identical). Second, the NCI investigators (Stewart et al., 1987, 1986) did their own
22 exposure monitoring at all the plants, including the Wallingford facility, in order to standardize
23 the data provided by the plants as well as to fill data gaps for certain jobs. There is no indication
24 that Marsh et al. (1996) made any additional measurements themselves. Third, although the
25 Marsh et al. (2002, 1996) papers are not entirely consistent on this point, those investigators
26 apparently assumed that the job-specific exposures at the plant were essentially constant over the
27 history of the plant, whereas the NCI team, based on interviews with plant personnel
28 knowledgeable about equipment and process changes, assumed that past exposures were higher.
29 In any event, despite the discrepancies in the absolute exposure values, the relative
30 exposures for both the Marsh et al. (2002, 1996) and NCI studies, as reflected in the exposure-
31 response relationships, are less subject to misclassification and are considered to be reliable.
32 The Wallingford plant is just 1 of the 10 plants in the NCI study (representing 4,389 of the
33 25,619 workers in the NCI cohort), but if the Marsh et al. (1996) exposure estimates, which are
34 roughly 10-fold lower than the NCI estimates, are closer to the actual exposures for those
35 workers, then the true potency of formaldehyde could be greater than that suggested by the unit
This document is a draft for review purposes only and does not constitute Agency policy.
5-86 DRAFT—DO NOT CITE OR QUOTE
-------
1 risk estimates calculated above based on the NCI data. Furthermore, if the NCI exposure values
2 were significantly overestimated across all 10 plants, then the actual potency could be higher
3 still.
4 With respect to the exposure-response model, the log-linear model used by Hauptmann
5 et al. (2003) for their trend tests (i.e., RR = epx) is a commonly used model for epidemiologic
6 data with exposure as a continuous variable. However, the actual exposure-response relationship
7 is unknown. Moreover, even if the correct exposure-response model were known, there would
8 be substantial uncertainty in estimating the model parameters because there are only nine NPC
9 deaths to model. Furthermore, Beane Freeman et al. (2009) reported that in the follow-up
10 through 2004 it was discovered that 1,006 deaths that occurred during the 1980 to 1994
11 follow-up period had not been included in the analyses of the 1994 follow-up study (Hauptmann
12 et al., 2004, 2003), for reasons that have not been identified. Because NPC is such a rare cancer,
13 it is not expected that many, if any, NPC deaths were among the 1,006 excluded deaths;
14 however, it is unknown how inclusion of the 1,006 deaths would have altered the overall
15 exposure-response relationship and, hence, the regression coefficient. Additionally, a 15-year
16 lag was used for all the NCI solid cancer models. The actual minimum latency is unknown;
17 however, the investigators reported that lag intervals between 2 and 20 years yielded similar
18 results.
19 Another potentially significant source of uncertainty is associated with the exposure
20 metrics. With the log-linear model used for modeling the occupational data, the peak exposure
21 metric gave the strongest exposure-response relationship between formaldehyde exposure and
22 increased risk of NPCs. However, it is unclear how to extrapolate RR estimates based on peak
23 exposure estimates to meaningful estimates of lifetime extra risk of cancer from environmental
24 exposure (i.e., extra risk from lifetime continuous low-level environmental exposures). The
25 cumulative exposure metric also yielded a statistically significant exposure-response relationship
26 and was used for the primary cancer risk calculations in this assessment. The "true" exposure
27 metric best describing the lexicologically relevant dose of formaldehyde for nasopharyngeal
28 carcinogenesis is unknown. If a peak-exposure type of metric is the best representative of the
29 lexicologically relevant dose, this suggests that there are dose-rate effects in the exposure-
30 response relationship for formaldehyde and NPC. If this is the case, the unit risk estimates
31 presented here, which are based on a linear low-dose extrapolation, may overestimate the true
32 risks to an unknown extent.
33 Hauptmann et al. (2004) gave a lot of consideration to potential confounding and
34 modifying factors in their analyses. The important factors of age, race, sex, calendar year, and
35 pay category were taken into account in their Poisson regression and trend analyses.
This document is a draft for review purposes only and does not constitute Agency policy.
5-87 DRAFT—DO NOT CITE OR QUOTE
-------
1 Furthermore, they used the low-exposure person-years, rather than the unexposed person-years,
2 as their referent group in an effort to minimize any potential confounding effects resulting from
3 differences in socioeconomic or other characteristics between exposed and unexposed workers.
4 When the slope estimate (i.e., regression coefficient) for the exposed person-years only was used
5 in the analyses presented here, the unit risk estimate was essentially identical to that calculated
6 from the slope estimate for all person-years (see Tables 5-10 and 5-11).
7 In addition, these investigators evaluated routine respirator use, exposure to
8 formaldehyde-containing particulates, durations of exposure to 11 other chemicals/substances in
9 the plants (antioxidants, asbestos, carbon black, dyes and pigments, hexamethylenetetramine,
10 melamine, phenol, plasticizers, urea, wood dust, and benzene), and duration of employment as a
11 chemist or laboratory technician. Only 133 workers ever routinely used a respirator (Hauptmann
12 et al., 2003). Hauptmann et al. (2004) reported that RR estimates for NPC changed when
13 adjusted for duration of melamine exposure, although trend tests remained significant for
14 cumulative formaldehyde exposure (p = 0.006). The investigators suggested that the association
15 with melamine may be spurious, and the regression coefficients (i.e., P estimates) used in this
16 assessment were not adjusted for melamine. RR estimates reportedly did not change
17 substantially when adjusted for exposure to any of the other 10 chemicals/substances. None of
18 the workers who died of NPC was identified as being exposed to wood dust. On the other hand,
19 each of the seven formaldehyde-exposed workers who died of NPC was also exposed to
20 particulates, and neither of the two workers who died of NPC but were not exposed to
21 formaldehyde was exposed to particulates. However, for those workers exposed to particulates,
22 NPC risk increased with increasing formaldehyde exposure, suggesting a formaldehyde-
23 associated effect. Nonetheless, because of the correspondence between formaldehyde and
24 particulate exposures within the workers who died of NPC, there is uncertainty as to whether or
25 not particulates were acting as a modifying factor. Adjusting for duration of time spent working
26 as a chemist or laboratory technician did not substantially alter the results (Hauptmann et al.,
27 2004).
28 Adjusting for plant may result in overadjustment because plant is highly correlated with
29 exposure. Moreover, Hauptmann et al. (2004) adjusted for important plant-related factors by
30 adjusting for the 11 chemicals/substances. Nonetheless, these investigators conducted analyses
31 adjusted for plant to address potential unmeasured confounders associated with plant, and they
32 reported that the association with NPC remained. As noted above, five of the nine NPC deaths
33 were from the Wallingford plant also studied by Marsh et al. (2006, 2002). Marsh et al. (2007)
34 hypothesized that the excess NPCs in the Wallingford plant could be due to external employment
This document is a draft for review purposes only and does not constitute Agency policy.
5-88 DRAFT—DO NOT CITE OR QUOTE
-------
1 in metal-working industries, but we found no evidence to support this supposition (see Section
2 4.1.1.1).
3 Although smoking data were not available for the cohort, smoking is unlikely to explain
4 the excesses in NPCs because there was no consistent increase for tobacco-related diseases,
5 including lung cancer, across the same exposure metrics. No information was available on
6 Epstein-Barr virus, a major risk factor for NPC, in the cohort.
7 Despite inevitable uncertainties, it is important not to lose sight of the strengths of the
8 NCI study. In addition to the use of internal analyses and the extensive exposure assessment and
9 consideration of potential confounding or modifying variables, the NCI study has a large cohort
10 that has been followed for a long time. The cohort included 25,619 subjects, 75% of whom
11 entered before 1960, contributing a total of 865,708 person-years (730,312 for the exposed
12 workers) to the 1994 follow-up. Duration of follow-up in 1994 ranged up to 58 years, with a
13 median of 35 years. Duration of exposure ranged up to 46 years, with a median of 2 years.
14 Additional uncertainties are not so much inherent in the exposure-response modeling or
15 in the epidemiologic data themselves but rather stem from the process of obtaining more general
16 EPA risk estimates from these specific results. EPA cancer risk estimates typically represent a
17 plausible upper bound on increased risk of cancer incidence in the general population for all
18 tissue sites potentially affected by an agent. For experimental animal studies, this is
19 accomplished by using tumor incidence data and summing across all the tumor sites that
20 demonstrate significantly increased incidences, generally using data from the most sensitive sex
21 and species. However, in estimating comparable risks from the NCI epidemiologic data, certain
22 limitations are encountered. First, the NCI study is a retrospective mortality study, and cancer
23 incidence data are unavailable for the cohort. Second, these occupational epidemiology data
24 represent a worker cohort that is generally healthier than the general population
25 (e.g., SMRs < 1) (see Table 2 of Hauptmann et al. [2004]).
26 The first limitation was addressed quantitatively in the calculation of cancer incidence
27 risk estimates from the mortality results, and, even though there are assumptions made in using
28 incidence data this way, the incidence-based estimates are believed to be better estimates of
29 cancer incidence risk than the mortality-based estimates. With respect to the second limitation,
30 the healthy worker effect is often an issue in occupational epidemiology studies, and it is
31 difficult to know to what extent there is a healthy worker effect with respect to the development
32 of NPC in this study. As discussed above, Hauptmann et al. (2004) sought to minimize potential
33 confounding effects resulting from differences in socioeconomic or other characteristics between
34 exposed and unexposed workers by using the low-exposure person-years, rather than the
35 unexposed person-years, as their referent group. Nonetheless, when the slope estimates for the
This document is a draft for review purposes only and does not constitute Agency policy.
5-89 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposed person-years only were used in the analyses in this assessment, unit risk estimates
2 essentially identical to those calculated from the slope estimates for all person-years were
3 obtained (see Tables 5-10 and 5-11). In terms of representing the general population, the NCI
4 cohort was somewhat diverse, but the workers were predominantly white males (81%) then
5 white females (12%), black males (7%), and black females (<1%), and they were all adults.
6 Finally, NPC is just one of the upper respiratory tract cancers concluded to be causally
7 associated with formaldehyde exposure (see Section 4.5). These upper respiratory tract cancers
8 are rare cancers and are difficult to detect in cohort studies. Thus, although NPC was the only
9 such cancer with an exposure-response relationship amenable to the derivation of a unit risk
10 estimate, additional, unquantified risk may exist for the other upper respiratory tract cancers. If
11 there was a strong exposure-response relationship between these cancers and formaldehyde
12 exposure, a more apparent association in the Hauptmann et al. (2004) study might have been
13 expected, as was seen for NPC, despite the rare nature of these cancers. Thus, the exposure-
14 response relationship for these other upper respiratory tract cancers is likely modest, at best, and,
15 because these are rare cancers, the contribution of the risk for these cancers to the total cancer
16 risk from formaldehyde exposure is not expected to be large. Nonetheless, with such rare
17 cancers, there is uncertainty regarding the extent to which the estimate based on NPC may
18 underestimate the risk for all upper respiratory tract cancers.
19 In summary, the inhalation cancer unit risk estimate of 1.1 x 10~2 per ppm for NPC is
20 based on human data from a high-quality epidemiologic study with individual exposure
21 estimates for each worker. A major uncertainty is the appropriate model/exposure metric for
22 extrapolation to environmental exposures.
23
24 5.2.3. Lymphohematopoietic Cancer
25 5.2.3.l.Exposure-Response Modeling of the National Cancer Institute Cohort
26 The results of NCI's internal analyses for lymphohematopoietic cancers using the peak
27 exposure, average intensity, and cumulative exposure metrics from the follow-up through 2004
28 are reported by Beane Freeman et al. (2009). There was reportedly no evidence of associations
29 with duration of exposure, and those results were not presented. For the peak exposure metric,
30 statistically significant log-linear trends were observed for all lymphohematopoietic cancers,
31 Hodgkin lymphoma, and leukemia (the latter only when the unexposed person-years were
32 included). There was also evidence for potential associations with myeloid leukemia
33 specifically, especially when risks were viewed over time, and with multiple myeloma. Using
34 the average exposure metric, there was a significant trend for Hodgkin lymphoma. With the
35 cumulative exposure metric, there were no statistically significant trends; however, the Hodgkin
This document is a draft for review purposes only and does not constitute Agency policy.
5-90 DRAFT—DO NOT CITE OR QUOTE
-------
1 lymphoma trend results had/?-values not much greater than 0.05 (p trends = 0.06 and 0.08 with
2 and without the unexposed person-years, respectively), as did the leukemia trend results
3 (p trends = 0.08 and 0.12 with and without the unexposed person-years, respectively). As
4 discussed above with NPC, it is not clear how to extrapolate RR estimates based on the peak
5 exposure estimates to meaningful estimates of lifetime extra risk of cancer from environmental
6 exposures. The average exposure metric is also problematic because it suggests that duration of
7 exposure is not important (e.g., exposure to a given exposure level for 1 year conveys the same
8 amount of risk as exposure to the same level for 70 years). Cumulative exposure is generally the
9 preferred metric for quantitative risk assessment for environmental exposure to carcinogens, and,
10 because the Hodgkin lymphoma and leukemia trend results had ^-values not much greater than
11 0.05 using the cumulative exposure metric and the elevations in risk with that metric were
12 consistent with significant elevations observed with the peak exposure (for Hodgkin lymphoma
13 and leukemia) and average exposure (for Hodgkin lymphoma) metrics (see Table 5-12), a
14 determination was made to calculate unit risk estimates for Hodgkin lymphoma and leukemia
15 based on cumulative exposure. There is also support for associations between formaldehyde
16 exposure and both Hodgkin lymphoma and leukemia from other studies (see Section 4.5.2). No
17 other lymphohematopoietic cancer responses provided adequate exposure-response data with the
18 cumulative formaldehyde exposure metric in the NCI cohort from which to derive unit risk
19 estimates.
20 As for the NPC results discussed in Section 5.2.2, the RR estimates in Table 5-12 were
21 derived using log-linear Poisson regression models stratified by calendar year, age, sex, and race
22 and adjusted for pay category (salary/wage/unknown). The NCI investigators used the low-
23 exposure category as the reference category to "minimize the impact of any unmeasured
24 confounding variables since nonexposed workers may differ from exposed workers with respect
25 to socioeconomic characteristics" (Hauptmann et al., 2004). A 2-year lag interval was used to
26 determine exposures in order to account for a minimal latency period for lymphohematopoietic
27 cancers.
28 Dr. Beane Freeman provided EPA with the regression coefficient estimates for Hodgkin
29 lymphoma and leukemia mortality from the log-linear trend test models for cumulative exposure
30 (i.e., RR = epx, with exposure [X] as a continuous variable) used in the NCI analyses (personal
31 communication from Laura Beane Freeman, NCI, to John Whalan, EPA, August 26, 2009).
32 These estimates are presented in Table 5-13. As with the NPC calculations in Section 5.2.2, the
33 nonexposed person-years were included in the primary unit risk estimate derivations in order to
34 Table 5-12. Relative risk estimates for mortality from Hodgkin lymphoma (ICD-8
35 code 201) and leukemia (ICD-8 codes 204-207) by level of formaldehyde exposure
This document is a draft for review purposes only and does not constitute Agency policy.
5-91 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
for different exposure metrics
Cancer type
Relative risk (number of deaths)
p trendb
p trend0
Peak exposure (ppm)
Hodgkin
lymphoma
Leukemia
0
0.67 (2)
0.59 (7)
>0 to <2.0a
1.0(6)
1.0(41)
2.0 to <4.0
3.30(8)
0.98 (27)
>4.0
3.96(11)
1.42(48)
0.004
0.02
0.01
0.12
Average intensity (ppm)
Hodgkin
lymphoma
Leukemia
0
0.53 (2)
0.54 (7)
>0 to <0.5
1.0(10)
1.0(67)
0.5 to <1.0
3.62(9)
1.13(25)
>1.0
2.48 (6)
1.10(24)
0.03
0.50
0.05
>0.50
Cumulative exposure (ppm x years)
Hodgkin
lymphoma
Leukemia
0
0.42 (2)
0.53 (7)
>0 to <1.5
1.0(14)
1.0(63)
1.5 to <5.5
1.71 (7)
0.96 (24)
>5.5
1.30(4)
1.11 (29)
0.06
0.08
0.08
0.12
4
5
6
7
8
9
10
11
12
13
14
15
"Reference category for all categories.
bLikelihood ratio test (1 degree of freedom) of zero slope for formaldehyde exposure (continuous variable, except for
peak exposure metric) among all (nonexposed and exposed) person-years.
'Likelihood ratio test (1 degree of freedom) of zero slope for formaldehyde exposure (continuous variable, except for
peak exposure metric) among exposed person-years only.
Source: Beane Freeman et al. (2009).
Table 5-13. Regression coefficients for Hodgkin lymphoma and leukemia
mortality from NCI trend test models3
Cancer type
Hodgkin lymphoma
Leukemia
Person-years
All
Exposed only
All
Exposed only
P
(per ppm x year)
0.02959
0.02879
0.01246
0.01131
Standard error
(per ppm x year)
0.01307
0.01333
0.006421
0.00661
16
This document is a draft for review purposes only and does not constitute Agency policy.
5-92 DRAFT—DO NOT CITE OR QUOTE
-------
1 "Models were stratified by calendar year, age, sex, and race and adjusted for pay category; exposures included a
2 2-year lag interval.
O
4 Source: Personal communication from Laura Beane Freeman to John Whalan (August 26, 2009).
5
6
7 be more inclusive of all the exposure-response data. Final results for the exposed person-years
8 only are presented for comparison.
9
10 5.2.3.2.Prediction of Lifetime Extra Risks for Hodgkin Lymphoma and Leukemia Mortality
11 Extra risk estimates for Hodgkin lymphoma and leukemia mortality were calculated
12 using the same general methodology described above for the NPC mortality estimates (see
13 Section 5.2.2.2), with the following exceptions. U.S. age-specific 2006 all-cause mortality rates
14 (NCHS, 2009) and NCHS age-specific 2002-2006 background mortality rates for Hodgkin
15 lymphoma and leukemia (http://seer.cancer.gov/csr/1975_2006/) for all race and gender groups
16 combined were used in the life-table programs. In addition, a 2-year lag period was used instead
17 of a 15-year lag period.
18 The resulting point estimates and one-sided 95% UCLs for the extra risk of Hodgkin
19 lymphoma mortality associated with varying levels of continuous exposure to formaldehyde are
20 presented in Table 5-14. The results for leukemia are shown in Table 5-15. In both cases, the
21 models predict extra risk estimates that are fairly linear for exposures below about 0.01-0.1 ppm
22 but not for exposures above 0.1 ppm.
23 As discussed in Section 5.2.2.2 above, 1% extra risk levels are typically used as the basis
24 for the POD for low-dose extrapolation from epidemiologic data. As for NPC, however,
25 Hodgkin lymphoma has a very low background mortality rate (e.g., lifetime background risk is
26 about 0.00038), and the 1% level of risk is associated with RR estimates that are substantially
27 higher than those observed in the epidemiology study. Hence, a 1% extra risk level would be an
This document is a draft for review purposes only and does not constitute Agency policy.
5.93 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 5-14. Extra risk estimates for Hodgkin lymphoma mortality from
various levels of continuous exposure to formaldehyde
Exposure concentration
(ppm)
0.0001
0.001
0.01
0.1
1
10
Extra risk
2.04 x 10~7
2.05 x KT6
2.10x 10~5
2.79 x 10^
1.63 x 10"1
9.89 x 10"1
95% UCL on extra risk
3.53 x 10~7
3.55 x 10~6
3.71 x 10~5
6.17 x 10~4
8.36 x 10"1
9.90 x 10"1
4
5
6
7
Table 5-15. Extra risk estimates for leukemia mortality from various levels
of continuous exposure to formaldehyde
Exposure concentration
(ppm)
0.0001
0.001
0.01
0.1
1
10
Extra risk
1.64x KT6
1.64x 10~5
1.66x 10^
1.87x 10~3
8.07 x 10~2
9.80 x 10"1
95% UCL on extra risk
3.02 x 10~6
3.03 x 10~5
3.10x 10~4
3.90 x 10~3
5.19 x 10"1
9.89 x 10"1
9
10
11
12
13
14
15
16
17
18
19
20
21
22
upward extrapolation. Based on the life-table program, the RR estimate associated with an extra
risk of 1% for Hodgkin lymphoma mortality is 27. Even 0.1% yields an RR estimate at the
higher end of what was observed in the epidemiology study (RR = 3.6) (note that our primary
analyses include the nonexposed workers, and thus the 0-exposure group becomes the referent
group and the RR estimates presented for Hodgkin lymphoma and cumulative exposure in
Table 5-12 would be adjusted upward [about 2.4-fold] relative to the 0-exposure group). A
0.05% extra risk level yields an RR estimate of 2.3, which better corresponds to the RRs at the
lower end of the observable range. Thus, 0.05% extra risk was selected for determination of the
POD for Hodgkin lymphoma, and, consistent with EPA's Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 2005a), the LEG value corresponding to that risk level was used as the
POD.
This document is a draft for review purposes only and does not constitute Agency policy.
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
RR estimate (2.5) that would be above the highest categorical result reported, even after
adjusting the RR estimates upward relative to the 0-exposure group (see above paragraph). A
0.5% extra risk level yields an RR estimate of 1.8, which better corresponds to the RRs in the
range of the data. Thus, the LEG value corresponding to 0.5% extra risk was selected for the
POD for leukemia.
Because formaldehyde is a mutagenic carcinogen and the weight of evidence supports the
conclusion that formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic
MOA (see Section 4.5), a linear low-dose extrapolation was performed, also in accordance with
EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a). The ECooos, LECooos, and
inhalation unit risk estimates for Hodgkin lymphoma mortality are presented in Table 5-16, and
the ECoos, LECoos, and inhalation unit risk estimates for leukemia mortality are presented in
Table 5-17.
Table 5-16. ECooos? LECooos? and inhalation unit risk estimates for Hodgkin
lymphoma mortality from formaldehyde exposure based on Beane Freeman
et al. (2009) log-linear trend analyses for cumulative exposure
Person-years
All
Exposed only
ECooos
(ppm)
0.151
0.155
LECooos
(ppm)
0.0875
0.0881
Unit risk3
(Ppm"1)
5.7 x 10~3
5.7 x 10~3
18
19
20
21
22
aUnit risk = 0.0005/LEC00o5.
This document is a draft for review purposes only and does not constitute Agency policy.
5.95 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
Table 5-17. ECoos, LECoos, and inhalation unit risk estimates for leukemia
mortality from formaldehyde exposure based on Beane Freeman et al. (2009)
log-linear trend analyses for cumulative exposure
Person-years
All
Exposed only
ECoos
(ppm)
0.224
0.246
LECoos
(ppm)
0.121
0.126
Unit risk"
(Ppm"1)
4.1 x 10~2
4.0 x 10~2
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
aUnit risk = 0.005/LEC005.
5.2.3.3.Prediction of Lifetime Extra Risks for Hodgkin Lymphoma and Leukemia Incidence
As for NPC, both Hodgkin lymphoma and leukemia have substantial survival rates
(84.7% at 5 years for Hodgkin lymphoma [http://seer.cancer.gov/statfacts/html/hodg.html] and
53.1% at 5 years for leukemia [http://seer.cancer.gov/statfacts/html/leuks.html], based on
1999-2005 SEER data); thus, it is preferable to derive incidence estimates. Unit risk estimates
for Hodgkin lymphoma and for leukemia incidence were calculated as described above for the
NPC incidence estimates (see Section 5.2.2.3). Age-specific background incidence rates for
2002-2006 for Hodgkin lymphoma and for leukemia from SEER17, a registry covering about
26% of the U.S. population, were obtained from the SEER Web site
(http://seer.cancer.gov/csr/1975_2006/). The incidence-based calculation relies on the
assumptions that Hodgkin lymphoma (and leukemia) incidence and mortality have the same
exposure-response relationship for formaldehyde exposure and that the incidence data are for
first occurrences of Hodgkin lymphoma (and leukemia) or that relapses provide a negligible
contribution. The first assumption is more uncertain for leukemia because it is a grouping of
subtypes with different survival rates (see Section 5.2.3.4 for further discussion). The
calculation also relies on the fact that Hodgkin lymphoma (and leukemia) incidence rates are
small compared with the all-cause mortality rates. The resulting ECooos, LECooos, and inhalation
unit risk estimates for Hodgkin lymphoma incidence are presented in Table 5-18, and the ECoos,
LECoos, and inhalation unit risk estimates for leukemia incidence are presented in Table 5-19.
The unit risk estimate for Hodgkin lymphoma incidence is about threefold higher than the
corresponding mortality-based estimate, for all person-years. This sizeable discrepancy can be
attributed to the high survival rates for Hodgkin lymphoma. For leukemia, the incidence unit
risk estimate is about 40% higher than the mortality-based estimate. This difference is lower
than the twofold
This document is a draft for review purposes only and does not constitute Agency policy.
5-96 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
Table 5-18. ECooos, LECooos, and inhalation unit risk estimates for Hodgkin
lymphoma incidence from formaldehyde exposure, based on Beane Freeman
et al. (2009) log-linear trend analyses for cumulative exposure
Person-years
All
Exposed only
ECooos
(ppm)
0.0515
0.0529
LECooos
(ppm)
0.0298
0.0301
Unit risk3
(Ppm"1)
1.7x 10~2
1.7x 10~2
5
6
7
8
9
10
11
12
aUnit risk = 0.0005/LECooos
Table 5-19. ECoos, LECoos, and inhalation unit risk estimates for leukemia
incidence from formaldehyde exposure based on Beane Freeman et al. (2009)
log-linear trend analyses for cumulative exposure
Person-years
All
Exposed only
ECoos
(ppm)
0.162
0.178
LECoos
(ppm)
0.0875
0.0909
Unit risk"
(Ppm"1)
5.7 x 10~2
5.5 x 10~2
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
"Unit risk = 0.005/LEC005.
difference seen with NPC estimates, despite comparable survival rates, probably because of
different age distributions of the mortality and incidence rates.
The preferred estimate for the inhalation cancer unit risk for Hodgkin lymphoma is the
estimate of 1.7 x 10~2 per ppm derived using incidence rates for the cause-specific background
rates, for all person-years. Similarly, the preferred estimate for leukemia is the estimate of
5.7 x 10~2 per ppm derived using incidence rates, for all person-years. In both cases, the results
from the exposed person-years only are essentially identical.
Because Hodgkin lymphoma is a rare cancer, with a relatively low number of cases
occurring per year in the United States (according to SEER statistics, an estimated 8,510 people
were diagnosed with Hodgkin lymphoma in the United States in 2009
[http://seer.cancer.gov/statfacts/html/hodg.html]), a rough calculation was done to assure that the
unit risk estimate derived for Hodgkin lymphoma incidence is not implausible in comparison to
actual case numbers. For example, assuming an average constant lifetime formaldehyde
exposure level of 5 ppb for the U.S. population, the inhalation unit risk estimate for Hodgkin
lymphoma equates to a lifetime extra risk estimate of 8.5 x 10~5. Assuming an average lifetime
This document is a draft for review purposes only and does not constitute Agency policy.
5.97 DRAFT—DO NOT CITE OR QUOTE
-------
1 of 75 years (this is not EPA's default average lifetime of 70 years but rather a value more
2 representative of actual demographic data) and a U.S. population of 300,000,000, this lifetime
3 extra risk estimate suggests a crude upper-bound estimate of 340 incident cases of Hodgkin
4 lymphoma attributable to formaldehyde exposure per year. Alternatively, assuming an average
5 constant lifetime formaldehyde exposure level of 20 ppb, the calculation suggests a crude upper-
6 bound estimate of 1,360 incident cases of Hodgkin lymphoma per year. Both upper bound
7 estimates, using different assumed lifetime exposure levels, are well below the estimated
8 8,510 total incident Hodgkin lymphoma cases diagnosed per year in the United States.12
9
10 5.2.,3.4.Sources of'Uncertainty
11 By and large, the sources of uncertainty discussed above (see Section 5.2.2.4) for the
12 NPC risk estimates, such as high-to-low dose extrapolation, retrospective exposure estimation,
13 exposure metric/model uncertainties, and application of data from a "healthy" worker cohort to
14 the more diverse general population also apply to the Hodgkin lymphoma and leukemia risk
15 estimates. The Hodgkin lymphoma risk estimates are based on 27 deaths, which is more than
16 were available for the NPC risk estimates, but 27 is still a small number for exposure-response
17 modeling. The leukemia risk estimates are based on 123 deaths, so there is less uncertainty with
18 the parameter estimation from the exposure-response modeling for that cancer type, although
19 uncertainties still exist about the general model form. A 2-year lag interval was used for
20 lymphohematopoietic cancers versus the 15-year lag for NPC. Beane Freeman et al. (2009)
21 evaluated lag intervals between 2 and 25 years and reported that lag intervals of about 18 years
22 provided the best fit to the lymphohematopoietic cancer data but did not change the risk
23 estimates; thus, they retained the 2-year lag interval that was used in the previous follow-up
24 (Hauptmann et al., 2003). The most appropriate lag intervals for Hodgkin lymphoma and
25 leukemia are unknown, but alternate lags are unlikely to have a large impact on the results.
12 With the application of age-dependent adjustment factors (see Section 5.4.4), the lifetime unit risk estimate for
Hodgkin lymphoma would increase by a factor of 1.66, and the crude upper-bound estimates of the incident cases
per year attributable to formaldehyde exposure would similarly increase by a factor of 1.66. The resulting adjusted
estimates of 564 and 2,260 for 5 ppb and 20 ppb exposure levels, respectively, are still well below the estimated total
number of 8,500 incident cases per year in the United States.. Similar calculations for leukemia yield even lower
relative upper-bound estimates of cases attributable to formaldehyde exposure, in comparison to estimated total
incident cases, because, although the unit risk estimate for leukemia is about 3.3 times the unit risk estimate for
Hodgkin lymphoma, the total estimated number of incident leukemia cases in the United States, is 5.3 times the
estimate for Hodgkin lymphoma (an estimated 44,790 cases diagnosed in the U.S. for 2009, according to SEER
[http://seer.cancer.gov/statfacts/html/leuks.html]). For leukemia, crude upper-bound ADAF-adjusted estimates of
the incident cases per year attributable to formaldehyde exposure levels of 5 ppb and 20 ppb are 1900 and 7,580,
respectively, which are well below the estimated total number of 44,790 incident cases per year in the United States.
This document is a draft for review purposes only and does not constitute Agency policy.
5-98 DRAFT—DO NOT CITE OR QUOTE
-------
1 The same potential confounding or modifying factors that were investigated for NPC and
2 the other solid cancers, as discussed in Section 5.2.2.4 above, were evaluated for the
3 lymphohematopoietic cancers. Beane Freeman et al. (2009) reported that controlling for
4 duration of exposure to the 11 other substances that they considered (see Section 5.2.2.4) or for
5 working as a chemist or laboratory technician "did not meaningfully change results"; results
6 were not shown. The investigators also reported that excluding the 586 individuals with
7 exposure to benzene, a known leukemogen, did not change the RR estimates for myeloid or
8 lymphoid leukemia in the highest peak exposure category. Furthermore, Beane Freeman et al.
9 (2009) found no evidence of heterogeneity of RR estimates for lymphohematopoietic cancers by
10 race, sex, or pay category, and adjusting for plant reportedly did not substantively change results.
11 A further uncertainty is which lymphohematopoietic cancer types are linked to
12 formaldehyde exposure. As discussed in Section 4.5.2, lymphohematopoietic cancers are a
13 diverse group of cancers with different etiologies, and the epidemiologic database suggests
14 associations with multiple different subtypes of these cancers. Section 4.5 concludes that
15 formaldehyde is causally associated with all lymphohematopoietic cancers as a group and with
16 leukemias as a group (with the strongest evidence for myeloid leukemia). However, at present,
17 exactly which subtypes are etiologically linked to formaldehyde exposure is unknown. Cancer
18 risk estimates were derived for Hodgkin lymphoma and leukemia because, in addition to support
19 for an association between these lymphohematopoietic cancer subtypes and formaldehyde
20 exposure with other exposure metrics and from other studies, these had the strongest associations
21 with cumulative exposure in the Beane Freeman et al. (2009) update of the large, high-quality
22 NCI study. However, it is unknown whether these two subtypes best represent the total
23 lymphohematopoietic cancer risk.
24 In addition, leukemia itself is a grouping of diverse (e.g., acute lymphocytic, chronic
25 lymphocytic, acute myeloid, chronic myeloid) subtypes, and using this grouping injects
26 additional uncertainty into the derivation of cancer incidence estimates. One of the assumptions
27 that the incidence-based calculation relies on is that the cancer incidence and mortality have the
28 same exposure-response relationship for formaldehyde exposure. This assumption may be
29 problematic for the leukemia incidence estimates if not all of the leukemia subtypes represented
30 in the grouping are associated with formaldehyde exposure to the same extent. This is because
31 different leukemia subtypes have different survival rates, so if a subtype with a relatively high
32 survival rate is included in the background incidence rates while not actually being associated
33 with formaldehyde exposure or being associated to a lesser extent than other subtypes, then the
34 incidence risk will be overestimated. The mortality risk calculations are not similarly affected
This document is a draft for review purposes only and does not constitute Agency policy.
5.99 DRAFT—DO NOT CITE OR QUOTE
-------
1 by including subtypes that may not actually be associated with formaldehyde exposure because
2 background mortality for the subtypes is already taken into account in the regression coefficient.
3 Figure 5-10 shows the mortality versus incidence rates for all leukemia and the two main
4 subtypes, myeloid leukemia and lymphoid leukemia. This figure does not show the acute versus
5 chronic myeloid and leukemia subtypes or the monocytic or other leukemia subtypes; however,
6 it serves to illustrate the impact of using rates for groupings that contain subtypes with different
7 survival rates. For example, if lymphoid leukemia is the predominant subtype associated with
8 formaldehyde exposure, then using the leukemia grouping for the incidence rates may
9 underestimate the cancer incidence risk because the incidence rates for leukemia (relative to the
10 mortality rates) are diluted with inclusion of the incidence rates for myeloid leukemia, which has
11 a smaller incidence-to-mortality ratio (i.e., poorer survival). On the other hand, if myeloid
12 leukemia is the predominant subtype associated with formaldehyde exposure, then using the
13 leukemia grouping for the incidence rates may overestimate cancer incidence risk. If incidence
14 risks are being overestimated, the effect should be minimal because the incidence risk estimates
15 for leukemia calculated in Section 5.2.3.3 are not that much greater (about 40%) than the
16 mortality-only estimates.
17 Finally, as for the NPC risk estimates, when the slope estimates for the exposed person-
18 years only were used for the Hodgkin lymphoma and leukemia risk calculations, unit risk
19 estimates similar to those calculated from the slope estimates for all person-years were obtained
20 (see Tables 5-16 to 5-19); thus, the impacts of including the unexposed person-years are
21 minimal.
22 As discussed in Section 5.2.2.4, despite inevitable uncertainties, it is important not to lose
23 sight of the strengths of the NCI study. In addition to the use of internal analyses and extensive
24 exposure assessment and consideration of potential confounding or modifying variables, the NCI
25 study has a large cohort that has been followed for a long time. With the additional follow-up
26 through 2004, reflected in the lymphohematopoietic cancer results of Beane Freeman et al.
27 (2009), the median duration of follow-up was 42 years, and the 25,619 cohort members had
28 accrued 998,106 person-years of follow-up. Over half of the cohort was deceased, and there was
29 a substantial number of lymphohematopoietic deaths (319 total; 286 in the exposed workers).
30 In summary, the inhalation cancer incidence unit risk estimates of 1.7 x 10~2 per ppm for
31 Hodgkin lymphoma and 5.7 x 10~2 per ppm for leukemia are based on human data from a high-
32 quality epidemiologic study with individual exposure estimates for each worker. The major
33 source of uncertainty in both risk estimates is the extrapolation to environmental exposures.
This document is a draft for review purposes only and does not constitute Agency policy.
5-100 DRAFT—DO NOT CITE OR QUOTE
-------
s. 5^
§
§
fj\
o
80.0
70.0
0.0
— —all leuk mort
all leuk inc
— —myeloid mort
myeloid inc
lymphoid mort
— lymphoid inc
0.0 10.0 20.0 30.0
40.0 50.0
age (years)
60.0 70.0 80.0
Figure 5-W. Age-specific mortality and incidence rates for myeloid, lymphoid, and all leukemia.
O
O
-------
1 5.2.4. Conclusions on Cancer Unit Risk Estimates Based on Human Data
2 In this assessment, a (plausible upper bound) lifetime extra cancer unit risk of
3 5.4 x 1CT3 per ppm of continuous formaldehyde exposure was estimated using a
4 life-table program and linear low-dose extrapolation of the excess NPC mortality and log-linear
5 modeling results (for cumulative exposure) reported in a high-quality occupational
6 epidemiologic study (based on nine NPC deaths). Applying the same regression coefficient and
7 life-table program to background NPC incidence rates yielded a lifetime extra cancer unit risk
8 estimate of 1.1 x 1CT2 per ppm (8.8 x 1CT6 per |ig/m3).
9 Using similar methods and data for Hodgkin lymphoma (27 deaths) and leukemia
10 (123 deaths) mortality based on the cumulative exposure metric, from a further follow-up of the
11 same cohort study, (plausible upper bound) lifetime extra cancer risk estimates of 1.7 x 1CT2 per
12 ppm (1.4 x 1CT5 per |ig/m3) and 5.7 x 1CT2 per ppm (4.6 x 1CT5 per |ig/m3) for Hodgkin
13 lymphoma incidence and leukemia incidence, respectively, were derived.
14 To estimate the total cancer risk from formaldehyde exposure, risk estimates for these
15 three cancer types (NPC, Hodgkin lymphoma, and leukemia) were combined, although, as
16 discussed above, these three cancer types may not fully reflect the total cancer risk for all
17 cancers thought to be causally associated with formaldehyde exposure. For an approximate
18 estimate of the combined (upper bound) risk, risk estimates were combined assuming a normal
19 distribution. For comparability, risk estimates for formaldehyde were combined at a common
20 level of 0.1 ppm. This level was selected because it is close to the PODs (LECooss) used above
21 for leukemia mortality (0.121 ppm) and leukemia incidence (0.0875 ppm), and leukemia is the
22 predominant cancer type in terms of extra risk. Note that unit risk estimates for the different
23 cancer types calculated at 0.1 ppm will differ slightly from those reported above (see Sections
24 5.2.2 and 5.2.3) because they are calculated at a level other than the PODs used in the above
25 calculations. To derive the combined risk, maximum likelihood estimates (MLEs) of risk and
26 their 95% upper bounds (UCLs) were calculated for each cancer type using the same methods
27 and life-table programs employed in sections 5.2.2 and 5.2.3. The standard errors (SEs) were
28 then estimated from the risk estimates using the equation: UCL = MLE + 1.645 x SE. The
29 variances can then be calculated from the SEs according to the equation: Variance = SE2. The
30 sum of the variances then provides an estimate of the variance for the sum of the MLEs, and the
31 95% upper bound on the sum of the MLEs can be estimated by applying the above equations in
32 reverse. Tables 5-20 and 5-21 provide a summary of the results of these calculations for the
33 combined cancer mortality and incidence risks, respectively.
34
35
This document is a draft for review purposes only and does not constitute Agency policy.
5-102 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 5-20. Calculation of combined cancer mortality unit risk estimate at
0.1 ppm
Cancer type
NPC
Hodgkin lymphoma
Leukemia
Sum
Combined risk
Combined unit riska
(per ppm)
MLE of risk
2.63 x l(T4
2.79 x l(T4
1.87 x l(T3
2.41 x 1(T3
95% upper bound
on risk
5.75 x l(T4
6.17x l(T4
3.90 x l(T3
5.09 x 10~3
4.49 x 10~3
4.49 x 10~2
SE
1.90 x 10~4
2.05 x 10~4
1.23 x 10~3
1.27 x 10~3
Variance
3.60 x 10~8
4.22 x 10~8
1.52 x 10~6
1.60 x 10~6
4
5
6
7
"Unit risk = 95% upper bound on combined risk/0.1 ppm.
Table 5-21. Calculation of combined cancer incidence unit risk estimate at
0.1 ppm
Cancer type
NPC
Hodgkin lymphoma
Leukemia
Sum
Combined risk
Combined unit riska
(per ppm)
MLE of risk
7.56 x 10~4
1.10 x 10~3
2.84 x 10~3
4.70 x 10~3
95% upper bound
on risk
1.62x 10~3
2.35 x 10~3
5.89 x 10~3
9.86 x 10~3
S.lOx 10~3
S.lOx 10~2
SE
5.25 x 10~4
7.60 x 10~4
1.85 x 10~3
2.07 x 10~3
Variance
2.76 x 10~7
5.77 x 10~7
3.44 x 10~6
4.29 x 10~6
10
11
12
13
14
15
16
17
18
19
"Unit risk = 95% upper bound on combined risk/0.1 ppm.
As can be seen from the results in Table 5-20, the upper bound risk estimates for cancer
mortality for the individual cancer types at 0.1 ppm are within 10% of the values that would be
obtained from the unit risk estimates derived in sections 5.2.2 and 5.2.3 (see Tables 5-10, 5-16,
and 5-17). Furthermore, the combined unit risk estimate for mortality for the three cancer types
(4.5 x 1Q~2 per ppm) is appropriately bounded by the mortality unit risk estimate for leukemia
(4.1 x icr2 per ppm), which has the highest individual mortality unit risk estimate, and by the
sum (5.2 x 1Q~2 per ppm) of the individual unit risk estimates presented in sections 5.2.2 and
This document is a draft for review purposes only and does not constitute Agency policy.
5-103 DRAFT—DO NOT CITE OR QUOTE
-------
1 5.2.3. Similarly, the combined risk calculated at 0.1 ppm is necessarily bounded by the sum of
2 the MLEs and the sum of the 95% upper bounds for the individual risks calculated at 0.1 ppm.
3 Thus, the value of 4.5 x 1CT2 per ppm (3.7 x 1CT5 per |ig/m3) calculated at 0.1 ppm for the
4 combined unit risk is a reasonable estimate for the total cancer mortality unit risk (based on the
5 three cancer types considered).
6 As can be seen from the results in Table 5-21, the upper bound risk estimates for cancer
7 incidence for the individual cancer types at 0.1 ppm are within 33% of the values that would be
8 obtained from the unit risk estimates derived in sections 5.2.2 and 5.2.3 (see Tables 5-11, 5-18,
9 and 5-19). Furthermore, the combined (incidence) unit risk estimate for the three cancer types
10 (8.1 x 1CT2 per ppm) is appropriately bounded by the unit risk estimate for leukemia
11 (5.7x 1CT2 per ppm), which has the highest individual unit risk estimate, and by the sum
12 (8.6 x 1CT2 per ppm) of the individual unit risk estimates presented in sections 5.2.2 and 5.2.3.
13 Similarly, the combined risk calculated at 0.1 ppm is necessarily bounded by the sum of the
14 MLEs and the sum of the 95% upper bounds for the individual risks calculated at 0.1 ppm.
15 Thus, the value of 8.1 x 10"2 per ppm (6.6 x 10~5 per |ig/m3) calculated at 0.1 ppmforthe
16 combined unit risk is a reasonable estimate for the total cancer unit risk (based on the three
17 cancer types considered).
18 As documented in Section 4.5, formaldehyde is a mutagenic carcinogen and the weight of
19 evidence supports the conclusion that formaldehyde carcinogenicity can be attributed, at least in
20 part, to a mutagenic MOA. Therefore, since there are no chemical-specific data to evaluate
21 susceptibility of different life stages, increased early-life susceptibility should be assumed, and,
22 if there is early-life exposure, the age-dependent adjustment factors (ADAFs) should be applied
23 in accordance with EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
24 Exposure to Carcinogens (U.S. EPA, 2005b). See Section 5.4.4 below for more details on the
25 application of the ADAFs.
26
27 5.3. DOSE-RESPONSE MODELING OF RISK OF SQUAMOUS CELL CARCINOMA
28 IN THE RESPIRATORY TRACT USING ANIMAL DATA
29 In the previous section, dose-response analyses based on human data for
30 lymphohematopoietic cancer and NPC were presented. The dose-response analyses of cancer
31 risk presented in this section are based on nasal tumor data from laboratory bioassays using
32 F344 rats. Because the analyses involved are extensive, most of the details are provided in the
33 appendices.
34 An increased incidence of nasal squamous cell carcinoma (SCC) was seen in two long-
35 term bioassays using F344 rats (Monticello et al., 1996; Kerns et al., 1983). Although other
This document is a draft for review purposes only and does not constitute Agency policy.
5-104 DRAFT—DO NOT CITE OR QUOTE
-------
1 studies in laboratory animals exist, these two studies, when combined, provide the most robust
2 data for analyses. These inhalation data on nasal SCC tumor incidence were used to estimate
3 human respiratory cancer risk in the nose and were also extrapolated to the entire respiratory
4 tract; in other words, a site concordance between rat and human is not assumed. This is
5 reasonable because the respiratory and transitional epithelial cell types considered to be at risk of
6 SCC in the upper respiratory tract are also prevalent in the lower human respiratory tract, and
7 there is greater penetration of formaldehyde flux posteriorly in the nose and in the rest of the
8 human respiratory tract relative to that of the rat. These considerations are strengthened by the
9 findings of DNA-protein cross-links (DPXs) in the proximal portions of the rhesus monkey
10 lower respiratory tract (Casanova et al., 1991). In addition, some epidemiologic studies
11 (Gardner et al., 1993; Blair et al., 1990, 1986) reported an increase in lung cancer associated
12 with formaldehyde exposure, while others (Collins et al., 1997; Stayner et al., 1988) reported no
13 such increases.
14 EPA's cancer guidelines (U.S. EPA, 2005a) suggest using a BBDR model for
15 extrapolation when data permit. A BBDR model for formaldehyde was developed by scientists
16 at the CUT Centers for Health Research (see Appendix D) (Conolly et al., 2004, 2003, 2000;
17 Kimbell et al., 2001a, b; Overton et al., 2001; CUT, 1999), which interfaced several models to
18 combine the extensive mechanistic information available in studies involving the F344 rat and
19 rhesus monkey and time-to-tumor incidence data in long-term bioassays, as shown by the
20 schematic in Figure 5-11. This mechanistic information included formaldehyde and DPX
21 dosimetry in the F344 rat, rhesus monkey, and human airways and cell proliferation data in the
22 F344 rat nasal lining. This document presents extensive evaluation of the underlying models and
23 data and of the alternative parametrizations of the models that were also explored for the purpose
24 of the current assessment (see Appendix E, Appendix F). A summary of conclusions is
25 presented in Section 5.3.3. In particular, the following conclusions by EPA were critical in
26 determining how the models could be used to inform the quantitative dose-response assessment:
27
28 • When used to model the dose-response in the range of the available data, the BBDR
29 models were judged to have the advantage of being more accurate and biologically based
30 (than purely statistical descriptions such as the multistage-weibull model) and allowing
31 utilization of various data in an integrated manner.
32 • Variations to modeling the F344 rat tumor incidence data in Conolly et al. (2003) were
33 examined. Given the data, each of these models, including the modeling in Conolly et al.
34 (2003), was judged to be just as biologically plausible as the other. Each of the models
35 described the rat tumor incidence equally well, was based on different characterizations
This document is a draft for review purposes only and does not constitute Agency policy.
5-105 DRAFT—DO NOT CITE OR QUOTE
-------
s. 5^
Gc TO
TO a
§
Regional respiratory tract dosimetry model
PBPK models for DPX
DPX in Rat
m N,
DPXi
DPX in Monkey
Figure 5-11. Schematic of integration of pharmacokinetic and pharmacodynamic components in the CUT
model.
fe
H
Note: |3 = death rate; (j, = mutation rate per cell division; aN, N(t), UN are informed (partially or fully) by empirical data;
other parameters are estimated by fitting to tumor incidence data.
O
O
-------
1 of the same empirical cell kinetic data, and was based on the same empirical data on DPX
2 measurements. However, the added human risk over baseline levels estimated by these
3 models (including the original model) were extremely different, and ranged from
4 negative to large positive values at environmental exposure concentrations.
5 • When used for the purpose of extrapolating risk, the BBDR models did not appear to
6 reasonably constrain either risk estimates extrapolated to human exposures or risk
7 estimates for the F344 rat when they were extrapolated below the range of observable
8 data.
9 • Human respiratory cancer risk calculated in Conolly et al. (2004) was numerically
10 unstable. Therefore, clonal growth modeling was not found to be a useful approach for
11 human extrapolation of rodent risk estimates.
12 • Thus, the biologically based derivation of human risk estimates in Conolly et al. (2004)
13 cannot be characterized as a plausible upper bound in the face of model uncertainties (a
14 key conclusion of those authors).
15
16 For all these reasons, the BBDR modeling of the rat data
17
18 • was employed in this assessment to derive multiple PODs (for SCC in the respiratory
19 tract) in the range of the observed data, using model-derived internal dose estimates,
20 • but was not used to extrapolate far below the observed data.
21
22 The inhalation unit risk estimates of SCC in the human respiratory tract were derived by
23 using multiple methods to model the F344 tumor incidence data as follows:
24
25 1. conventional multistage Weibull time-to-tumor modeling
26 2. variations of the model for rat tumor incidence implemented in Conolly et al. (2003) that
27 were considered in the process of the evaluation.
28
29 PODs were calculated as exposure concentrations corresponding to the 95% statistical
30 upper bound extra risks of 0.005, 0.01, and 0.05 (0.005 used only with BBDR modeling). The
31 inhalation unit risk for SCC in the human respiratory tract (upper and lower) derived from the
32 above animal bioassay data was then calculated by linear extrapolation to the origin from the
33 POD. Linear extrapolation is supported in part by the proven genotoxicity of the chemical and
34 the observation of cytogenetic effects in human occupational exposures (see chapter 4). In
35 particular, the formation of DPXs on formaldehyde interaction with DNA has been observed at
36 doses well below those considered cytotoxic (see Section 5.3.1.2). In results obtained in some
37 implementations of the biologically based models, formaldehyde-induced mutagenicity (modeled
This document is a draft for review purposes only and does not constitute Agency policy.
5-107 DRAFT—DO NOT CITE OR QUOTE
-------
1 as proportional to DPX concentration) was found to be a critical determinant of its
2 tumorigenicity, both at the low dose pertaining to human exposure concentrations as well as in
3 the dose range in which formaldehyde is considered to be cytotoxic.
4 The human equivalent concentration was calculated by assuming that continuous lifetime
5 exposure to a given steady-state flux of formaldehyde (expressed in pmol/mm2-hour) leads to
6 equivalent risk of nasal cancer across species. Risk per respiratory or transitional epithelial cell
7 with replicative potential was computed as a function of formaldehyde flux in the nasal region
8 and extrapolated to the rest of the respiratory tract.
9
10 5.3.1. Long-Term Bioassays in Laboratory Animals
11 This section briefly describes the various animal data and dosimetry information utilized
12 in the above (but not in all) models, based on which estimates for the inhalation unit risk are
13 derived later in this chapter.
14
15 5.3.1.1. Nasal Tumor Incidence Data
16 Various bioassays have reported the effects of formaldehyde on rats, mice, and rhesus
17 monkeys and have been discussed at length earlier in this document. Two of these bioassays
18 (Monticello et al., 1996; Kerns et al., 1983), when combined, allow for the most robust
19 characterization of the long-term dose response in a laboratory species; therefore, the focus here
20 is on these bioassays, combined. These long-term bioassays found an increased incidence of
21 nasal SCCs in rats exposed to formaldehyde by the inhalation route. In these combined data, rats
22 were exposed to 0, 0.7, 2.0, 6.01, 9.93, and 14.96 ppm (0, 0.86, 2.5, 7.4, 12.2, and 18.4 mg/m3)
23 exposure concentrations of formaldehyde. SCCs were observed only at 6.01 ppm and higher
24 exposure concentrations. Table 5-22 provides a summary of the tumors from these bioassays,
25 and the time-to-tumor characteristics are as shown by the data in Figure 5-12 (in Section 5.3.3).
26 Other tumor bioassays were also conducted by various researchers and have been detailed in
27 chapter 4.
28
This document is a draft for review purposes only and does not constitute Agency policy.
5-108 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
Table 5-22. Summary of tumor incidence in long-term bioassays on
F344 rats
Formaldehyde exposure, ppm
0.0
0.7
2
6.01
9.93
14.96
Number of animals
341
107
353
343
103
386
Number with SCC
0
0
0
3
22
162
Percent with SCC
0
0
0
0.87
21.4
42.0
4
5
6
7
Sources: Combined data from Monticello et al. (1996) and Kerns et al. (1983).
This document is a draft for review purposes only and does not constitute Agency policy.
5-109 DRAFT—DO NOT CITE OR QUOTE
-------
0.00001
0 ppm
0.20-,
0.15-
0.10-
0.00
200 400 600 800 1000 0
6 ppm
200 400 600 800 1000
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
1.0-,
0.8-
0.6-
0.4-
0.2-
0.0
10 ppm
1.0-,
0.6-
0.2-
0
200 400 600 800 1000
Age (days)
15 ppm
0 200 400 600 800 1000
Age (days)
-Kaplan-Meier Conollyfit Ourfit
Figure 5-12. Fit to the rat tumor incidence data using the model and
assumptions in Conolly et al. (2003).
Note: Fitting was performed on data of Kerns et al. (1983) and Monticello et al. (1996)
combined with ALL NTP historical controls under the assumption that all SCCs are fatal.
Figure compares the fit obtained by Conolly et al. (2003) with the reproduction of these
results under identical conditions, inputs, and assumptions by Subramaniam et al. (2007).
There were minor residual differences among the implementations; see the appendix in
Subramaniam et al. (2007) for explanation.
Source: Subramaniam et al. (2007). Reprint permission required.
5.3.1.2. Mechanistic Data
The Kerns et al. (1983) and Monticello et al. (1996) tumor studies were accompanied or
followed by additional studies that provided extensive mechanistic information on both
pharmacokinetics and pharmacodynamics. These studies have been summarized elsewhere in
this document and in other reviews (CUT, 1999; Monticello and Morgan, 1997; Morgan, 1997;
Heck et al., 1990). In addition to the tumor incidence data, the following data and mechanistic
information (some of which were model derived) are used in the quantitative models utilized in
This document is a draft for review purposes only and does not constitute Agency policy.
5-110 DRAFT—DO NOT CITE OR QUOTE
-------
1 this chapter. Additional data for the rhesus monkey are also available that inform the hazard
2 assessment but which have not been explicitly used in deriving the inhalation unit risk. Rhesus
3 monkey data have been discussed in chapter 4 and chapter 3 (DPX and formaldehyde
4 dosimetry).
5
6 • DPX: Formaldehyde interacts with DNA to form DPXs. These cross-links are
7 considered to induce mutagenic as well as clastogenic effects. Casanova et al. (1994,
8 1989) carried out two studies of DPX measurements in F344 rats. In the first study, rats
9 were exposed to concentrations of 0.3, 0.7, 2, 6, and 10 ppm for 6 hours and DPX
10 measurements were made over the whole respiratory mucosa of the rat, while, in the
11 second study, the exposure was to 0.7, 2, 6, or 15 ppm formaldehyde for 3 hours and
12 measurements were made at "high" and "low" tumor sites. DPX formation was observed
13 at all exposure concentrations in both studies (0.3 ppm-15 ppm); the DPX levels were
14 statistically significantly elevated at concentrations >2 ppm, with the trend also
15 indicating elevated DPXs at 0.7 ppm. These data were used in the development of a
16 PBPK model for predicting DPX levels in the nasal lining (see chapters 3 and 4).
17 • Cell labeling index data: Male F344 rats were exposed to formaldehyde gas over a range
18 of concentrations (0, 0.7, 2, 6, 10, or 15 ppm) in two phases of a labeling study. The first
19 phase (Monticello et al., 1991) employed injection labeling with a 2-hour pulse labeling
20 time, and animals were exposed to formaldehyde for periods of 1, 4, and 9 days and
21 6 weeks. The second phase (Monticello et al., 1996) used osmotic minipumps for
22 labeling with a 120-hour release time to quantify labeling in animals exposed for 13, 26,
23 52, and 78 weeks. These data have been analyzed at length in Appendix E.
24 • Airflow models: Physical and computer models of airflow in anatomically realistic
25 representations of the F344 rat and human upper respiratory tract have been constructed
26 (Kimbell et al., 1993, 1997a; Kepler et al., 1998; Subramaniam et al., 1998; see
27 Chapter 3).
28 • Formaldehyde dosimetry: Regional uptake of formaldehyde has been calculated for the
29 upper respiratory tract of the rat and human by using the above computer representations
30 and for the lower respiratory tract of the human by using an idealized representation of
31 the human lower respiratory tract (Kimbell et al., 2001a; Overton et al., 2001; also see
32 chapter 3 and further discussion of uncertainties in Appendix F).
33
34 5.3.2. The CUT Biologically Based Dose-Response Modeling
35 The studies mentioned above in 5.3.1.1 and 5.3.1.2 were generated at the CUT Centers
36 for Health Research and led to the development of a biologically motivated dose-response model
37 for formaldehyde-induced cancer as represented in a series of papers and in a health assessment
38 report (CUT model) (Conolly et al., 2004, 2003, 2000; Conolly, 2002; Kimbell et al., 2001a, b;
39 Overton et al., 2001; CUT, 1999). EPA's cancer guidelines (U.S. EPA, 2005a) suggest using a
40 BBDR model for extrapolation when data permit since it facilitates the incorporation of MO A in
This document is a draft for review purposes only and does not constitute Agency policy.
5-111 DRAFT—DO NOT CITE OR QUOTE
-------
1 risk assessment. The CUT modeling and available data were evaluated in a series of peer-
2 reviewed papers (Klein et al., 2010; Crump et al., 2008; Subramaniam et al., 2008, 2007) and
3 debated further in the literature (Conolly et al., 2009; Crump et al., 2009). Alternatives to the
4 parametrization and model structure in the CUT biological modeling (but based on that original
5 model) are further explored and evaluated in this assessment (Appendix E). Appendix F carries
6 out a sensitivity analysis of the human risk estimates in Conolly et al. (2004) based on key
7 uncertainties evaluated in Appendix E. These BBDR models are used in this assessment to
8 calculate PODs from the dose-response curve for the F344 rat nasal tumor risk; extrapolation to
9 human is then carried out by using EPA's baseline ("default") approach (U.S. EPA, 1994) but
10 using model-derived internal dose metrics for rat and human. See Section 5.3.3 for rationale
11 supporting these decisions.
12 First, the key features of the BBDR modeling in Conolly et al. (2003, 2004) are briefly
13 described, and the following notation is used throughout this section: N cell = normal cell; I cell
14 = initiated cell; LI = labeling index and is equal to the number of labeled cells/(number labeled
15 + unlabeled cells); ULLI = unit length LI equal to the number of labeled cells/length of basement
16 membrane; aN = division rate of normal cells (hour1); JIN = rate at which an initiated cell is
17 formed by mutation of a normal cell (per cell division of normal cells).
18 In Conolly et al. (2003), tumor incidence data in the Kerns et al. (1983) and Monticello et
19 al. (1996) long-term bioassays were modeled by using an approximation of the two-stage clonal
20 growth model (Moolgavkar et al., 1988) and allowing formaldehyde to have a direct mutagenic
21 action. Conolly et al. (2003) combined these data with historical control data on 7,684 animals
22 obtained from National Toxicology Program (NTP) bioassays. These models are based on the
23 Moolgavkar, Venzon, and Knudson (MVK) stochastic two-stage model of cancer (Moolgavkar
24 et al., 1988; Moolgavkar and Knudson, 1981; Moolgavkar and Venzon, 1979), which accounts
25 for growth of a pool of normal cells, mutation of normal cells to initiated cells, clonal expansion
26 and death of initiated cells, and mutation of initiated cells to fully malignant cells.
27 The MVK model for formaldehyde accounted for two MO As as follows that may be
28 relevant to formaldehyde carcinogenicity:
29 1. An indirect MOA in which the regenerative cell proliferation in response to
30 formaldehyde cytotoxicity increases the probability of errors in DNA replication. This
31 MOA was modeled by using labeling data on normal cells in nasal mucosa of rats
32 exposed to formaldehyde.
33 2. A possible direct mutagenic MOA, based on information indicating that formaldehyde is
34 mutagenic (Speit and Merk, 2002; Heck et al., 1990; Grafstrom et al., 1985), was
35 modeled by using rat data on formaldehyde production of DPXs (Monticello et al., 1996,
This document is a draft for review purposes only and does not constitute Agency policy.
5-112 DRAFT—DO NOT CITE OR QUOTE
-------
1 1991). In Conolly et al. (2003), the intracellular dose that induces mutations is
2 considered proportional to the local DPX dose.
3
4 The human model for formaldehyde carcinogenicity (Conolly et al., 2004) is
5 conceptually very similar to the rat model. The model uses, as input, results from a dosimetry
6 model for an anatomically realistic representation of the human upper airways and an idealized
7 representation of the lower airways. However, the model does not incorporate any data on
8 human responses to formaldehyde exposure.
9 A novel contribution of the CUT model, described by the schematic in Figure 5-11, is
10 that cell replication rates and DPX concentrations are driven by local dose, which is
11 formaldehyde flux to each region of nasal tissue expressed as pmol/mm2-hour. This dosimetry is
12 predicted by computational fluid dynamics (CFD) modeling using anatomically accurate
13 representations of the nasal passages of a single F344 rat or Caucasian male human (see
14 chapter 3). Such a feature is important in incorporating site-specific toxicity in the case of a
15 highly reactive gas like formaldehyde for which uptake patterns are spatially localized and
16 significantly different across species (see chapter 3). In the CUT model, each of these
17 parameters is characterized by local flux (see Figure 5-11). The inputs to the two-stage cancer
18 modeling consisted of results from other model predictions as well as empirical data as follows:
19
20 • Regional uptake of formaldehyde in the respiratory tract was predicted by using CFD
21 modeling in the F344 rat and human (Kimbell et al., 1997a, 2001a, b; Overton et al.,
22 2001; Subramaniam et al., 1998).
23 • Replication rates for normal cells were inferred from LI data on rats exposed to
24 formaldehyde (Monticello et al., 1996, 1991, 1990).
25 • Concentrations of DPXs linked to the regional flux of formaldehyde were predicted by a
26 PBPK model (Conolly et al., 2000) calibrated to fit the DPX data in F344 rat and rhesus
27 monkey (Casanova et al., 1994, 1991) and subsequently scaled up to humans. The DPX
28 concentration levels were incorporated into the two-stage clonal expansion model by
29 defining mutation rate of normal and initiated cells as the same linear function of DPX.
30 That is,
31
32 HN = m = n^ai + KMU x DPX (5-1)
33
34 where (IN is the rate at which an initiated cell is formed by mutation of a normal cell (per
35 cell division of normal cells), and likewise jii is the rate at which a malignant cell is
36 formed by mutation of an initiated cell (per cell division of initiated cells). The unknown
This document is a draft for review purposes only and does not constitute Agency policy.
5-113 DRAFT—DO NOT CITE OR QUOTE
-------
1 constants UNbasai (the baseline rate) and KMU were estimated by fitting model predictions
2 to the tumor bioassay data.
3
4 The rat model in Conolly et al. (2003) involved six unknown statistical parameters that
5 were estimated by fitting the model to the rat formaldehyde bioassay data shown in Table 5-22
6 (Monticello et al., 1996; Kerns et al., 1983) plus data from several thousand control animals
7 from all the rat bioassays conducted by the NTP. These NTP bioassays were conducted from
8 1976 through 1999 and included 7,684 animals with an incidence of 13 SCCs (i.e., 0.17%
9 incidence). The resulting model predicts the probability of a nasal SCC in the F344 rat as a
10 function of age and exposure to formaldehyde. The fit to the tumor incidence data is shown in
11 Figure 5-12 (in Section 5.3.3.). (For later reference in Appendix E, this figure compares the fit
12 to the data obtained by the modeling in Conolly et al. [2003] with that obtained by the
13 reimplementation of this model in Subramaniam et al. [2007].)
14 Subsequent to the BBDR model for modeling rat nasal cancer, Conolly et al. (2004)
15 developed a corresponding model for humans for the purpose of extrapolating the risk to humans
16 estimated by the rat model. Also, rather than considering only nasal tumors, the model is used to
17 predict the risk of SCC in the entire human respiratory tract. The human model for
18 formaldehyde carcinogenicity in Conolly et al. (2004) is conceptually very similar to the rat
19 model in Conolly et al. (2003) and follows the schematic in Figure 5-11. The following points
20 need to be noted:
21
22 • The model does not incorporate any data on human responses to formaldehyde exposure.
23 • The model is based on an anatomically realistic representation of the human nasal
24 passages (in a single individual) and an idealized representation of the lower respiratory
25 tract. Local formaldehyde flux to respiratory tissue is estimated by a CFD model for
26 humans (Kimbell et al., 2001a; Overton et al., 2001; Subramaniam et al., 1998).
27 • Rates of cell division and cell death are, with a minor modification, assumed to be the
28 same in humans as in rats.
29 • The concentration of formaldehyde-induced DPXs in humans is estimated by scaling up
30 from values obtained from experiments in the F344 rat and rhesus monkey (Conolly
31 et al., 2000, and also discussed further in Section 3.6.6 of this document). The human
32 value for KMU in Equationl is obtained by assuming that the ratio KMU/|ibasai is
33 invariant across species. The other statistical parameters for the human model are either:
34 (a) estimated by fitting the model to the human background incidence of tumors, (b)
35 assumed to have the same value as that obtained in the rat model, or, (c) in one case,
36 fixed at a value suggested by the epidemiologic literature.
37
This document is a draft for review purposes only and does not constitute Agency policy.
5-114 DRAFT—DO NOT CITE OR QUOTE
-------
1 Some further clarification pertaining to the structure and calibration of the models in
2 Conolly et al. (2004, 2003) that are key to understanding model assumptions is provided in
3 Appendix D.
4
5 5.3.2.1. Major Results of the CUT Modeling Effort
6 Based on the biologically based modeling of the rat SCC data, CUT (1999) and Conolly
7 et al. (2004, 2003) presented the following major conclusions. The evaluation of the strength of
8 these conclusions is summarized in Section 5.3.3., and as addressed in that section, this current
9 assessment is not in agreement with these conclusions.
10
11 • The putative, directly mutagenic action of formaldehyde "does not play a significant role
12 in the tumor response in the rat (and also in the human), [and such a conclusion] should
13 be robust for any potentially mutagenic effect of formaldehyde with a time course similar
14 tothatofDPX."
15 • Respiratory cancer risks associated with inhaled formaldehyde are de minimis (10~6 or
16 less) at relevant human exposure levels. This was based on using an upper bound on the
17 model estimate for the directly mutagenic action of formaldehyde.
18
19
inwu^i v^duinaLV^ j-wi uiv^ UIIV^LIJ iiiuLa^v^iiiv^ av^uwii wi j.wniiaiu^iij uv^.
Therefore, exposure standards protective of effects of formaldehyde-induced cytotoxicity
should be sufficient to protect from its potential carcinogenic effects.
20 • The human risk estimates in Conolly et al. (2004) were judged by the authors to be
21 conservative in the face of model uncertainties because the model: (a) included a hockey-
22 stick model for normal cell replication rates when the cell replication dose-response
23 curve as averaged by the authors had a J shape, (b) used overall respiratory tract cancer
24 incidence data in humans, and (c) evaluated the model at the statistical upper bound of
25 the proportionality parameter relating DPXs to the probability of mutation.
26 • The dose-response assessment in Conolly et al. (2004) did not explicitly evaluate the risk
27 of lymphohematopoietic cancers. However, Conolly et al. (2004) argued that
28 formaldehyde was unlikely to cause the cancers reported in Hauptmann et al. (2003).
29 Their reasoning was based on the steepness of the dose-response curve predicted in
30 Conolly et al. (2004) for respiratory cancer at exposures of 1 ppm and above, and the
31 conclusions in Heck and Casanova (2004).
32
33 5.3.3. This Assessment's Conclusions from Evaluation of Dose-Response Models of DPX,
34 Cell-Replication and Genomics Data, and of BBDR Models for Risk Estimation
35 The CUT modeling of the rat tumor incidence and mechanistic information detailed in
36 Section 5.3.1 and alternative models that were developed based on the conceptual framework in
37 the CUT modeling were extensively evaluated for this assessment. These results are presented in
38 Appendices D, E (BBDR modeling of the rat data), and F (sensitivity analysis of BBDR model
This document is a draft for review purposes only and does not constitute Agency policy.
5-115 DRAFT—DO NOT CITE OR QUOTE
-------
1 results for human risk). In particular, Table E-l in Appendix E and Table F-l in Appendix F
2 tabulate all the uncertainties and assumptions that were examined along with results of that
3 evaluation. The quantitative and qualitative characterization of the cell replication data from
4 Monticello et al. (1996, 1991) are presented in Appendix E. The most significant conclusions
5 resulting from these various analyses, focusing on the ones that have maximal impact on the
6 dose-response assessment, are presented below.
7
8 Description of Time-to-Tumor Data
9 The overall approach and use of data in Conolly et al. (2004, 2003) have substantial
10 advantages to offer in describing the dose response observed in animal bioassays. The authors'
11 model provides a good statistical description of the time-to-tumor data. The fit to the data was
12 found to be superior to that obtained by using multistage-Weibull time-to-tumor modeling of the
13 tumor incidence data (comparison based on visual inspection [see Figure 5-12 in this section and
14 Figures 5-17, 5-18, 5-19 in Section 5.3.4]).
15
16 Integration of Various Relevant Data
17 The model framework integrates various pharmacokinetic and pharmacodynamic
18 components (regional formaldehyde flux, DPX, cell-replication, and tumor incidence data)
19 within a single conceptual framework and thus facilitates description of the tumor dose response
20 that utilizes the extensive mechanistic information available for formaldehyde.
21
22 Regional Dosimetry
23 Regional (site-specific) dosimetry in the upper respiratory tract is considered important
24 for understanding the tumorigenicity of a reactive chemical like formaldehyde. The regional
25 dosimetry models discussed in chapter 3 compute local formaldehyde flux to the tissue and are
26 based on anatomically realistic constructions of the nasal airways in each species. The other
27 relevant mechanistic data, DPX and cell replication, are expressed as a function of this local
28 formaldehyde flux.
29
30 Confidence in Dosimetry
31 Model predictions of formaldehyde flux to the respiratory lining have not been verified
32 experimentally, and such verification would present formidable experimental challenges.
33 Overall, the formaldehyde dosimetry modeling utilized in the CUT modeling presents a
34 reasonable level of confidence, as detailed in chapter 3, Section 3.6, by virtue of agreement
35 among multiple model predictions (models that predict airflow profiles as well as a PBPK model
This document is a draft for review purposes only and does not constitute Agency policy.
5-116 DRAFT—DO NOT CITE OR QUOTE
-------
1 for DPX, which uses the calculated formaldehyde flux as input) and various kinds of available
2 data. These data comprise airflow profiles in physical casts of the nasal cavity of an F344 rat
3 (Kimbell et al., 1997a, 200la), a human (Subramaniam et al., 1998), and a rhesus monkey
4 (Kepler et al., 1998); DPX data (see discussion of Cohen-Hubal et al. [1997] in chapter 3); and
5 qualitative concordance between uptake patterns and cell proliferation (Morgan et al., 1997;
6 Monticello et al., 1996). The CFD models of formaldehyde flux represent only an individual of
7 each species. However, considerable interindividual differences are to be expected in the
8 regional dosimetry, particularly in the human (Garcia et al., 2009: Subramaniam et al., 2008).
9 This is discussed briefly in Chapter 3 (see Section 3.6) and further in Appendices B and F.
10
11 Control Tumor Data
12 In developing their model, Conolly et al. (2004, 2003) included control rats from all NTP
13 cancer bioassays—a total of 7,684 rats. As elaborated in Appendix E, lumping all NTP
14 historical control animals along with the control animals in the Kerns et al. (1983) and
15 Monticello et al. (1996) inhalation bioassays does not appear to be supportable and substantially
16 alters dose-response predictions (Crump et al., 2009, 2008; Subramaniam et al., 2008, 2007).
17 There are legitimate questions regarding comparability of results in rats from different stocks,
18 studied at different times, in different laboratories, and by different routes of exposure and
19 evaluated by using somewhat different pathological procedures (Haseman, 1995; Rao et al.,
20 1987). If historical controls are used from only those inhalation studies that present a low
21 potential for genetic and time-related variations in tumor incidence and survival of animals or if
22 only concurrent controls are used, the model for extrapolation of risk to humans (the human
23 BBDR model) becomes numerically unstable. In such a model, it is not possible to bound
24 human risk by using the extrapolation approach applied in the CUT model. When the included
25 NTP control data were restricted to those from NTP inhalation studies, the upper bound human
26 risk estimate obtained by Conolly et al. (2004) (i.e., with everything else in their modeling
27 retained unchanged) was increased by 50-fold (Crump et al., 2008).
28
29 Cell Replication Dose Response
30 As discussed in chapter 4, characterization of the uncertainties and variability in the cell
31 replication dose response is crucial to understanding formaldehyde carcinogenicity. Analyses of
32 the data in Monticello et al. (1991, 1996) to derive dose response curves for cell replication are
33 presented in Appendix E and are partly published in Subramaniam et al. (2008). The raw
34 individual animal data from this bioassay were made available to EPA. The analyses
35 demonstrate the following:
This document is a draft for review purposes only and does not constitute Agency policy.
5-117 DRAFT—DO NOT CITE OR QUOTE
-------
2 • Sustained exposure to formaldehyde affects cell division rates (compared to baseline
3 levels) over a continuum of formaldehyde flux to the nasal lining that includes flux levels
4 below those thought to be cytotoxic.
5 • Given the qualitative and quantitative uncertainties in the data and in their interpretation,
6 a variety of cell replication dose-response models are plausible as reasonable
7 characterization of the data. Cell replication response differs substantially among nasal
8 sites and over time during the course of the bioassay. In consideration of these
9 differences, the dose response for cell replication included shapes that were monotonic
10 increasing as well as nonmonotonic at low dose (also see Meng et al. 2010 and a
11 discussion of their data in Appendix F). For example, rather different statistical
12 descriptions of the data result depending on whether
13 i. different sites and exposure times were modeled separately;
14 ii. all exposure times were pooled to model the response at each site;
15 iii. the labeling index was time-weighted and averaged over all sites;
16 iv. flux and labeling index were weighted by the number of cells at a given site;
17 v. the short exposure durations in Monticello et al. (1991) were examined separately.
18 In addition, transient increases in cell turnover at subcytotoxic doses are seen in
19 other experiments in rats exposed to formaldehyde (see chapter 4).
20 • At higher, cytolethal formaldehyde flux levels, regenerative hyperplasia-induced cell
21 proliferation clearly takes over.
22
23 Genotoxicity
24 Chapter 4 provides multiple lines of evidence to characterize formaldehyde as a
25 genotoxicant. Of particular note is the observation of cytogenetic effects at human occupational
26 exposures and the formation of DPXs upon formaldehyde interaction with DNA at doses well
27 below those considered cytotoxic. As noted earlier, DPX formation was detected in rats at
28 exposures ranging from 0.3 ppm to 15 ppm. These DPX levels are seen to be statistically
29 significantly increased over baseline levels at 2 ppm and above. The DPX measured at 0.7 ppm
30 shows a trend that is consistent with an increase at this dose (see chapter 3), and it is critical to
31 consider "trend" when analyzing low-dose data.
32
This document is a draft for review purposes only and does not constitute Agency policy.
5-118 DRAFT—DO NOT CITE OR QUOTE
-------
1 Inferences on MOAfrom Modeling the Data
2 The highly curvilinear nature of dose responses associated with DPX formation, LI data,
3 and tumor response, as well as mechanistic interpretation of these observed data, have provided
4 grounds for arguments in the literature that formaldehyde tumorigenicity (at exposures >6 ppm)
5 should be uncoupled from its potential carcinogenicity in the low-dose region. Furthermore,
6 some researchers have argued that any potential low-dose risk is due to formaldehyde's
7 mutagenicity, that this mutagenic potential is too weak to be of significance, and that the
8 observed risk is entirely due to cell proliferation induced by regenerative hyperplasia in response
9 to cell injury at cytotoxic doses (i.e., without a relevant role for the direct mutagenic action of
10 formaldehyde). Conolly et al. (2004, 2003) represented a quantitative expression of this point of
11 view. However, alternative parametrizations of the model used in Conolly et al. (2004, 2003)
12 have shown that the mutagenic component can be important to explaining the observed tumor
13 incidence and that the risk at low dose due this mutagenicity can be significant (Subramaniam
14 et al., 2007; Appendix E).
15 As mentioned in 5.3.3.6, analysis of the considerable uncertainty-variability in the cell
16 labeling data indicates that, upon exposure to formaldehyde, cell replication is significantly
17 altered over a continuum that includes low and high concentration levels. At high dose, the
18 effect on cell replication is regenerative. At lower doses, the data indicate that both monotonic
19 and nonmonotonic dose-response curves for cell replication are plausible. Various plausible
20 dose-response curves for cell replication were incorporated into the alternate BBDR models
21 evaluated in this assesment (see Appendix E) and were seen to strongly influence the low-dose
22 response curves for risk. The following exercise was particularly instructive in illuminating the
23 uncertainty in the shape of the dose-response curve at low dose. The BBDR models were
24 exercised with normal cell replication rates considered to be less than (nonmonotonic) or equal
25 to (threshold) baseline rates over a segment of the low-dose range. Such a scenario did not
26 necessarily lead to lower than baseline or threshold in formaldehyde respiratory cancer risk in
27 the rat in that low-dose range13. This is partly because there are no data to inform how
28 formaldehyde-induced mutation might alter cell replication and apoptotic rates (in particular if
29 the mutation is to be construed as an initiating event in the carcinogenesis).
30 Accordingly, the dose-response assessment in this document does not treat formaldehyde
31 as a threshold carcinogen.
32
13 all the models reproduced the chronic time-to-tumor data well
This document is a draft for review purposes only and does not constitute Agency policy.
5-119 DRAFT—DO NOT CITE OR QUOTE
-------
1 Kinetics of Initiated Cells
2 Modeling results are hypersensitive to the division and death rate of initiated cells that
3 cannot be further inferred by the available empirical cell labeling data (Conolly et al., 2009;
4 Crump et al., 2009, 2008). Several plausible alternate model structures for describing initiated
5 cell kinetics, none of which degrade the agreement of the model with the underlying data used to
6 construct the model originally, led to low-dose risk estimates in the rodent that varied by many
7 orders of magnitude, including negative values (see Figures E-5A,B and E-6A,B in Appendix E).
8 Extremely small perturbations in the division rate (and, likewise, of death rates) of
9 initiated cells in the model lead to human risk estimates ranging anywhere from negative values
10 to +0.01 at 0.01 ppm (see Crump et al. 2008 and Appendix F, Figure F-5). These perturbations
11 were small compared with the normal variation in the division rates of normal cells.
12 The sensitivity analyses on the basis of which these conclusions were reached have been
13 criticized as resulting in implausible risk estimates (given the epidemiologic data) as a
14 consequence of implementing model variations that are not biologically reasonable (Conolly
15 et al. 2009). This criticism was rebutted by Crump et al. (2009) on biological and
16 epidemiological grounds. These debates are discussed fully in Appendix F.
17 In addition, there are major qualitative uncertainties in extrapolating normal cell
18 replication rates from the rat to human (see Table F-l in Appendix F, and Subramaniam et al.
19 [2008]). Subramaniam et al. (2008) examined the inferences that arise from the assumptions in
20 the CUT model on initiated cell replication and death rates and concluded that several inferences
21 were not supportable on the basis of available biological information (see Appendix E, Section
22 E.3.3.1 for a summary).
23
24 Risk Extrapolation
25 The modeling approach in the human formaldehyde model of Conolly et al. (2004) and
26 the variations examined showed extreme sensitivity, including numerical instability, to uncertain
27 model assumptions. This model, and the alternative BBDR models examined, were therefore
28 determined not to be informative for extrapolation from animal to human at any exposure
29 concentration. In the face of model uncertainties, the biologically based derivation of human
30 risk estimates of 10"6 or less at exposures of 0.1 ppm and below in Conolly et al. (2004) or CUT
31 (1999) cannot be characterized as a plausible upper bound.
32 The use of clonal growth modeling for extrapolation of risk from high to low exposures
33 in the rodent followed by a conventional (default) approach to extrapolate the low-dose animal
34 risk to the low-dose human risk was next evaluated. However, as explained earlier, the models
35 do not adequately constrain risk in the rodent. For example, various model representations as
This document is a draft for review purposes only and does not constitute Agency policy.
5-120 DRAFT—DO NOT CITE OR QUOTE
-------
1 shown in Figure E-6A,B in Appendix E were used to evaluate added MLE risk at the 10 5 level
2 (see Figure F-5A,B in Appendix E) in the F344 rat. Human exposures were then calculated that
3 would result in equivalent lifetime risk by using formaldehyde flux estimated in each species as
4 the dosimeter and conventional extrapolation methods (U.S. EPA, 1994b). A 25-fold difference
5 was found between the different models in the equivalent exposure concentration so derived.
6 Model uncertainty was substantially higher than the statistical uncertainty arising out of a given
7 model specification. Therefore, this avenue was also found not to be informative.
8 Consequently, the CUT model or its variations were not used in this assessment as a
9 biologically-based or biologically-motivated means of extrapolating outside the observed dose-
10 response in the F344 rat.
11 Thus, in view of all the above considerations and in accordance with EPA's cancer
12 guidelines (U.S. EPA, 2005a), the derivation of unit risk for human respiratory cancer from
13 animal bioassay data in this document is based on a linear extrapolation to the origin from a
14 POD on the dose-response curve. Low-dose linearity was exhibited by the risk estimates from
15 most of the models that were examined in the sensitivity analysis (see discussion surrounding
16 Figure E-5A,B in Appendix E).
17
18 BBDR Modeling for Deriving an "Integrated" POD
19 The CUT BBDR modeling approach provides a good fit to the time-to-tumor data and
20 therefore allows for an appropriate determination of a POD while at the same time incorporating
21 a large amount of mechanistic information in an integrated manner and allowing the use of
22 model-derived internal dose estimates. Thus, use of this model provides an alternative to
23 developing separate PODs based on several of the underlying components of the data, such as
24 DPX, flux, and labeling data. Accordingly, the model is used in this assessment to derive a POD
25 from a dose response, based on the nasal cancers in rats. Uncertainties in the derivation of the
26 POD were represented by using the variations of the CUT model examined in this chapter.
27 These POD calculations as well as others are detailed below.
28
This document is a draft for review purposes only and does not constitute Agency policy.
5-121 DRAFT—DO NOT CITE OR QUOTE
-------
1 Genomics Data
2 The genomics data of Thomas et al. (2007) and Andersen et al. (2008) provide additional
3 insight into formaldehyde's biological effects in the URT and the steep dose-response curve for
4 tumorigenesis. However, as summarized in a review by Chiu et al. (2010), there are various
5 limitations in the interpretation of these genomics data and their relevance for the pathways
6 contributing to the disease process in humans. In particular, the data from these studies, as
7 analyzed, do not inform the critical MOA questions pertaining to formaldehyde carcinogenicity.
8 These insights have been elaborated in Section 4.4.5, and the difficulties in the use and
9 interpretation of the quantitative modeling of these data, as presented in these studies, are
10 detailed at length in Appendix G.
11
12 5.3.4. Benchmark Dose Approaches to Rat Nasal Tumor Data
13 This section describes various BMD analyses to determine PODs for low-dose
14 extrapolation of SCC risk in the human respiratory tract (upper and lower).
15
16 5.3.4.1 .Benchmark Dose Derived from BBDR Rat Model and Flux as Dosimeter
17 5.3.4.1.1. Response for benchmark dose.
18 Typically, the BMD is calculated at the 5 or 10% response level. However, it appears
19 appropriate to consider the benchmark response (BMR) at lower levels in exceptional cases that
20 are supported by empirical data. In the case of data combined from the Kerns et al. (1983) and
21 Monticello et al. (1996) bioassays, the lowest observed tumor response of SCC was below the
22 1% level (at 0.85%) (see Table 5-22). Additionally, the BBDR modeling incorporates precursor
23 response in the form of LI data. Therefore, it was determined that it would also be appropriate to
24 evaluate the POD at the 0.5% level while still staying in the neighborhood of the experimentally
25 observed response.
26 The various data presented earlier in this chapter point to highly curvilinear dose
27 responses for formaldehyde-induced tumor incidence as well as DPX and cell replication. This
28 is also borne out by dose-response information based on gene array data (Thomas et al. 2007;
29 Andersen et al. 2008). Cytotoxicity-driven regenerative replication and epithelial degeneration
30 play a critical role in the steeply rising nature of the tumor dose-response. These observations
31 raise the concern that cancer potency derived by straight-line extrapolation from the low end of
32 observed tumor data (roughly at the 1% response) has the potential to be a significant
33 overestimate for a reasonable upper bound. The pertinent question then is: what is a low-dose
34 linear dose-response modeling of the data that is statistically consistent with the uncertainties in
35 the observed time-to-tumor data. To address this question, the risk estimate based on the linear
This document is a draft for review purposes only and does not constitute Agency policy.
5-122 DRAFT—DO NOT CITE OR QUOTE
-------
1 extrapolation from a POD (based on the statistical upper confidence bound on risk) to the origin
2 is compared with that predicted at the low-dose end by the Multistage-Weibull model fitted to
3 the observed time-to-tumor data. The unit risk based on this model is obtained by calculating
4 ql *, the 95% statistical upper bound on the coefficient associated with the linear term in the
5 multistage model polynomial. This model fits the data reasonably well, reflects the highly
6 curvilinear shape of the dose-response because of its mathematical flexibility, and allows for the
7 possibility of low-dose linearity. Thus, for comparison the following estimates of unit risk are
8 also presented (in addition to the unit risks calculated at the 1%, 5% and 10% response levels):
9
10 1. Unit risk that is based on ql *, which is derived from fitting the multistage Weibull model
11 to the observed data.
12 2. Unit risk based on low-dose linear extrapolation from a POD at the 0.5% level.
13
14 5.3.4.1.2. Dose metric.
15 The dose metric used for the extrapolation was the average wall mass flux of
16 formaldehyde (expressed in pmol/mm2-hour to the entire surface of the airway lining but
17 excluding tissue lined by nonmucus-coated squamous tissue, which was considered to not absorb
18 formaldehyde). The use of flux as a dosimeter is similar to the calculation of a regional gas dose
19 ratio (RGDR) as proportional to minute volume divided by the surface area in the given species
20 and is thus in line with EPA's guidance for calculating a dosimetric adjustment factor (DAF) for
21 category 1 gases, whose effects are presumed to be at the POE (U.S. EPA, 1994b) (i.e., ratio of
22 average flux over the same respiratory region in each species = ratio of the quantity [minute
23 volume/surface area of the region] between the two species). This lends support to an
24 interspecies extrapolation based on the equivalence of formaldehyde flux as a determinant of
25 risk.
26 The spatial distribution of formaldehyde over the nasal lining was characterized by
27 partitioning the nasal surface by formaldehyde flux to the tissue, resulting in 20 "flux bins" (see
28 Figure 5-13). Each bin is comprised of elements (not necessarily contiguous) of the nasal
29 surface that receive a particular interval of formaldehyde flux per ppm of exposure concentration
30 (Kimbell et al., 2001b). The spatial coordinates of elements comprising a particular flux bin are
31 fixed for all exposure concentrations, with formaldehyde flux in a bin scaling linearly with
32 exposure concentration (ppm). The number of cells at risk varies across the bins, as shown in
33 Figure 5-14.
34
This document is a draft for review purposes only and does not constitute Agency policy.
5-123 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
5.3.4.1.3
two altei
were use
opposed
. Extrapolation to humans. For linear extrapolation from the 0.5 and 1% levels,
•native versions of the biologically based model in Conolly et al. (2003) for the F344 rat
d. In both cases, only the historical control data from NTP inhalation studies (as
to all NTP studies) were added to the concurrent controls and weekly averaged DPX
concentrations as calculated by Subramaniam et al. (2007) (who implemented a variant of the
PBPK model in Conolly et al. 2000) were used. Both models provided good fits to the tumor
incidence data, similar to the fit shown in Figure 5-12. Neither model could be considered better
than the other on the basis of model description of tumor incidence data. The values of the
parameters in these models and their fit to the data are provided in Tables E-4 and E-5 of
Appendix E.
3000-
£ 2500 -
o
O „
Flux Per Unit Exposure
[pmol/(mm2-h-pprr
o 8 8 8
O O O O O
. ' ^ 3e+7 I
'= 1
•s \
. u> 2e+7 |
« 1
•s 1
d 1e+7 I
z I
• c> |- A r\ * r on
u o nu lo AJ
5 10 15 20
c, o;^ Flux Bin No.
13
14
15
16
17
18
19
20
Figure 5-13. Spatial distribution of
formaldehyde over the nasal lining, as
characterized by partitioning the nasal
surface by formaldehyde flux to the
tissue per ppm of exposure
concentration, resulting in 20 flux bins.
Source: Subramaniam et al. (2008).
Figure 5-14. Distribution of cells at risk
across flux bins in the F344 rat nasal
lining.
Source: Subramaniam et al. (2008).
In Model 1 the normal cell replication dose response was described by the same
hockey-stick-shaped curve used in Conolly et al. (2003). The form of the dose-response
curves for initiated cell kinetics (division and death) was also the same as that considered by
Conolly et al. (2003). This model is the same as Model E in Table III of Subramaniam et al.
(2007) and values of the parameters and model fit to the data can be obtained from their
Table.
This document is a draft for review purposes only and does not constitute Agency policy.
5-124 DRAFT—DO NOT CITE OR QUOTE
-------
1 Model 2 was an alternative to the Conolly et al. (2003) model, and is denoted as
2 Model 15 in the sensitivity analysis described in Appendix E (see Figures E-6A and
3 Table E-4 for parameter values). The dose response for replication of normal cells was
4 monotone increasing and did not exhibit a threshold in dose. This was obtained by fitting
5 the 13-week cell replication data in Monticello et al. (1996). The raw replicate animal data
6 from this study was provided to EPA by the Hamner Institutes for Health Research. The
7 cell replication dose response for initiated cells was a sigmoidal-shaped curve, increasing
8 monotonically with flux from a background value up to an asymptotic value. The baseline
9 cell-replication for initiated cells was constrained to be equal to or greater than the baseline
10 rate of division of normal cells. Initiated cell death rate was considered proportional to
11 initiated cell birth rate. The biological rationale for these choices is given in Appendix E.
12 Models 1 and 2 predicted monotonic dose-response curves.
13 The sequence of steps in arriving at a unit risk for SCC in human nasal airways from
14 a given BBDR modeling of the F344 rat nasal tumor incidence data is outlined below.
15 Extrapolation to the lower respiratory tract is described later.
16
17 1. Calculate the MLE risk and 95% upper confidence bound on risk at various exposure
18 concentrations (dRAxin ppm) by exercising the two BBDR models. Here, the POD is
19 defined as dRAx for which the 95% upper bound added risk is either 0.005 or 0.01.
20 These values approximate the 95% lower bounds on the BMD corresponding to the
21 added risks (i.e., the BMDLRAx).
22 2. Using CFD modeling simulations in Kimbell et al. (2001b), calculate the average
23 flux over the entire rat nose at resting breathing rates corresponding to dRAx- Here,
24 the subscript "i" is over flux bins and N is the number of cells at risk in a given bin.
25
26 AvgFlux(d
RAT
-\RAT
27
(5-3)
28 3. The experiment exposure was for periods of 5 days/week, 6 hours/day. Therefore,
29 calculate the average daily exposure, obtained by making a 5/7 x 6/24 duration
30 adjustment; that is, 5/7 x 6/24 x AvgFlux(dRAx).
31 4. Now assume that lifetime exposure to similar levels of average formaldehyde flux to
32 cells at risk leads to similar lifetime risk (MLE or upper bound, respectively) of
33 tumor incidence across animal species. Also, in calculating human equivalent
This document is a draft for review purposes only and does not constitute Agency policy.
5-125 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
concentrations, EPA has traditionally assumed chronic animal laboratory exposure
scenarios to be equivalent to human lifetime exposures (U.S. EPA, 1994b).
5. Since a CFD model for a human upper respiratory tract is available (Subramaniam
et al., 1998), it is possible to determine the average wall mass flux in this particular
human nose for any specific breathing scenario. Likewise, a computational "single-
path" model to determine average mass flux at any specific lung depth was available
(Overton et al. 2001); however, risk in the lower respiratory tract will be addressed
later. From the human CFD simulations in Kimbell et al. (200la, b), the human
airborne exposure concentration level that would yield an average wall mass flux in
the human nose equal to [(5/7) x (6/24) x AvgFlux(dRAi)] is then calculated. In
other words, given a risk-specific dose in the rat, the equivalent human exposure
concentration is given by
6.
7.
d
HUMAN
= (5 / 7) x (6 / 24) x AvgFlux(dRAT ) x
ppm
(5-4)
HUMAN
To use this equivalent human exposure concentration, make the following
assumption: when humans are exposed to the above concentration of formaldehyde
(dHUMAN) throughout the course of a lifetime, the added risks are anticipated to be
similar to those experienced by the animal in the chronic bioassay.
Let/ denote the ratio of the average flux per ppm of exposure concentration in the
two species:
/ =
(5-5)
Now, the olfactory epithelium comprises a substantial fraction of nasal tissue in the
rat. Because the olfactory region in the rat projects directly in the path of main
airstreams (Kimbell et al., 1997a), a sizable flux of formaldehyde is delivered to this
region in the rat. Tumors were not observed in the olfactory tissue of the rat.
Therefore, since effects observed in the rat are being extrapolated to the human, cells
from olfactory tissue are excluded in calculating average flux in the rat in the
Equation 4. For the human, both volumetric flow (2.5%, Subramaniam et al. [1998])
This document is a draft for review purposes only and does not constitute Agency policy.
5-126 DRAFT—DO NOT CITE OR QUOTE
-------
1 and surface area (-5%, Kelly et al. [2000]) for the olfactory region are relatively
2 small, so inclusion of this region is not likely to make a difference of much
3 significance in the calculation of average flux in the human. Since data on
4 formaldehyde flux delivered to the human olfactory region were not readily
5 available, the olfactory region was not excluded for the human. The average human
6 flux calculated here uses a working level classification for the activity profile where
7 an individual spent equal amounts of time in a day at resting and light and moderate
8 activity levels, corresponding to minute volumes of 7.5, 9, and 25 L/minute,
9 respectively. This resulted in the following ratio14:
10
11 /= 444[rat]/956.4[human] = 0.46 (5-6)
12
13 8. The airborne exposure concentrations dnuMAN corresponding to a given MLE and
14 upper bound lifetime added risk levels are the human BMDHUMAN and BMDLHuMAN,
15 respectively. These are shown in Figure 5-15. (The rather sudden increase by
16 -0.0015 in the upper confidence bound on risk for model 1 for exposure exceeding
17 -0.41 ppm could not be explained. This jump was verified by repeated calculations
18 that used different initial simulation conditions and convergence criteria.)
19
20 Extrapolation to the human lower respiratory tract
21 Next, the human lower respiratory tract is also considered to be potentially at risk.
22 Therefore, the above calculations of BMD and BMDL need to be augmented to include the
23 lower respiratory tract for humans. This calculation was facilitated by dosimetry
24 calculations of formaldehyde wall mass flux to various depths in the lung by using a single
25 path model. Refer to Overton et al. (2001) for details on their dosimetry modeling. The
26 calculations for including the lower respiratory tract in determining an overall BMD and
27 BMDL involved the following steps:
28
29 a. As given by Equation 5-3, calculate dHuMAN for various MLE risk levels. This gives
30 a dose-response relationship for lifetime risk of SCC in the human nose due to
31 continuous exposure to airborne formaldehyde.
32 b. Express this dose-response relationship in terms of average flux over the entire
33 human nasal lining.
34 c. Next, express this dose-response relationship, calculated here for the entire nose, as
35 risk per nasal cell versus average flux.
36
14 This is to be contrasted with a corresponding value of 0.71 in Schlosser et al. (2003) who used only resting
inspiratory rates.
This document is a draft for review purposes only and does not constitute Agency policy.
5-127 DRAFT—DO NOT CITE OR QUOTE
-------
,*
/
/ *
y rf
S x
/* •
/ m
/ >*'' /
/ * /
/ * *
jS f //
r ^ .--' /
''''
_jk *** *' .*
**' - * * ,~^^
• • • model 2 UB
— •*— model 1 mle
• * * __»-^^
" * ^~--~~~~~
0.4 0.41 0.42 0.43 0.44 0.45 0.46 0.47 0.48 0.49
human exposure cone (ppm)
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Figure 5-15. MLE and upper bound (UB) added risk of SCC in the human
nose for two BBDR models.
Note: Airborne exposure concentrations dnuMAN corresponding to a given MLE
and upper bound lifetime added risk levels are the human BMDnuMAN and
BMDLHuMAN, respectively.
d. Now, if the respiratory and transitional cell types in the human lung and nose are
equally susceptible to formaldehyde-induced cancer risk (as is also assumed in
Conolly et al. [2004]), then it appears reasonable to assume that MLE risk per cell at
a given value of formaldehyde flux is the same in the lung as in the nose.
e. The number of cells and the average flux in a given flux bin in the lung are known
(Overton et al., 2001). Thus, at a given air concentration, the MLE risk due to cells
in the various flux bins of the lung is obtained.
f. One important feature of Overton et al. (2001) was that their flux bins mapped
physically with lung depth. Therefore, in addition to extrapolating risk to the entire
human lung, it was also relatively easy to extend the risk calculation in e. above as a
function of airway generation in the lung (corresponding to different lung depths).
g. The MLE value risk to the lower respiratory tract (as determined above in steps
a.-e.) was a small fraction of risk to the upper respiratory tract. This is because of
high formaldehyde reactivity and solubility at the POE. Therefore, it sufficed to
assume that the relative increase in upper bound risk for the combined upper
respiratory tract + lower respiratory tract compared to that for only the upper
respiratory tract would be the same as the corresponding relative increase in the
This document is a draft for review purposes only and does not constitute Agency policy.
5-128 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
33
34
value of the MLE risk. The upper bound risk to the entire respiratory tract and
consequently the BMDL value corresponding to a given response were thus
determined.
These calculations indicated that including the risk of SCC in the lower respiratory
tract resulted in at most a 3% increase in the added risk at the lower end of the human
exposure range in Figure 5-15 (i.e., at 0.42 ppm) and about a 1.5% increase at the higher end
of the range in that plot. Therefore, including the lower respiratory tract did not appreciably
alter the human BMDs and BMDLs at the 0.5 and 1% response levels. This occurs because
of the steepness in the dose-response curve in this exposure range and much lower risk in
the lung at any exposure concentration.
Unit risks of SCC in the human respiratory tract extrapolated in this manner are reported
in Table 5-23.
Table 5-23. BMD modeling of unit risk of SCC in the human respiratory
tract
Extra risk level
0.005
0.010
Benchmark levels (ppm)
BMD
0.415-0.450
BMDL
0.410-0.435
0.430-0.460
Unit risk3
(per ppm)
1.2x 10~2
2.2 x 10~2
"Obtained from the mean of the two BMDLs.
Note: Findings are based on nasal tumors in rats and formaldehyde flux to tissue as dosimeter, using dose-
response curves for the F344 rat predicted by clonal growth modeling. Two chronic bioassays (Monticello
et al., 1996; Kerns et al., 1983) were combined, and control animals from the historical NTP inhalation
bioassays were added to the control animals in these bioassays.
5.3.4.2. Comparison with Other Benchmark Dose Modeling Efforts
The CUT assessment (Schlosser et al., 2003; CUT, 1999) also presented, as their less preferred
option, a benchmark approach on the data set obtained by combining the two chronic bioassays
with similar protocols (Monticello et al., 1996; Kerns et al., 1983) along with data from
94 animals that had not been previously examined. These authors used two measures of
response: tumor incidence and cell proliferation. In each case, they used two dosimeters: DPX
and formaldehyde flux to the nasal lining.
The extrapolation to human was carried out by using a hybrid CFD and pharmacokinetic
model. The CFD model (Kimbell et al., 2001a, b; Kepler et al., 1998; Subramaniam et al., 1998)
This document is a draft for review purposes only and does not constitute Agency policy.
5-129 DRAFT—DO NOT CITE OR QUOTE
-------
1 enabled calculation of site-specific flux in the nose of the rat, monkey, and human species for
2 inhaled formaldehyde concentrations, and the PBPK model (Conolly et al., 2000) linked this flux
3 to predicted DPX levels. The models were constructed for anatomically realistic representations
4 of a single individual in each species. The CFD and PBPK modeling and uncertainties in these
5 estimates have been reviewed in the Modeling the Toxicokinetics of Formaldehyde and DPX
6 section of chapter 3.
7
8 5.3.4.2.1. Benchmark dose using administered concentration.
9 Schlosser et al. (2003) fit multistage, Weibull, polynomial, and log-probit quantal models
10 to the tumor data and exercised the models (except the log-probit) with and without requiring
11 that the fits pass through the origin. The log-probit fit passed through the origin (see
12 Figure 5-16). A fifth degree polynomial was used in the multistage model. The best fit was
13 obtained with the polynomial and Weibull models for the tumor incidence data with a nonzero
14 intercept (threshold) on the dose axis. Fits passing through the origin did not pass the statistical
15 goodness-of-fit criteria (p > 0.01) for models other than the log-probit. The dose response near
16 the lowest dose was steep, with the LEDios and LEDois for the administered concentrations
17 nearly the same for each model, at least to one significant figure, and ranged from 3.8 to 6.4
18 ppm.
19
20 5.3.4.2.2. Benchmark dose derived with internal dose (flux and DPX) as dose metrics in
21 Schlosser et al (2003).
22 Schlosser et al. (2003) used CFD simulations (Kimbell et al., 2001a, b) of mass flux of
23 formaldehyde delivered across the nasal lining. The dose metric used by Schlosser et al. (2003)
24 for the extrapolation was the average flux of formaldehyde, expressed in pmol/cm2-minute, to
25 the entire surface of the airway lining. This excluded tissue lined by nonmucus-coated
26 squamous tissue, which was considered not to absorb formaldehyde.
27 In the CFD model, flux in any region is linearly related to the airborne exposure
28 concentration (i.e., flux = f x Ca;r [ppm], where f is a constant of proportionality and Cair is the
29 exposure concentration). The ratio of f (rat)/f (human) was determined as given by Equation 5-4.
30 This ratio was equal to 0.71 and differed from the value of 0.46 used in this document (as
31 presented in Equation 4-5) because Schlosser et al. (2003) used resting inspiratory rates.
32 In the next level of dosimetric complexity, Schlosser et al. (2003) used DPX as the
33 relevant dosimeter based on values predicted by PBPK models developed by Conolly et al.
This document is a draft for review purposes only and does not constitute Agency policy.
5-130 DRAFT—DO NOT CITE OR QUOTE
-------
Probit Model with 0.95 Confidence Level
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
0,8
•g 0.6
0.2
0
Probit
.BMD.Lj JBMD.
0 2
15:1902/042004
10
12
14
16
dose
Figure 5-16. Replot of log-probit fit of the combined Kerns et al. (1983) and
Monticello et al. (1996) data on tumor incidence showing BMCio and
BMCL10.
Source: Adapted from Schlosser et al. (2003).
(2000). This expressed the local dose as pmol of formaldehyde equivalents covalently bound to
DNA per unit volume of nasal tissue. Human CFD and PBPK models were exercised to
determine the airborne concentration of formaldehyde that yields average DPX levels equal to
those in the rat at the BMC. This airborne concentration was then the HEC. The human
benchmark extrapolations in Schlosser et al. (2003) using flux and DPX are shown in
Table 5-24, located at the end of Section 5.4.
The assumption in using DPX data was that lifetime exposure to the same DPX
concentration for a given duration each day leads to equivalent risk across species. Table 5-24
shows their human benchmark calculations for a continuous environmental exposure. These
were exposures that resulted in the same steady-state DPX concentrations as the weekly TWA
DPX values in rats at the rat benchmark exposure concentrations.
5.3.4.2.3. Cell proliferation in CUT benchmark modeling.
Schlosser et al. (2003) also used cell proliferation as representing the adverse response.
The BMDs and BMDLs calculated with these data did not differ appreciably from their other
benchmark estimates. The use of cell proliferation as an end point is considered to have the
This document is a draft for review purposes only and does not constitute Agency policy.
5-131 DRAFT—DO NOT CITE OR QUOTE
-------
1 advantage that it represents an early step contributing to carcinogenesis. In this document, a
2 BMD or BMDL is not calculated based solely on cell replication as a response. Instead, cell
3 replication rates are used as input to the clonal growth model and a benchmark dose based on a
4 fit to the tumor response using that model is considered a better choice since it integrates cell
5 replication along with other relevant data, such as the number of cells at risk and DPXs.
6
7 5.3.4.3. Kaplan-Meier Adjustment
8 In the simplest consideration of the impact of competing risks on the nasal tumor incidence,
9 tumor incidences were adjusted for early deaths according to Kaplan-Meier (KM) survival
10 estimates (KS Crump Group, 2001). This procedure allows for the possibility that some tumors
11 may otherwise have developed in the animals that died early due to other causes. All the animals
12 in the study were considered except those that were kept past termination of exposure. A
13 comparison of the adjusted incidence data is presented below in Table 5-25. While the
14 adjustments have been provided in Table 5-25, it needs to be noted that the data allow for a full
15 time-to-tumor analysis as presented below.
16
17 5.3.4.4. EPA Time-to-Tumor Statistical Modeling
18 Instead of using the KM adjustment, EPA has used the multistage Weibull time-to-tumor
19 model (Portier et al., 1986; Krewski et al., 1983) in other assessments (e.g., ethylene oxide,
20 1,3-butadiene, chloroprene). This is a dose-response model that includes the exact time of
21 observation of the tumors and therefore gives appropriate weight to the amount of time each
22 animal was on study without a tumor and acknowledges earlier tumor incidence with increasing
23 dose level. The data used in this analysis were obtained from the appendix in Conolly et al.
24 (2003) with one crucial modification. These data combined the nasal squamous carcinoma data
25 of Kerns et al. (1983) and Monticello et al. (1996) along with results from an additional
26 94 animals not previously examined in the Monticello et al. (1996) study. Animals in some
27 exposure groups were held up to 6 months following the 24-month exposure period; these
28 animals were deleted from the analysis for the following reason: there were no tumors among
29 these animals, and inclusion of them would have required estimating an equivalent TWA
30 exposure over the entire study period for these animals (40 in 2 ppm group, 39 in 6 ppm group,
31 3 in 15 ppm group), whereas the other animals would be represented by their actual exposure
32 concentrations.
33
34
This document is a draft for review purposes only and does not constitute Agency policy.
5-132 DRAFT—DO NOT CITE OR QUOTE
-------
11
<~5 ^
KO
S3
Table 5-24. Human benchmark extrapolations of nasal tumors in rats by using formaldehyde flux and DPX
§
Model
Weibullc'd
(with threshold)
Multistage
Weibull (time-
to-tumorf48
BBDR models
(see Table
5-23)
Source
Schlosser et
al. (2003)
EPA (this
assessment)
EPA (this
assessment)
Rat benchmark levels
(ppm)
ED
LED
ED
LED
1%
5.91
5.58
4.28
3.51
5%
6.12
5.94
5.93
5.52
10%
6.40
6.22
6.84
6.41
Extrapolated human benchmark levels (ppm)
Dose metric1"
Fluxe
DPXf
Fluxh
ED
LED
ED
LED
ED
LED
1%
0.75
0.71
0.76
0.71
0.35
0.30
5%
0.78
0.76
0.79
0.76
0.49
0.46
10%
0.82
0.79
0.84
0.81
0.57
0.53
See Table 5-23 and associated text
Unit risk" (ppm) l
1%
1.4xlO~2
1.4xlO~2
3.4xlO~2
5%
6.6 x 10~2
6.6 x 10~2
1.1 x KT1
10%
1.3 x 10"1
1.2 x 10"1
1.9 x 10"1
ql* = 2.2 x 10~2
atl%:2.2xlO~2
at 0.5%: 1.2xKT2
Note 1: Combined tumor incidence data from Kerns et al. (1983) and Monticello et al. (1996) were used for response.
"Slope of straight line extrapolation from the POD of the dose-response curve at the 1, 5, and 10% extra risk level.
bFlux: CFD modeling. DPX: CFD + PBPK modeling.
°p Value for Weibull model fit = 0.90. For the time-to-tumor modeling, goodness-of-fit/> value was not provided by software package; therefore, fit was judged
by comparing fitted curve to KM survival estimates (see Figure 5-19).
dFor Weibull model, Schlosser et al. (2003) obtained best fit with a positive intercept on dose axis. For multistage Weibull model, curves pass through origin.
eHuman benchmark levels extrapolated using flux were multiplied by fncHo-Rat/fHCHo-Human (= 0.71) for interspecies extrapolation and multiplied by (6/24) x (5/7)
to adjust for continuous exposure.
fHuman benchmark levels using DPX were continuous environmental exposures that would result in steady-state DPX levels in humans equal to the weekly TWA
DPX levels in rats at the rat BMCs for 6 hours/day and 5 days/week.
gP(d,t) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk)* tz]. q0, qi, q2, qs, q4 were all taken to be zero. q5 = 2.9 x 10~22, z = 8.1.
hHuman benchmark levels extrapolated using flux were multiplied by fncHo-Rat/fHCHo-Human =0.46 for interspecies extrapolation and multiplied by (6/24) x (5/7) to
adjust for continuous exposure (see Section 5.3.6.2).
O
o
-------
1
2
Table 5-25. Formaldehyde-induced rat tumor incidences
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Exposure level (ppm)
0.0
0.7
2.0
6.0
10.0
15.0
KM adjusted incidence
0.0
0.0
0.0
0.02
0.61
0.83
Observed tumor/
number at risk3
0/242
0/70
0/254
3/120a
22/3 6a
57/190a
aKM adjusted. Numbers not indicated by footnote were not amenable to KM
adjustment because there were no tumors; these numbers at risk reflect all animals
surviving 1 year on study.
Source: Monticello et al. (1996); Kerns et al. (1983).
Due to earlier tumor occurrence with increasing exposure level and increased mortality
with increasing exposure level, methods that can reflect the influence of competing risks and
intercurrent mortality on site-specific tumor incidence rates are preferred. EPA has generally
used the multistage Weibull model because it incorporates the time at which death with tumor
occurred, giving appropriate weight to the amount of time each animal was on study without a
tumor; the model has the following form: P(d) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk) x
(t - to)z], wherep(d) represents the lifetime risk (probability) of cancer at dose d (i.e., human
equivalent exposure in this case); parameters qt > 0, for i = 0, 1, ..., k; t is the time at which the
tumor was observed; and z is a parameter estimated in fitting the model, which characterizes the
change in response with age. The parameter to represents the time between when a potentially
fatal tumor becomes observable and when it causes death.
A further consideration is the distinction between tumor types as being either fatal or
incidental in order to adjust for competing risks. Incidental tumors are those tumors thought not
to have caused the death of an animal (such as those observed during interim or terminal
sacrifices), while fatal tumors are thought to have resulted in animal death. For these data, nasal
tumors observed with early deaths were considered to be fatal.
The dose-response analyses (see Figures 5-17, 5-18, 5-19) were conducted by using the
computer software program TOX_RISK, version 5.3 (ICF, Fairfax, VA), which is based on
Weibull models drawn from Krewski et al. (1983). Parameters were estimated by using the
This document is a draft for review purposes only and does not constitute Agency policy.
5-134 DRAFT—DO NOT CITE OR QUOTE
-------
1 method of maximum likelihood. Specific multistage Weibull models were selected for the
2 individual tumor types for each sex, based on the values of the log likelihoods according to the
3 strategy used by EPA (2002b). If twice the difference in log-likelihoods was less than a %2 with
4 degrees of freedom equal to the difference in the number of stages included in the models being
5 compared, the models were considered comparable, and the most parsimonious model (i.e., the
6 lowest-stage model) was selected contingent on visual fits of the data as follows. For incidental
7 tumors, plots of model fits compared with Hoel-Walburg estimates of cumulative incidence were
8 also examined for goodness of fit in the lower exposure region of the observed data (Gart et al.,
9 1986) (see Figure 5-18). For fatal tumors, plots of model fits were compared with KM estimates
10 of cumulative incidence. If a model with one more stage fitted the low-dose data better than the
11 most parsimonious model, then the model with one higher stage was selected.
12 Due to the sharp increase in responses between 6 and 10 ppm, no adequate fit was
13 achieved. Data for the highest dose were dropped in an effort to focus the fitting process for this
14 empirical model on the low-dose region. The model that then provided the best overall fit
15 included five stages but with coefficients for the lower stages estimated to be zero (see
16 Figures 5-17, 5-18, 5-19). The parameter t0 was estimated to be zero, consistent with rapidly
17 fatal tumors. On the other hand, an alternate run treating all tumors as incidental to the death of
18 the affected animals yielded BMCLs and BMCs within 10% of these estimates (see Figure 5-18);
19 thus, tumor context is not a sensitive consideration for these data.
20 For the same reasons as discussed in Section 5.3.3 (the concluding discussion of the
21 BBDR modeling), a linear low-dose extrapolation approach was used to estimate human
22 carcinogenic risk associated with formaldehyde exposure. PODs for estimating low-dose risk
23 were identified at doses at the lower end of the observed data, corresponding to 1% extra risk,
24 defined as the extra risk over the background tumor rate [P(d) - P(0)]/[l - P(0)]. PODs
25 corresponding to 10% extra risk are also provided to facilitate comparison with other chemicals.
26 Rat benchmark levels obtained by analysis of the tumor data are shown in Table 5-24. PODs
27 were converted to continuous human-equivalent exposure levels by multiplying by
28 (5 days/7 days) x (6 hours/24 hours), or 0.178, and by multiplying by the ratio of fluxes
29 developed in Section 5.3.6.1.3. The lifetime continuous inhalation unit risk for humans is
30 defined as the slope of the line from the lower 95% bound on the exposure at the POD,
31 calculated by dividing the BMR level (1%) by the corresponding BMCL0i. This 95% UCL
32 represents a plausible upper bound on the true risk.
33
34
This document is a draft for review purposes only and does not constitute Agency policy.
5-135 DRAFT—DO NOT CITE OR QUOTE
-------
0.9
0.8
0.7 -
o 0.4
0.3
0.2
0.1 -
o4—O-
Kaplan-Meier adjusted
incidence, from Schlosser
et al. (2003)
Multistage-Weibull estimate
at 104 weeks, omitting high
dose from fit
i
6 8 10
Formaldehyde exposure (ppm)
12
14
16
1
2
3
4
5
6
7
Figure 5-17. EPA Multistage Weibull modeling: nasal tumor dose response.
Note: Time-to-tumor modeling of Kerns et al. (1983) and Monticello et al. (1996)
data compared with incidences adjusted by using KM estimates evaluated at
104 weeks.
Source: Adapted from Schlosser et al. (2003).
This document is a draft for review purposes only and does not constitute Agency policy.
5-136 DRAFT—DO NOT CITE OR QUOTE
-------
15:3602/26/2004
Incidental Graph
hchoS.ttd - nasal squarnous cell carcinomas
Model: Five Stage Weib
— Dose (ppm)=0.7
Dose (ppm)=2
Dose (ppm)=6
Dose (ppm)=10
Hoel Walburg (6)
HoelWalburg(IO)
40
60 80
Time (wks)
100
120
1
2
3
4
5
6
7
Figure 5-18. Multistage Weibull model fit.
Note: Data of Kerns et al. (1983) and Monticello et al. (1996) compared with
Hoel-Walburg estimates of tumor incidences occurring at interim and terminal
sacrifices.
15:3602/26/2004
Fatal Graph
hchoS.ttd - nasal squarnous cell carcinomas
Model: Five Stage Weib
1
0,8
0.6
0.4
0,2
Dose (pprn)=0.7
Dose (ppm)=2
Dose (ppm)=6
Dose (ppm)=10
Kaplan Meier (6)
Kaplan Meier (10)
20
40
60
Time (wks)
80
100
9
10
11
12
Figure 5-19. Multistage Weibull model fit of tumor incidence data compared
with KM estimates of spontaneous tumor incidence.
Source: Developed from data reported in Kerns et al. (1983) and Monticello et al.
(1996).
This document is a draft for review purposes only and does not constitute Agency policy.
5-137 DRAFT—DO NOT CITE OR QUOTE
-------
1 The extrapolation to humans in terms of using formaldehyde flux to tissue as the dose
2 metric is shown in Table 5-24, where unit risk in terms of ql *, the statistical upper bound on the
3 coefficient, ql, of the term linear in dose in the multistage model, is also presented, ql* is
4 presented even though this is no longer done, as per current EPA practice (see Section 5.3.6 for
5 discussion).
6 These results are to be compared with the preferred benchmark estimates obtained in
7 Table 5-23 by using the results of biologically based models. In summary, the unit risks
8 obtained by various methods, including the results in Schlosser et al. (2003), fall within a rather
9 tight range. In particular, ql * was obtained to within a factor of two of other values even though
10 ql itself was zero. The general result may be noted here, that even in cases where ql is zero, the
11 upper bound ql* is linear with dose (Subramaniam et al., 2006; Guess et al., 1977). The large
12 difference between ql and ql * aptly reflects the large uncertainty in the low-dose response.
13 5.4. CONCLUSIONS FROM THE QUANTITATIVE ASSESSMENT OF CANCER
14 RISK FROM FORMALDEHYDE EXPOSURE BY INHALATION
15 5.4.1. Inhalation Unit Risk Estimates Based on Human Data
16 As described in Section 5.2, a (plausible upper bound) lifetime extra cancer unit risk of
17 1.1 x 10~2 per ppm (8.8 x 10~6 per |ig/m3) of continuous formaldehyde exposure was estimated
18 for NPC incidence using the log-linear modeling results (for NPC mortality from cumulative
19 exposure) from a high-quality occupational epidemiologic study in a life-table analysis to obtain
20 a POD and then applying linear low-dose extrapolation from the POD. Using similar methods
21 and data from the same study for Hodgkin lymphoma and leukemia mortality from cumulative
22 formaldehyde exposure, (plausible upper bound) lifetime extra cancer risk estimates of
23 1.7 x 10~2 per ppm (1.4 x 10~5 per |ig/m3) for Hodgkin lymphoma incidence and
24 5.7 x 10~2 per ppm (4.6 x 10~5 per |ig/m3) for leukemia incidence were derived. Sources of
25 uncertainty in these estimates are discussed in sections 5.2.2.4 and 5.2.3.4. For the incidence
26 risk for these three cancer types combined, a total (upper bound) cancer unit risk estimate of
27 8.1 x 1Q~2 per ppm (6.6 x 10~5 per |ig/m3) was obtained (see Section 5.2.4).
28
29 5.4.2. Inhalation Unit Risk Estimates Based on Rodent Data
30 As described in Section 5.3, the unit risk derived for SCC in the upper and lower
31 respiratory tract (combined) based on linear extrapolation from PODs from several plausible
32 models, including purely statistical modeling (nose only, quantal and time-to-tumor modeling)
33 and biologically based modeling (entire respiratory tract), resulted in a narrow range of
This document is a draft for review purposes only and does not constitute Agency policy.
5-138 DRAFT—DO NOT CITE OR QUOTE
-------
1 1.2 x 10 2 to 2.2 x 10 2 per ppm. Risk to the lower respiratory tract was numerically
2 insignificant compared to the nasal cancer risk.
3 5.4.3. Summary of Inhalation Unit Risk Estimates
4 The epidemiologic and rodent inhalation data indicate multiple sites of concern. Unit
5 risk estimates calculated separately from these data are presented in Table 5-26.
6 As can be seen in the summary table (see Table 5-26), the unit risk estimate based on
7 human data for NPC is in the range of the estimates calculated for respiratory tract cancer from
8 the rodent nasal cancer data. The unit risk estimate for Hodgkin lymphoma is also in the same
9 range, while the unit risk estimate for leukemia and the total cancer unit risk estimate are up to
10 fourfold higher.
11 As noted in EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), when
12 high-quality human data are available, they are generally preferred over laboratory animal data
13 for quantitative risk assessment. Thus, the preferred (plausible upper bound) unit risk estimate in
14 this assessment is the value of 8.1 x 10~2 per ppm (6.6 x 10~5 per |ig/m3) based on human data
15 for NPC, Hodgkin lymphoma, and leukemia.
16 As documented in Section 4.5, formaldehyde is a mutagenic carcinogen and the weight of
17 evidence supports the conclusion that formaldehyde carcinogenicity can be attributed, at least in
18 part, to a mutagenic MOA. Therefore, since there are no adequate chemical-specific data to
19 evaluate the susceptibilities of different life stages by the inhalation route of exposure, increased
20 early-life susceptibility should be assumed, and, if there is early-life exposure, the ADAFs
21 should be applied, in accordance with EPA's Supplemental Guidance for Assessing
22 Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b). See Section 5.4.4
23 below for more details on the application of the ADAFs.
24 The inhalation unit risk estimates presented above, which are calculated based on a linear
25 extrapolation from the POD (95% lower confidence bound on the EC), are expected to provide
26 upper bounds on the risk of cancer incidence. However, for certain applications, such as benefit-
27 cost analyses, estimates of "central tendency" for the risk below the POD are desired. Extra risk
28 estimates per ppm based on linear extrapolation from the EC (e.g., 0.005/ECoos) for the cancer
29 responses based on the human data are reported in Table 5-27. Note that these extrapolated risk
30 estimates are not central tendency estimates in any statistical sense because once risk is linearly
31 extrapolated below the EC, it is no longer a function of the original (Cox regression) model
32 which generated the ECs and the LECs. These estimates are dependent on the suitability of the
33 EC estimates as well as on the applicability of the linear low-dose extrapolation. The
34 assumption of low-dose linearity is supported by the mutagenicity of formaldehyde (see Section
This document is a draft for review purposes only and does not constitute Agency policy.
5-139 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
4.5.3). [If
Table 5-26. Summary of inhalation unit risk estimates
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Cancer type3
Dose metric
Unit risk estimate
(ppnT1)
Based on epidemiologic data
Nasopharyngeal
Hodgkin lymphoma
Leukemia
Total cancer riskb
Cumulative exposure
Cumulative exposure
Cumulative exposure
Cumulative exposure
0.011
0.017
0.057
0.081
Based on experimental animal data
SCC of the respiratory
tract
Local dose (flux) of
formaldehyde in
pmol/mm2-hour
0.011-0.022
aThe unit risk estimates are all for cancer incidence.
bThe total cancer unit risk estimate is an estimate of the upper bound on the sum of risk estimates calculated
for the 3 individual cancer types (nasopharyngeal cancer, Hodgkin lymphoma, and leukemia); it is not
the sum of the individual (upper bound) unit risk estimates (see Section 5.2.4).
these estimates were to be used for benefit-cost analyses or some other purpose, ADAFs should
be applied, as appropriate, in accordance with EPA's Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b), as discussed above
and in Section 5.4.4.]
5.4.4. Application of Age-Dependent Adjustment Factors (ADAFs)
When there is sufficient weight of evidence to conclude that a mutagenic MOA is
operative in a chemical's carcinogenicity and there are inadequate chemical-specific data to
assess age-specific susceptibility, as is the case for formaldehyde (by inhalation exposure; see
Section 5.4.3), EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens (U.S. EPA, 2005b) recommends the application of default ADAFs to
adjust for potential increased susceptibility from early-life exposure (see U.S. EPA [2005b] for
detailed information on the general application of these adjustment factors). In brief, EPA
(2005b) establishes ADAFs for three specific age groups: 10 (for <2 years), 3 (for 2 to
<16 years), and 1 (for 16 years and above). For risk assessments based on specific exposure
assessments, the 10-fold and threefold adjustments to the unit risk estimates are to be
This document is a draft for review purposes only and does not constitute Agency policy.
5-140 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Table 5-27. Extra risk estimates per ppm based on ECsa
Cancer type
nasopharyngeal cancer
Hodgkin lymphoma
leukemia
Total cancerd
BMRb
0.0005
0.0005
0.005
Outcome
mortality
incidence
mortality
incidence
mortality
incidence
mortality
incidence
EC (ppm)c
0.15
0.074
0.15
0.051
0.22
0.16
BMR/ECuMR
(per ppm)c
3.3 x 1Q~3
6.8 x 1Q~3
3.3 x 1Q~3
9.8 x 1Q~3
2.3 x 1Q~2
3.1 x 1Q~2
2.4 x I0~2d
4.7 x I0~2d
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
aBased on all person-years. Values based on exposed person-years only would be virtually identical.
bBMR = benchmark response, i.e., extra cancer risk level used to calculate the ECs and LECs.
°To convert ppm to ug/m3, multiply by 1,230; to convert ppm"1 to (ng/m3)"1, divide by 1,230.
dThe extra risk estimates per ppm for total cancer are not derived from ECs but rather from the calculations of
combined cancer risk at 0.1 ppm presented in Section 5.2.4 (see Table 5-20 for mortality and Table 5-21 for
incidence). The sums of the MLEs of risk from Tables 5-20 and 5-21, multiplied by 10 to convert from per 0.1
ppm to per ppm, correspond to the extra risk estimates per ppm calculated from the ECs (in that they are based on
MLEs and not bounds) but they are not equivalent to the sum of the EC-based values because those are calculated
at different ECs and the MLEs of risk are all calculated at a common exposure level of 0.1 ppm.
combined with age-specific exposure estimates when estimating cancer risks from early-life
(<16 years age) exposure. The ADAFs and their age groups may be revised over time. The
most current information on the application of ADAFs for cancer risk assessment can be found at
www.epa.gov/cancerguidelines.
For inhalation exposures, assuming ppm equivalence across age groups (i.e., equivalent
risk from equivalent exposure levels, independent of body size) and using the preferred unit risk
estimate of 6.6 x 10~5 per ug/m3 from Section 5.4.3, the calculation is fairly straightforward. For
example, the ADAF-adjusted total cancer unit risk estimate for a constant lifetime exposure level
is calculated as shown in Table 5-28.
This 70-year risk estimate of 1.1 x 10~4 for a constant exposure of 1 ug/m3 calculated in
Table 5-28 is equivalent to a lifetime unit risk of 1.1 x 10~4 per ug/m3 (0.13/ppm), adjusted for
early-life susceptibility, assuming a 70-year lifetime and constant exposure across age groups.
As mentioned above, for risk assessments based on specific exposure assessments, application of
Table 5-28. Total cancer risk from exposure to a constant formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
5-141 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
exposure level of 1 ug/m from ages 0-70 years
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Age group
0 to < 2 years
2 to < 16
years
> 16 years
ADAF
10
3
1
Unit risk
(per ug/m3)
6.6 x 10~5
6.6 x 10~5
6.6 x 10~5
Exposure
concentratio
n (ug/m3)
1
1
1
Duration
adjustment
2 years/70 years
14 years/70 years
54 years/70 years
Total risk =
Partial risk
1.9 x 1Q~5
4.0 x 10~5
5.1 x 10~5
1.1 x 1(T4
(Note that the partial risk for each age group is the product of the values in columns 2-5 [e.g.,
10 x (6.6 x l(T5) x i x 2/70 = 1.9 x io~5], and the total risk is the sum of the partial risks.)
the ADAFs is to be combined with age-specific exposure estimates when estimating cancer risks
from early-life (<16 years age) exposure. Further example calculations can be found in EPA's
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens
(U.S. EPA, 2005b).
In addition to the uncertainties discussed above for the inhalation unit risk estimate, there
are uncertainties in the application of ADAFs to adjust for potential increased early-life
susceptibility. The ADAFs are general default factors, and it is uncertain to what extent they
reflect increased early-life susceptibility for exposure to formaldehyde, if, in fact, early-life
susceptibility is increased as assumed. To some extent, the unit risk estimates for Hodgkin
lymphoma and leukemia already reflect some partial increased risk from early-life exposure
because the life-table programs include background rates for childhood cancers. However, the
impact of this partial increased risk is negligible compared to the effect of the ADAFs on the
final risk estimate. For example, eliminating the background rates up to age 16 from the life-
table programs decreases the lifetime extra risks at the PODs by about 0.5% for leukemia and
about 1.2% for Hodgkin lymphoma. The ADAFs, on the other hand, increased the lifetime unit
risk estimate by about 66%.
5.4.5. Conclusions: Cancer Inhalation Unit Risk Estimates
As presented in Section 5.4.3, the preferred (plausible upper bound) cancer unit risk
estimate for formaldehyde exposure in this assessment is the total cancer risk estimate of
8.1 x 10~2 per ppm (6.6 x 10~5 per ug/m3) based on (adult) human data for NPC, Hodgkin
lymphoma, and leukemia
This document is a draft for review purposes only and does not constitute Agency policy.
5-142 DRAFT—DO NOT CITE OR QUOTE
-------
1 In addition, as described in Section 5.4.4, because the weight of evidence supports the
2 conclusion that formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic
3 MOA and there are inadequate chemical-specific data to assess age-specific susceptibility,
4 increased early-life susceptibility should be assumed and, if there is early-life exposure, ADAFs
5 should be applied, in accordance with EPA's Supplemental Guidance for Assessing
6 Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b). Consequently,
7 applying the ADAFs to the preferred unit risk estimate to obtain a full lifetime unit risk
8 estimate yields
9
10 0.081/ppm x [(10 x 2 years/70 years) + (3 x 14/70) + (1 x 54/70)]
11 = 0.13/ppm = 1.1 x 10~4%g/m3)
12
13 Using the above full lifetime unit risk estimate of 0.13 per ppm, the lifetime chronic
14 exposure level of formaldehyde corresponding to an increased cancer risk of 10~6 can be
15 estimated as follows: (10~6)/(°-13/PPm) = 7.7 x 10"6 ppm = 0.008 ppb = 0.009 ug/m3. Similarly,
16 the lifetime chronic exposure level of formaldehyde corresponding to an increased cancer risk of
17 10~4 is 0.8 ppb, or 0.9 ug/m3. (Note that for less-than-lifetime exposures scenarios [or for
18 exposures that vary with age], the adult-based combined estimate of 0.081 per ppm should be
19 used, but if there is early-life exposure, the ADAFs should be applied in accordance with EPA's
20 Supplemental Guidance [see Section 5.4.4]).
This document is a draft for review purposes only and does not constitute Agency policy.
5-143 DRAFT—DO NOT CITE OR QUOTE
-------
1 6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND
2 DOSE-RESPONSE
3
4
5 6.1. SUMMARY OF HUMAN HAZARD POTENTIAL
6 6.1.1. Exposure
7 Formaldehyde (CH2O) occurs as a gas at room temperature. It is highly reactive and
8 dissolves readily in water. Formaldehyde is present in a wide variety of products including
9 plywood adhesives, abrasive materials, insulation, insecticides and embalming fluids (IPCS,
10 2002a; Agency for Toxic Substances and Disease Registry [ATSDR], 1999). The major sources
11 of anthropogenic emissions of formaldehyde are motor vehicle exhaust, power plants,
12 manufacturing plants that produce or use formaldehydes or substances that contain formaldehyde
13 (i.e., glues), petroleum refineries, coking operations, incineration, wood burning, and tobacco
14 smoke (INEG, 2003). Reported outdoor air concentrations of formaldehyde in urban and
15 suburban areas are near 3 ug/m3 (~ 3 ppb) (U.S. EPA, 2008) and indoor residential levels are
16 approximately 10 times higher (Health Canada and Environment Canada, 2001).
17 Limited United States data indicate that the upper end of the formaldehyde concentration
18 range in drinking water is approximately 10 |ig/L IPCS, 2002a). Formaldehyde is a natural
19 component of a variety of foodstuffs (IARC, 1995; IPCS, 1989). In addition, foods may also be
20 contaminated with formaldehyde as a result of fumigation (e.g., grain, seeds), cooking (as a
21 combustion product), and release from formaldehyde resin-based tableware (IARC, 1995).
22 Limited data measuring formaldehyde in food indicates that the concentration range is
23 <0.03-14 mg/kg Health Canada and Environment Canada, 2001). Daily intake of formaldehyde
24 has been estimated to be between 1.5 and 14 mg/day for an average adult (IPCS, 1989; Fishbein,
25 1992).
26
27 6.1.2. Absorption, Distribution, Metabolism, and Excretion
28 In water, less than 0.1% of formaldehyde exists unhydrated, with the majority reported to
29 be in the hydrated form, methylene glycol (CH2(OH)2) (Priha et al., 1996). Formaldehyde is a
30 reactive molecule that can react with both low molecular weight cellular components (e.g., GSH)
31 as well as high molecular weight components. It is also a well-known cross-linking agent.
32 Further, formaldehyde is a product of normal cellular metabolic processes. Endogenous
33 formaldehyde is a constituent of the one-carbon pool. It is thought that most endogenous
34 formaldehyde exists in a form that is reversibly bound to nucleophiles (Heck et al., 1990; 1982)
35 and that the formaldehyde hemithioacetal adduct formed with glutathione could account for
36 50-80% of the total formaldehyde normally present in cells (Heck et al., 1982).
This document is a draft for review purposes only and does not constitute Agency policy.
6-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 Inhaled formaldehyde is efficiently absorbed ("scrubbed") in the upper respiratory tract.
2 The fraction that is absorbed was determined to be approximately 97% in rats (Morgan et al.,
3 1986), and 85% and 90% respectively in computer simulations of one rhesus monkey and human
4 at rest (Kepler et al., 1998; Kimbell et al., 2001b). As the inspiratory rate increased, this fraction
5 decreased to about 70% during light exercise and to 58% during heavy exercise conditions in the
6 human (Kimbell et al. 2001). During heavy exercise, the absorption of formaldehyde in the first
7 six to eight generations of the tracheobronchial airways is estimated to be comparable to that in
8 the nasal region (Overton et al., 2001).
9 Airway geometry is an important determinant of inhaled-formaldehyde dosimetry in the
10 respiratory tract. There are large differences across species in the anatomy of the upper
11 respiratory tract and in airflow patterns. Using computer simulation, the regional uptake patterns
12 of formaldehyde in the upper respiratory tract are observed to be spatially nonhomogeneous and
13 to exhibit strong species differences. Airflow patterns are also significantly different as
14 breathing patterns and activity profiles change, depending on whether breathing is oral or nasal.
15 The overall information on the disposition of inhaled formaldehyde comes from many
16 studies using different experimental methods including: [14C] radiolabeling, gas
17 chromatography-mass spectroscopy (GC-MS), dual isotope labeling (3H, 14C) and high-
18 performance liquid chromatography (HPLC) studies. In a study of rats following exposure to
19 radiolabeled formaldehyde, the radioactivity was very high in the nasal mucosa but was also
20 extensively distributed to various tissues including the bone marrow (Heck et al., 1983). The
21 elevated 14C in various tissues was thought unlikely to be due to free formaldehyde but instead to
22 arise from either rapid metabolic incorporation or formation of covalent adducts or incorporation
23 via carboxylation reactions of the 14CC>2 formed during metabolism (Heck et al., 1983;
24 Casanova-Schmitz et al., 1984). Studies using the GC-MS method indicate that exposure to
25 formaldehyde over a wide range of exposure concentrations and durations does not result in
26 elevated levels in blood, above those of endogenous formaldehyde levels in rats, rhesus monkeys
27 and humans (Heck et al., 1985; Casanova et al., 1998). These GC-MS measurements are
28 consistent with the conclusions that formaldehyde does not appreciably reach the blood, is
29 rapidly metabolized, interacts with macromolecules when it escapes metabolism, or is otherwise
30 undetected.
31 In further studies on the disposition of inhaled formaldehyde, Casanova-Schmitz et al.
32 (1984) and Casanova-Schmitz and Heck (1983) used dual-isotope labeling of inhaled
33 formaldehyde as an approach to distinguish between formaldehyde adduct formation and
34 metabolic incorporation. These were followed by more sensitive experiments using HPLC
35 measurements in rats and rhesus monkeys exposed to radiolabeled formaldehyde (Casanova et
This document is a draft for review purposes only and does not constitute Agency policy.
6-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 al. 1989, 1991). Results from this sets of experiments found that labeling in the nasal mucosa
2 was due to both covalent binding and metabolic incorporation and labeling of bone marrow
3 macromolecules was found to be entirely due to metabolic incorporation. Overall, Heck,
4 Casanova-Schmitz, and their coworkers interpreted the results of these experiments to indicate
5 that inhaled formaldehyde does not reach distant sites (beyond the portal of entry) at detectable
6 levels.
7 Formaldehyde is primarily metabolized by glutathione-dependent formaldehyde
8 dehydrogenase. In humans this enzyme is referred to using the protein code of ADH3. The
9 major factor in the disposition of formaldehyde is metabolic clearance by oxidation to formate,
10 which is either further metabolized to CO2 and water, incorporated into the one-carbon pool,
11 and/or eliminated in the urine as a sodium salt.
12 In radiolabeling studies, Heck et al. (1983) determined that the relative contributions of
13 various excretion pathways in F344 rats following inhalation exposure to formaldehyde were
14 independent of exposure concentration. Nearly 40% of inhaled [14C] -formaldehyde appeared to
15 be eliminated via expiration, presumably as CO2, while about 17% and 5% was eliminated in the
16 urine and feces, respectively. Nearly 40% of inhaled [14C] -formaldehyde remained in the
17 carcass, presumably due to metabolic incorporation. For exposure via the oral route, absorption
18 of [14C] -formaldehyde (7 mg/kg) in rats resulted in 40% exhaled (as 14CO2), 10% excreted in
19 urine, 1% excreted in feces, and much of the remaining 49% retained within the carcass,
20 presumably due to metabolic incorporation (IARC, 1995; Buss et al., 1964).
21 Several human and animal studies have reported formaldehyde in exhaled breath (see
22 Section 3.6.2). However, limitations of analytical techniques employed for breath analysis can
23 only tentatively identify formaldehyde (Spanel and Smith, 2008, Wehinger et al., 2007). A
24 recent study has illustrated that the use of proton transfer reaction in SIFT-MS may result in false
25 positive results for formaldehyde as the characteristic analytical product ion for formaldehyde is
26 also produced from methanol and ethanol (Spanel and Smith, 2008). Therefore, ethanol and
27 methanol in exhaled breath will contribute to the analytical product tentatively identified as
28 formaldehyde in the existing literature. Additionally, some studies do not have appropriate
29 control samples to define formaldehyde levels for inhaled air prior to breath analysis. Therefore,
30 the two major limitations of available studies of formaldehyde levels in human breath include the
31 potential for false positives for formaldehyde from the primary analytical technique for breath
32 analysis and the need for concurrent room air controls.
33 Although several studies of healthy subjects report levels of formaldehyde between the
34 detection limit and 12 ppb (Wang et al., 2008; Cap et al., 2008 and Kushch et al., 2008), there
35 was no adjustment for an artifact in the analytical method that makes it impossible to distinguish
This document is a draft for review purposes only and does not constitute Agency policy.
6-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 between formaldehyde and reaction products for 1% of exhaled methanol and ethanol which are
2 detected at the same mass to charge ratio as formaldehyde in these analytical techniques (Spanel
3 and Smith, 2008). To date, there is no published study of formaldehyde in exhaled breath which
4 makes this adjustment for reporting formaldehyde levels. Therefore, reports of formaldehyde in
5 exhaled breath should be carefully interpreted as the mass reported as formaldehyde—is only
6 tentatively identified as formaldehyde. A review of the data where methanol and ethanol levels
7 are also provided, indicate that levels of formaldehyde (tentatively identified as m/z = 31) may
8 reflect a significant contribution from reaction products of methanol and ethanol (see Section
9 3.6.2). In summary, there are insufficient data at this time to confidently establish a
10 concentration of formaldehyde in exhaled breath that can be attributed to endogenous sources.
11 This assessment identifies a critical research need for further studies on the measurement of
12 exhaled formaldehyde.
13
14 6.1.3. Noncancer Health Effects in Humans and Laboratory Animals
15 A wide variety of human clinical and observational epidemiology and animal studies
16 provide evidence for health effects in response to formaldehyde exposure. Some of these health
17 effects are commonly noted at the portal of entry, as expected for exposure to a reactive gas. In
18 addition, effects on the nervous and reproductive systems, developmental effects, and
19 immunomodulation have been reported. The overall weight of evidence (WOE) of human and
20 animal studies for the hazard potential of formaldehyde is discussed below, along with
21 information on plausible modes of action (MOAs).
22
23 6.1.3.1. Sensory Irritation
24 Formaldehyde, a chemical irritant, binds to protein receptors of the trigeminal nerve,
25 triggering a burning and painful sensation in humans. This process is distinct from taste and
26 smell (Nielsen 1991; Cometto-Muniz and Cain, 1992). The trigeminal nerve, which has
27 three branches (ophthalmic, maxillary and mandibular), not only acts as an afferent nerve
28 relaying these sensations to the central nervous system, but also has efferent nerve activity
29 (Stedman's Medical Dictionary: Meggs, 1993). Stimulation of the trigeminal nerve may result
30 in reflex responses including lacrimation, coughing, and sneezing. Both the reflex responses as
31 well as sensations such as burning, pain, and itching of the eyes, nose, and throat are considered
32 adverse.
33 Formaldehyde-induced eye, nose, and throat irritation has been well documented in a
34 wide range of epidemiologic studies. Common effects of chemically-induced sensory irritation
35 include lacrimation, burning of the eyes and nose, rhinitis, burning of the throat, and cough
This document is a draft for review purposes only and does not constitute Agency policy.
6-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 (Feron et al., 2001). Studies examining these endpoints were either controlled chamber studies
2 with a defined population (e.g., healthy volunteers or sensitive individuals), worker/student
3 studies, or general population studies (e.g., residential). Chamber studies, by design, are acute
4 studies, although some researchers have investigated the outcomes after repeated exposures.
5 Occupational, student, and residential exposures are generally of longer duration, although there
6 is variability in exposure level and duration among subjects. The endpoints for assessing
7 irritation include self-reporting of symptoms (e.g., pain, burning, itching) and objective measures
8 of irritation (e.g., eye-blink counts, lacrimation).
9 Eye irritation is the most sensitive of reported effects in human studies. Two different
10 short-term chamber studies provide similar 10% BMDLs for eye irritation of 560 ppb and
11 240 ppb for 3 and 5 hour exposures, respectively (Kulle, 1993; Andersen and Molhave, 1983,
12 modeled by Arts et al., 2006b). Various occupational studies have noted increased eye irritation
13 for average exposures ranging from 180 ppb to 690 ppb (Horvath et al., 1988, Alexandersson
14 and Hedenstiera, 1998; Holmstrom and Wilhelmsson, 1988). The results of residential studies,
15 where in-home formaldehyde levels are used to document exposure, indicate eye irritation may
16 increase with increasing exposure from 70 to 200 ppb for these chronic exposure scenarios
17 (Ritchie and Lehnen, 1987, Hanrahan et al., 1984; Liu et al., 1991.)
18 When a rodent is exposed to an irritant, the inhaled dose and pattern of deposition can be
19 profoundly affected by reflex bradypnea, a protective reflex observed in rodents but not in
20 humans. Reflex bradypnea is manifest as markedly decreased activity or prostration, reduced
21 metabolism, hypothermia (as much as 5°C), significantly reduced respiratory rate and minute
22 volume, and altered blood and brain chemistry. Reflex bradypnea can occur when the trigeminal
23 nerve is exposed to a sufficient concentration of an irritant, such as formaldehyde. Because of
24 their small size, rodents are able to rapidly lower their metabolism and body temperature and
25 therefore their oxygen demand. The consequence is that their inhaled dose of an irritating
26 chemical is dramatically lowered. Reflex bradypnea is quantified as the RDso, which is the
27 concentration of a chemical that results in a 50% decrease in respiratory rate (see Tables 4-7 and
28 4-8). After the irritant exposure is removed, it can take up to two hours for rodents to fully
29 recover from the effects of reflex bradypnea. Even though humans do not exhibit reflex
30 bradypnea, involvement of trigeminal nerve stimulation, which is the mechanism for reflex
31 bradypnea in rodents, may be relevant to MO As for formaldehyde in other species, such as
32 primates and humans. For example, trigeminal nerve stimulation has been associated with
33 sensory irritation in humans, highlighting the relevance of this effect.
34
This document is a draft for review purposes only and does not constitute Agency policy.
6-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 6.1.3.2. Respiratory Tract Pathology
2 Formaldehyde-induced respiratory tract pathology includes inflammation, rhinitis, goblet
3 cell hyperplasia, metaplastic changes, squamous cell hyperplasia, and impaired mucociliary
4 transport. Formaldehyde binding to the trigeminal nerve triggers the release of neurogenic
5 mediators of inflammation resulting in tissue edema, lacrimation, mucus production, and
6 leukocyte infiltration. Therefore, observed pathological changes may be directly related to
7 neurogenic inflammation from activation of the trigeminal nerve or result, at least in part, from
8 formaldehyde-induced cell damage to the mucosal tissue. A series of exposures has also been
9 positively associated with reduced mucociliary clearance, and the induction of histopathologic
10 lesions in the nose in both human and animal studies assessing formaldehyde-induced changes in
11 the nasal mucosa suggest that these changes may be, at least in part, a protective or adaptive
12 response and that increased mucus flow and metaplastic changes would progress in relation to
13 the concentration and duration of exposure protecting the underlying tissue (Swenberg et al.,
14 1983).
15 In rodent studies, formaldehyde-induced histopathological lesions ranging from
16 inflammation to ulceration, necrosis, and metaplasia have been frequently reported in nasal
17 turbinates, maxilloturbinates, and in goblet and microvilli cells (e.g., Bhalla et al., 1991;
18 Monteiro-Riviere and Popp, 1986; Cassee and Feron, 1994; lonescu et al., 1978; Schreibner
19 et al.,1979; Monticello et al., 1989). These effects were observed after a variety of exposure
20 scenarios (e.g., 10 ppm for 4 hrs (Bhalla et al., 1991), 0.5 or 2 ppm for 6 hrs/day for 1 or 4 days
21 and 6 or 15 ppm for 6 hrs/day for 1 or 2 days (Monteiro-Riviere and Popp, 1986), 3.6 ppm
22 intermittently for 3 days (Cassee and Feron, 1994), 3% aerosols of formaldehyde for 3 hrs/day
23 for 50 days (lonescu et al., 1978)). The progressive pathology of the nasal passages from
24 formaldehyde inhalation exposure is dependent on increasing concentration and duration of
25 exposure, as well as from proximal to distal regions of the nasal cavity. For example, some
26 lesions may be transient (e.g., low-exposure cell proliferation), while others may have a
27 maximum response and be irreversible (e.g., allergic rhinitis). The nasal epithelium responds
28 with both adaptive and adverse epithelial changes. As respiratory epithelium transitions to
29 squamous metaplasia, the effective tissue dose of formaldehyde increases posterior to these
30 lesions. As epithelial barriers degrade (e.g., squamous metaplasia, keratinization), formaldehyde
31 penetrates more deeply into the nasal passages. Therefore, the relationship between
32 concentration and duration of exposure and health outcomes has been difficult to define and, in
33 fact, may be different for various health effects. Formaldehyde-related histopathological lesions
34 of the nasal mucosa have been observed at concentrations as low as 2 ppm for chronic exposure
This document is a draft for review purposes only and does not constitute Agency policy.
6-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 and after a duration as short as 6 hrs at higher concentrations (e.g., 6 ppm) (see Table 4-32,
2 Table 4-38).
3 Similar pathology has been reported for workers exposed to formaldehyde, including loss
4 of cilia, goblet cell hyperplasia, and cuboidal and squamous cell metaplasia and dysplasia, and
5 these pathology scores were significantly elevated in workers over controls (Holmstrom and
6 Wilhelmsson, 1988; Edling et al., 1988; and Boysen et al., 1990). Holmstrom and Wilhelmsson
7 (1988) reported associations between the mean daily exposure of 240 ppb (8hr TWA) and these
8 changes. Edling et al. (1988) reported that workers experienced a range of exposures
9 (80-900 ppb), with peak exposures of 4,000 ppb. Boysen et al. (1990) provided a range of
10 estimated exposures from 500 ppb to more than 2,000 ppb for workers with elevated mean
11 pathology scores. One controlled chamber study indicated formaldehyde-induced inflammatory
12 changes which persisted for 18 hours in adults exposed at 400 ppb for only 2 hours (Pazdrak
13 etal., 1993).
14 Short-term formaldehyde exposure also impairs the function of the mucociliary apparatus
15 which is a critical defensive barrier for the upper respiratory tract. Numerous laboratory animal
16 studies have reported impaired mucociliary clearance activity associated with formaldehyde
17 exposures as low as 500 ppb (see Table 4-10). Low-concentration or short-term exposures first
18 lead to an increased rate of ciliary beat, followed by impaired mucus flow, with slowed rate of
19 ciliary beat and eventual mucostasis (lack of mucus flow) and ciliastasis (lack of ciliary beat)
20 occurring at higher doses or longer exposure times. These effects have been shown to be both
21 concentration- and duration-dependent and to occur within 15 minutes after the initial exposure.
22 Morgan et al. (1983c) suggested that the initial stimulation of ciliary activity may be a defensive
23 response to the irritant gas, at which time some penetration of formaldehyde to the underlying
24 epithelial cells may occur. Later effects of mucostasis and ciliastasis may occur as a result of
25 formaldehyde-induced glycoprotein cross-links, creating a rigid mucus that effectively stops
26 mucus flow.
27 Formaldehyde-induced cell proliferation has been demonstrated in nasal epithelium in
28 animal studies after a range of exposure conditions (e.g., Swenberg et al., 1986; Cassee and
29 Feron, 1994; Reuzel et al., 1990; Woutersen et al., 1987) (see Table 4-43). Formaldehyde-
30 induced histopathology and mitogenesis may occur as a direct effect of exposure (Tyihak et al.,
31 2001) or as a secondary effect resulting from adaptive responses and/or compensatory tissue
32 repair that can occur after formaldehyde exposure (Swenberg, 1983). In a study of Rhesus
33 monkeys Monticello et al. (1996) noted that increased cell proliferation was seen in locations
34 with minimal histological changes in the respiratory tract indicating that cell proliferation may
35 be a more sensitive predictor of more severe health effects due to formaldehyde exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
6-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 Cellular proliferative responses may initiate lesion formation. A number of studies illustrate that
2 the duration of repeated exposures may be an important determinant of cell proliferation rates
3 (Wilmer et al., 1987; Swenberg et al., 1986). Reduced mucociliary clearance and the induction
4 of histopathologic lesions in the nose effects have been noted in human formaldehyde studies.
5 Histopathological lesions and biochemical changes have been reported in the lung
6 following formaldehyde inhalation exposure in experimental animal studies (Kamata et al.,
7 1996a; lonescu et al., 1978) following high exposure levels (128.4 or 294.5 ppm formaldehyde).
8
9
10 6.1.3.3. Effects on Pulmonary Function
11 The potential of formaldehyde exposure to cause pulmonary functional deficits in
12 humans has been examined on several time scales. The epidemiologic literature includes studies
13 of acute exposures among naively exposed anatomy graduate students (Kriebel et al., 1993;
14 2001), anatomy graduate students with several weeks of episodic exposure (Kriebel et al., 1993),
15 and post-shift versus pre-shift worker pulmonary function among those with regular
16 occupational exposure (Malaka and Kodama, 1990; Herbert et al., 1994; Alexandersson et al.,
17 1982; Alexandersson and Hedenstierna, 1989). Depending on whether the exposures are naive
18 or not, the epidemiologic studies that assessed the pulmonary effects after acute exposures to
19 formaldehyde are assessing different biological responses, namely, the acute effect alone or the
20 acute effect(s) in people who may have already been sensitized to different and unknown
21 degrees.
22 The observed effects in the previously unexposed anatomy students provide additional
23 information on acute exposures in two naive populations (Kriebel et al., 1993; 2001), as well as
24 insight into the possible intermediate stages of sensitization (Kriebel et al., 1993). Kriebel and
25 colleagues (1993) examined the prelaboratory and postlaboratory peak expiratory flow (PEF) in
26 students attending anatomy classes once a week. They found the strongest pulmonary response
27 when examining the average cross-laboratory decrement in peak expiratory flow in the first
28 2 weeks of the study when formaldehyde concentrations collected in the breathing zones had a
29 geometric average concentration of 0.73 ppm. Overall, the students exhibited a 2% decrement in
30 PEF, while the students with any history of asthma showed a 7.3% decrement in PEF. These
31 findings of acute decreases in PEF following students' initial formaldehyde exposure were
32 corroborated by the Kriebel et al. (2001) study, using a similar study design applied to a separate
33 class of anatomy students. Similar findings have been reported for low-level residential
34 formaldehyde exposure including decreased peak expiratory flow rates (PEFRs) (Krzyzanowski
35 et al., 1990). Workers chronically exposed to formaldehyde have exhibited signs of reduced
This document is a draft for review purposes only and does not constitute Agency policy.
6-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 lung function consistent with bronchial constriction, inflammation, or chronic obstructive lung
2 disease. Lung function deficits have been reported both in preshift versus postshift
3 measurements and_as a result of chronic exposures (Malaka and Kodama, 1990; Herbert et al.,
4 1994; Pourmahabadian et al., 2006, Alexandersson et al., 1982; Alexandersson and Hedenatiena
5 1989). Decreases in spirometric values, including vital capacity (VC), forced expiratory volume
6 (FEV), forced vital capacity (FVC) and FEV/FVC have been reported in humans. Chronic
7 studies also reported increased respiratory symptoms such as cough, increased phlegm, asthma,
8 chest tightness and chest colds in exposed workers (Malaka et al., 1990; Herbert et al., 1994;
9 Pourmahabadian et al., 2006, Alexandersson et al., 1982; Alexandersson and Hedenatiena 1989).
10 Similar findings have been reported following low-level residential formaldehyde exposure
11 including decreased PEFRs (Krzyzanowski et al., 1990).
12 Worker exposures associated with cross-shift differences in spirometric values are
13 consistent with formaldehyde-induced sensory irritation. Concordance has also been reported
14 between subjective irritant response and measured changes in pulmonary function further
15 supporting the possibility that cross-shift and short-term evidence of bronchial constriction may
16 be a reflexive response to sensory irritation.
17 A well-conducted residential epidemiology study by Krzyzanowski et al. (1990) was
18 considered to be the strongest among the candidate studies on the adverse pulmonary function
19 effects of formaldehyde for the purposes of deriving an RfC.
20
21 6.1.3.4. Asthmatic Responses and Increased Atopic Symptoms
22 The health effects of respiratory function, asthma and increased atopic response, have
23 been shown to be clinically related. For example, asthma affects pulmonary function and may be
24 triggered by an allergic response. These and other data suggest that there may be mechanistic
25 links between these two health effects. Formaldehyde-induced sensitization (see Section 4.2.1.5)
26 may enhance the asthmatic response or may enhance an individual's response to an allergen (see
27 Section 4.4). In both cases, sensitization results in phenotypic switching—or an individual
28 exhibiting clinical symptoms of a predisposition to asthma or atopy. Because of the connection
29 between the two endpoints, they are considered together herein.
30 Several cross-sectional studies have described a positive association between
31 formaldehyde concentration and asthma prevalence. A study on risk factors for the initial
32 physician diagnosis of asthma has shown concentration-dependent associations between
33 formaldehyde exposure and asthma (Rumchev et al., 2002). In a categorical analysis, Rumchev
34 et al. (2002) observed statistically significant effects above in-home formaldehyde
35 concentrations of 60 ug/m3, with increased but nonsignificant effects at 50-59 ug/m3 that were
This document is a draft for review purposes only and does not constitute Agency policy.
6-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 consistent with a concentration-response relationship. No effect was apparent at concentrations
2 in the next lower interval between 30-49 ug/m3. Garrett et al. (1999 a,b) reported a borderline
3 statistically significant association between bedroom formaldehyde concentrations and an
4 increased risk of atopy. The authors computed a respiratory symptom score for each child based
5 on the frequency of each of eight respiratory symptoms and this score was substantially and
6 statistically significantly higher among the asthmatic children compared to nonasthmatic
7 children. Health effects were reported at formaldehyde concentrations greater than 50 ug/m3 but
8 the lowest formaldehyde concentration interval at which health effects were observed was
9 20-50 ug/m3. The findings of Garrett et al. (1999 a,b) are supported by the results of a chamber
10 study reported by Casset et al. (2006) of 19 sensitized adult asthmatics exposed to formaldehyde
11 at a concentration of 100 ug/m3 for 30 minutes. Casset and colleagues observed an increased
12 bronchial responsiveness to mite allergen exposure (p = 0.05) and noted the provocative dose
13 (PD20 for FEV1) for mite allergen was 34.3 ng after formaldehyde exposure and 45.4 ng after
14 air exposure. However, in study by Ezratty et al. (2007) exposure to 500 ug/m3 formaldehyde
15 did not affect an allergen-induced increase in responsiveness to methacholine (p = 0.42) and
16 there was no formaldehyde-associated effect on the airway inflammatory response.
17 These observed health effects in humans are similar to the outcome of studies in
18 laboratory animals that show that formaldehyde can exacerbate existing immunogenic
19 hypersensitivity to known allergens (Sadakane et al., 2002; Tarkowski and Gorski, 1995; Riedel
20 etal., 1996). While potentiation varied based on sensitization protocols and formaldehyde
21 exposure regimens, the results support the finding that formaldehyde exposure can aggravate a
22 Type-I hypersensitivity response and may do so via a neurogenically initiated response.
23 Formaldehyde itself does not function as an allergen recognized by the immune system (Lee
24 etal., 1984) and does not appear to trigger formation of formaldehyde-specific IgE. Although
25 formaldehyde exposure has been reported to alter cytokine levels and immunoglobulins in some
26 experimental systems (Fujimaki et al., 2004a; Ohtsuka et al., 2003), these effects do not support
27 an immunogenically mediated type-I hypersensitivity. In studies in which either egg protein
28 (ovalbumin, OVA)-sensitized or dust mite (DerF)-sensitized animals were exposed to
29 formaldehyde, OVA-specific and DerF-specific antibody production was increased over
30 sensitization alone, suggesting that formaldehyde may potentiate sensitization responses (Riedel
31 et al., 1996; Sadakane et al., 2002). Formaldehyde-induced sensitivity responses may be
32 neurogenic in origin based on findings that neurogenic factors such as nerve growth factor
33 (NGF) and substance P were associated with formaldehyde exposure in sensitization protocols
34 (Fujimaki et al., 2004b).
This document is a draft for review purposes only and does not constitute Agency policy.
6-10 DRAFT—DO NOT CITE OR QUOTE
-------
1 6.1.,3.,5.Effects on the Immune System
2 Formaldehyde-induced systemic immunomodulation in laboratory animals has been
3 documented in the literature (Leach et al., 1983; Dean et al. 1984; Adams et al., 1987). A
4 number of studies have evaluated the ability of formaldehyde to induce systemic immunotoxic
5 effects in humans (Ohtani et al., 2004a, b; Erdei et al., 2003; Thrasher et al., 1990, 1987; Pross et
6 al., 1987). Some studies have reported altered innate immune responses associated with
7 formaldehyde exposure (Erdei et al., 2003), while others have noted adaptive immune response
8 suppression associated with formaldehyde exposure (Thrasher et al., 1990, 1987) and changes
9 associated with alterations to a predominant T-lymphocyte helper 2 (Th2) pattern (Ohtani et al.,
10 2004a, b). In contrast, Pross et al. (1987) did not observe formaldehyde-associated changes in
11 systemic immune function.
12 Diverse studies have investigated the possibility that formaldehyde exposure leads to
13 increased respiratory tract infections (Lyapina et al., 2004; Krzyzanowski et al., 1990; Holness
14 and Nethercott, 1989). Lyapina et al. (2004) reported increased respiratory tract infections and
15 decreased neutrophil respiratory burst activity (NRBA) in formaldehyde-exposed workers (at
16 722 ppb TWA). Incidences of doctor-diagnosed chronic bronchitis were more prevalent in
17 children under age 15 living in homes with higher formaldehyde (>60 ppb) readings in the
18 kitchen (p < 0.001) (Krzyzanowski et al., 1990). Holness and Nethercott (1989) also report
19 increased chronic bronchitis in formaldehyde-exposed funeral workers (380 ppb average
20 exposure).
21
22 6.l.3.6.Neurological Effects
23 Formaldehyde exposure via inhalation has been shown to adversely impact nervous
24 system function in laboratory animals and humans, although human data for formaldehyde-
25 induced neurological effects are limited. Studies in formaldehyde-exposed histology technicians
26 provide evidence of neurological impairment, including lack of concentration, impaired memory,
27 disturbed sleep, impaired balance, variations in mood and irritability. These effects were
28 significantly correlated with increasing duration of exposure to formaldehyde, but the findings
29 are not conclusive due to confounding by concomitant exposures to other neurotoxic solvents
30 (Kilburn et al., 1985, 1987). In a prospective study, Weisskopf et al. (2009) found a strong
31 association between duration of formaldehyde exposure and death from amyotrophic lateral
32 sclerosis (ALS), but information regarding exposure levels was not available. Short-term studies
33 with controlled exposure to humans (chamber studies) also provide limited support for changes
34 in cognitive function immediately following a single, controlled formaldehyde exposure (Bach
35 et al., 1990; Lang et al. 2008).
This document is a draft for review purposes only and does not constitute Agency policy.
6-11 DRAFT—DO NOT CITE OR QUOTE
-------
1 Available animal data provide substantial evidence of behavioral changes in animals
2 following single or short-term repeated inhalation exposures to relatively low levels of
3 formaldehyde. Among the animal studies, none of the available studies examined effects on
4 nervous system function following chronic formaldehyde inhalation, however.
5 Reported perturbations in nervous system function following formaldehyde exposure in
6 animal studies include reductions in motor activity, lack of habituation, impairment in
7 acquisition of a new learning task, deficits in retention of a previously learned task, increases in
8 corticosterone levels, sensitization to cocaine-induced locomotor activity, and enhanced fear
9 conditioning using an olfactory conditioned stimulus (CS) (see Table 4-57). Behavioral effects
10 have been seen in multiple laboratories and in studies conducted by different investigators using
11 a variety of testing paradigms. Many of these effects were observed at acute exposure levels at
12 or below 1.0 ppm, and some persisted days to weeks after termination of exposure.
13 More limited data indicate possible effects on the development of the nervous system,
14 including changes in brain structure and in the behavior of offspring (see Table 4-57). Similarly,
15 there is very little information regarding the mechanism by which effects on the nervous system
16 might be produced. The data regarding behavioral sensitization provide some support for a
17 stress-related mechanism for those specific findings, but the applicability of this mechanism to
18 the behavioral changes seen in the other studies, including the learning deficits and
19 developmental findings, has not been evaluated. Although there are data supporting stimulation
20 of the trigeminal nerve by formaldehyde (and documenting the relevance of this interaction to
21 the sensory irritation caused by formaldehyde), there are no data supporting a causal relationship
22 between irritant properties of formaldehyde and the behavioral and neurodevelopmental effects
23 in humans that occur following formaldehyde exposure. In summary, none of the available data
24 provide sufficient information to allow a determination of the mode of action for effects of
25 formaldehyde on the adult or developing nervous system.
26
27 6.1.3.7.Reproductive and Developmental Effects
28 Formaldehyde inhalation exposure has been associated with adverse developmental and
29 reproductive outcomes in both epidemiologic studies and experimental animal studies. Observed
30 developmental outcomes include fetal loss, structural alterations, growth retardation, and delays
31 in functional development.
32 Several occupational studies found an increased risk of spontaneous abortions among
33 formaldehyde-exposed women (Taskinen et al., 1999, 1994; John et al., 1994; Seitz and Baron,
34 1990; Axelsson et al., 1984). The Taskinen et al. (1999) study examined several reproductive
35 outcomes in women employed in the wood-processing industry, with a range of average daily
This document is a draft for review purposes only and does not constitute Agency policy.
6-12 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde exposures. The authors found that formaldehyde was associated with a more than
2 three-fold increased risk of spontaneous abortion, and with a nearly 50% decrease in a measure
3 of delayed conception indicating reduced fertility, an increased time to pregnancy, and an
4 increased risk for endometriosis in this study. In experimental animal studies, early fetal death
5 was noted following maternal formaldehyde exposures (Kitaev et al., 1984; Sheveleva, 1972),
6 supporting the epidemiologic findings that the spontaneous abortion is likely related to
7 formaldehyde exposure. Kitaev et al. (1984) hypothesized that formaldehyde may affect
8 reproductive function by stimulating the hypothalamus-pituitary-gonadal (HPG) axis, based on
9 their observations of increased ovary weight, increased number of ovulating cells, and changes
10 in blood levels of gonadotropins (LH and FSH) in female rats. Additionally, Maronpot et al.
11 (1986) reported endometrial hypoplasia with a lack of ovarian luteal tissue in formaldehyde-
12 exposed female rats. This finding may be relevant to the increased risk for endometriosis noted
13 in the Taskinen et al. (1999) study. However, additional human and animal studies are needed to
14 better understand the effects of inhalation exposure to formaldehyde on developmental outcomes
15 after early gestational windows of exposure or on the female reproductive system.
16 The findings of some occupational studies have suggested formaldehyde-related
17 associations with congenital malformations and low birth weight. In numerous experimental
18 animal studies, developmental effects have been noted following inhalation exposures to
19 formaldehyde (see Table 4-68). Exposure of rat dams to formaldehyde during pregnancy has
20 been shown to result in significantly decreased fetal weight gain (Martin, 1990; Saillenfait et al.,
21 1989; Kilburn and Moro, 1985). Other studies have noted changes in relative organ weight,
22 undescended testes, biochemical changes (e.g., ascorbic acid), and blood acidosis (Senichenkova
23 and Chebotar, 1996; Senichenkova, 1991; Kilburn and Moro, 1985; Gofmekler and
24 Bonashevskaya, 1969; Gofmekler, 1968; Pushkina et al., 1968).
25 Studies designed to assess adult male reproductive system toxicity in rats following
26 repeated inhalation exposures to formaldehyde have found concentration-dependent decreases in
27 Ley dig cell number and quality, degeneration of seminiferous tubules, decreases in testes weight,
28 alterations in sperm measures, decreased testosterone levels, alterations in trace metals in the
29 testes, and/or dominant lethal effects (Guseva, 1972; Ozen et al., 2002, 2005; Sarsilmaz et al.,
30 1999; Xing et al., 2007; Zhou et al., 2006) (see Table 4-71).
31
32 6.l.3.8.Effects on General Systemic Toxicity
33 Extrapulmonary effects such as changes in liver function enzymes and focal, chronic
34 inflammation in the heart and kidney have been observed due to formaldehyde exposure in
35 experimental animal studies. Most of these changes occurred at exposures of 20 ppm, and those
This document is a draft for review purposes only and does not constitute Agency policy.
6-13 DRAFT—DO NOT CITE OR QUOTE
-------
1 that occurred at lower formaldehyde exposures (3.7 ppm) were confounded by coexposures. The
2 underlying modes of action of liver, kidney, and cardiac effects have not been elucidated, and the
3 human relevance is unknown.
4
5 6.,1.3.9. Summary
6 Formaldehyde-induced eye, nose and throat irritation, decreased pulmonary function,
7 decreased mucociliary clearance and histopathological lesions have been extensively
8 documented in human and laboratory animal studies. These health effects are commonly noted
9 at the portal of entry as expected for exposure to a reactive gas. In addition, effects on immune
10 system responses and on the nervous and reproductive systems, including developmental effects,
11 have also been reported. An association between formaldehyde exposure and increased
12 incidence and severity of response to allergens (i.e., asthma and atopy) has been noted in
13 humans. This effect, which has also been studied in laboratory animals, might occur via a
14 neurogenic mode of action. A limited database of information that evaluates neurological effects
15 in humans following formaldehyde exposure demonstrates a potential for adverse outcomes, and
16 studies in laboratory animals have reported a variety of formaldehyde-induced neurobehavioral
17 and neurodevelopmental effects. Formaldehyde has also been associated with adverse
18 reproductive outcomes. Epidemiology studies have reported an association between
19 formaldehyde exposure and decreased fertility as well as an increased risk of spontaneous
20 abortions. Other epidemiology studies have suggested formaldehyde-related associations with
21 congenital malformations, low birth weight, and endometriosis. Animal studies have noted a
22 variety of developmental effects, including fetal death, structural alterations, and growth
23 retardation (e.g., delayed fetal skeletal ossification and decreased fetal body weight) following
24 inhalation exposure to formaldehyde, and adverse reproductive effects have been observed in
25 both males and females.
26
27 6.1.4. Carcinogenicity in Humans and Laboratory Animals
28
29 6.1.4.1.Carcinogenicity in Humans
30
31 Upper respiratory tract cancers:
32
33 Epidemiologic studies of formaldehyde-exposed workers provide sufficient evidence of a
34 causal association between formaldehyde exposure and nasopharyngeal cancer (see
35 Section 4.1.2.1.1) as well as nasal and paranasal cancers (see Section 4.1.2.1.2). The
36 epidemiologic evidence of association between formaldehyde exposure and other upper
This document is a draft for review purposes only and does not constitute Agency policy.
6-14 DRAFT—DO NOT CITE OR QUOTE
-------
1 respiratory tract cancers (see Section 4.1.2.1.3) is consistent with, and supportive of, a causal
2 association but insufficient on its own to reach a causal conclusion. However, taken together
3 with the causal evidence of an association between formaldehyde and nasopharyngeal cancer and
4 sinonasal cancer in neighboring tissues of the upper respiratory tract and sites of first contact
5 with inhaled formaldehyde, along with the strongly supportive evidence of association in
6 animals, the evidence is sufficient to conclude that formaldehyde is causally related to cancers of
7 the upper respiratory tract as a group.
8 Based on the total weight of evidence, including the results from a large and well-
9 followed longitudinal cohort study of 25,619 industrial workers and several case-control studies,
10 the epidemiologic evidence is sufficient to characterize the association between formaldehyde
11 nasopharyngeal cancer as causal in humans (Hauptmann et al., 2004; Hildesheim et al., 2001;
12 Vaughan et al., 2000). As further evaluated below, the evidence supporting a positive
13 association between formaldehyde exposure and NPC is unlikely due to chance, bias or
14 confounding. However, it should be noted that other smaller studies of formaldehyde-exposed
15 workers did not document increased NPC mortality (e.g., Coggon et al., 2003; Pinkerton et al.,
16 2004). These smaller study sizes yielded effect estimates with wide confidence intervals that
17 were not statistically inconsistent with the increased risk of mortality from nasopharyngeal
18 cancer reported in Hauptmann et al. (2004).
19 Luce et al. (2002) evaluated pooled data from 12 case-control studies conducted in
20 seven countries using a common job-exposure matrix and demonstrated a statistically significant
21 increased risk between formaldehyde exposure and sinonasal cancer exhibiting a concentration-
22 response relationship providing further causal evidence of carcinogenicity. This analysis was
23 based on a very large dataset of 930 cases and 3,136 controls, enabling the investigators to
24 control for multiple potential sources of bias and confounding and to conduct separate analyses
25 by histological type. These results are particularly convincing, as the association was
26 consistently seen for a rare subtype of sinonasal cancer which normally accounts for only 10% of
27 the reported cases.
28 In addition to the evidence of formaldehyde carcinogenicity in the nasopharynx, nose and
29 sinuses, other upper respiratory tract sites of direct contact with formaldehyde upon inhalation
30 (i.e., larynx, mouth and salivary gland) also showed evidence of increasing relative risk with
31 increasing average intensity and peak exposure in a large cohort study with exposure estimates
32 for the individual workers, although these trends did not reach the level of statistical significance
33 (Hauptmann et al., 2004). However, Hauptmann and colleagues (2004) concluded that in spite
34 of the small numbers of deaths from these rare cancers of the upper respiratory tract, the positive
35 associations of increased cancer risk with increased formaldehyde exposure were consistent with
This document is a draft for review purposes only and does not constitute Agency policy.
6-15 DRAFT—DO NOT CITE OR QUOTE
-------
1 the carcinogen!city of formaldehyde at these sites of first contact. Case-control studies also
2 provide evidence of an association between formaldehyde exposure and oral squamous cell
3 carcinoma (SCC), esophageal, and laryngeal cancers, and hypopharyngeal cancer (Gustavsson
4 et al., 1998; Laforest et al., 2000.)
5 The finding that formaldehyde inhalation causes nasal squamous cell carcinoma in
6 rodents (see Section 4.2.1.2) further supports the determination of a causal association of
7 formaldehyde exposure and increased risk of upper respiratory tract cancer in humans. Both
8 humans and animals developed tumors within the upper respiratory tract, the site expected to
9 receive direct exposure to formaldehyde.
10 Several researchers have argued that the relationship between formaldehyde exposure
11 and nasopharyngeal cancer based on existing studies has not been determined. Several
12 limitations, such as the rarity of the cancer and the imprecise estimates of exposure, are often
13 inherent in epidemiologic methods and exposure assessment. These constraints limit the ability
14 of epidemiologic studies to statistically detect associations and can lead to false negatives. The
15 results of the largest cohort study of nasopharyngeal cancer (Hauptmann et al., 2004) showed
16 statistically significant concentration-response relationships with increased risk of cancer
17 associated with increased formaldehyde exposure. However, even though this study was based
18 on 25,619 workers, only 9 cases of nasopharyngeal cancer were observed, compared to an
19 expected number of 5 cases, for a relative rate of 2.1 (with a confidence interval of 1.05-4.21)
20 (Hauptmann et al., 2004).
21 The next largest cohort study of nasopharyngeal cancer was based on 14,014 workers
22 (Coggon et al., 2003) and reported only 1 case compared to an expected number of 2 cases, for a
23 relative risk of 0.5 (with an estimated 95% confidence interval of 0.07 - 3.55; see Bosetti et al.,
24 2008). To put this finding into perspective, it is helpful to note not only the relative risk but also
25 that this effect estimate is highly unstable due to a lack of statistical power. The large width of
26 this interval (0.07 - 3.55) indicates that the range of possible true values includes both increased
27 and decreased NPC mortality and therefore does not contradict the evidence of elevated risk of
28 nasopharyngeal cancer mortality associated with formaldehyde exposure reported by Hauptmann
29 et al. (2004). The even smaller study of 11,039 textile workers by Pinkerton et al. (2004)
30 reported no cases of nasopharyngeal cancer compared to an expected number of one
31 case—yielding an effective relative risk of zero with a highly unstable 95% confidence interval
32 estimated at 0 - 3.00 (see Bosetti et al., 2008). While true that Pinkerton et al. (2004) did not
33 report an increased risk of nasopharyngeal cancer, this study did not have sufficient statistical
34 power to rule out a true association with less than a 3-fold increase in risk and therefore is
35 likewise not inconsistent with the finding by Hauptmann et al. (2004). Thus, results from these
This document is a draft for review purposes only and does not constitute Agency policy.
6-16 DRAFT—DO NOT CITE OR QUOTE
-------
1 cohort studies, with limited power to detect the relatively rare upper respiratory tract cancers
2 (e.g., NPC), are given less weight in the overall evaluation.
3 The largest occupational cohort study, conducted by the NCI (Hauptman et al., 2004), did
4 report statistically significant associations of formaldehyde exposure with carcinogenicity at the
5 sites of first contact with sufficient statistical power to rule out the null hypothesis of no
6 association. The NCI investigations controlled for potential selection bias due to the healthy
7 worker effect and for several potential confounders, including calendar year, age, sex, race, and
8 pay category. However, other potential sources of bias or confounding have been suggested with
9 respect to the strength of these data to support a causal conclusion.
10 Following reports of increased risk of NPC associated with formaldehyde exposure, a
11 series of analyses of similar data were undertaken by Marsh and coworkers (Marsh et al., 2007a,
12 b, 2002, 1996; Marsh and Youk, 2005). Briefly, these studies focused on the specific findings
13 from a single plant in the NCI cohort (Wallingford, Connecticut) that generated the majority of
14 the NPC cases. Marsh et al. (1996) confirm a significant adverse association of formaldehyde
15 with nasopharyngeal cancer but note the effects are predominantly among workers at the
16 Wallingford plant with less than one year employment. Marsh et al. (2002) report a five-fold
17 excess in risk of nasopharyngeal cancer associated with formaldehyde in both short-term and
18 long-term workers but note that the increase was concentrated among workers hired during
19 1947-1956. Marsh and Youk (2005) re-evaluated the same Wallingford workers and reported a
20 regional rate-based standardized mortality ratio (SMR) of 10.32 (95% CI = 3.79 - 22.47)
21 compared to 0.65 (95% CI = 0.08 - 2.33) for workers at the nine other plants combined.
22 However, Marsh and Youk (2005) also show that rate-based mortality ratios standardized to both
23 United States and local populations were elevated (nonsignificantly) not only at the Wallingford
24 plant but individually at each of the four other plants at which a single case of nasopharyngeal
25 cancer was reported: Plant 2 (SMRUS = 5.35), Plant 3 (SMRUS = 1.99), Plant 7 (SMRUS = 1.06),
26 and Plant 10 (SMRUS = 1.44). It should be noted that Plant 1 (Wallingford) and Plant 2 had both
27 the two highest median formaldehyde exposures and the two highest reported excess risks
28 (Marsh and Youk, 2005).
29 In another reanalysis of the NCI cohort data on the workers at the Wallingford plant,
30 Marsh and coworkers (2007a) suggested that an imprecise assessment of formaldehyde exposure
31 and an inability of the study to separate formaldehyde exposure from other potential chemical or
32 particulate exposures may have confounded the observed association between formaldehyde and
33 cancer. However, there was no evidence of any differential measurement error that could have
34 produced the observation of a spurious association. Any nondifferential exposure measurement
35 error (i.e., random error in the exposure assessment) would likely have led to an attenuated
This document is a draft for review purposes only and does not constitute Agency policy.
6-17 DRAFT—DO NOT CITE OR QUOTE
-------
1 observed effect of formaldehyde that was less than that which would otherwise have been
2 observed in the absence of measurement error.
3 The potential for confounding by particulates was explicitly examined by Hauptmann et
4 al. (2004) and it was shown that there was an exposure-response relationship with formaldehyde
5 among individuals with high particulate exposures—alleviating the potential for confounding
6 and thereby strengthening the causal interpretation of the formaldehyde relationship with an
7 increased risk of NPC. Marsh and coworkers (Marsh et al., 2007b) later suggested the reported
8 formaldehyde association was confounded by an association between silversmithing and NPC.
9 However, careful examination of that analysis (Marsh et al., 2007a) suggests that multiple
10 comparisons may have led to the reported observation with silversmithing. Additionally, the
11 reported effect was inconsistently reported between the results and the abstract sections using
12 different confidence intervals, and both sets of confidence intervals around the reported
13 association were extremely unstable spanning up to several hundred-fold. No prior studies
14 identified an associated between silversmithing and NPC. Thus it may be that silversmithing is
15 an artifactual potential confounder.
16 The increased NPC mortality observed in the NCI cohort (Hauptmann et al., 2004) has
17 been thoroughly examined for sources of bias and confounding by both the primary researchers
18 and Marsh and coworkers (Marsh et al., 2007a, b, 2002, 1996; Marsh and Youk, 2005). Despite
19 the extensive scrutiny of these results, no convincing and consistent alternative hypothesis of
20 causation has been identified. Taken together with the statistically significant association
21 demonstrating an exposure-response relationship within exposed workers, these data support the
22 conclusion that the association between formaldehyde exposure and increased risk of NPC is
23 causal.
24 Therefore, after a thorough examination of potential confounders, the association
25 between formaldehyde exposure and NPC mortality in the NCI cohort remains significant and
26 provides a positive exposure-response relationship. Additionally, case-control studies, which
27 have greater statistical power than cohort studies for rare diseases, provide strong additional
28 evidence in support of a causal association between formaldehyde exposure and the incidence of
29 NPC (Hildesheim et al., 2001; Vaughan et al., 2000). As these studies draw from different
30 demographic groups, regions of the world, and evaluate various confounding factors, there is
31 little potential for these consistently reported associations to be artifactual, confounded by
32 common exposures, or a result of bias or chance.
33
34 Lymphohematopoietic cancers:
This document is a draft for review purposes only and does not constitute Agency policy.
6-18 DRAFT—DO NOT CITE OR QUOTE
-------
1 Numerous epidemiologic studies have also reported an association between
2 formaldehyde-exposed workers, especially "professional" workers (e.g., pathologists,
3 embalmers, and funeral directors), and increased risk of lymphohematopoietic cancers (see
4 Table 4-90). Positive associations between formaldehyde exposure and lymphohematopoietic
5 cancers have been reported for chemical workers (Wong et al., 1983; Bertazzi et al., 1986),
6 embalmers (Walrath and Fraumeni, 1983, 1984; Hayes et al., 1990), anatomists and pathologists
7 (Harrington and Shannon 1975; Hall et al., 1991; Levine at al., 1984; Stroup et al., 1986;
8 Matanoski et al., 1989). However, clear associations (in terms of overall standardized mortality
9 ratios (SMRs) or proportional mortality ratios (PMRs) were not reported in analyses for garment
10 workers, iron-foundry workers, and a large US industrial cohort (Pinkerton et al., 2004;
11 Andjelkovich et al., 1995; Beane Freeman et al., 2009; Marsh et al., 1996), although associations
12 were observed in some of these studies when exposure-response relationships were considered.
13 Several published meta-analyses are available which more formally assess the strength of
14 association between formaldehyde exposure and mortality from all lymphohematopoietic cancers
15 (see Section 4.1.2.2.1.3). Pooled SMRs indicate stronger associations for professional workers
16 (embalmers, anatomists and pathologists) than industry workers (see Table 4-90). Bosetti et al.
17 (2008) found similar relationships, with a pooled SMR of 1.31 (95% CI 1.16-1.47) for
18 'professionals' (i.e., embalmers, anatomists and pathologists) versus a pooled estimate of 0.85
19 (95% CI 0.74-0.96) for industrial workers. A recent metaanalysis by Zhang et al. (2009) reports
20 a summary relative risk of 1.25 (95% CI 1.09-1.43) for both professional and industry workers
21 for all lymphohematopoietic cancers (ICD 9 codes 200-209).
22 Two well-designed cohort studies found significant positive associations between
23 formaldehyde-exposed professional workers and lymphohematopoietic cancer, particularly
24 leukemia, using cumulative exposure measures not previously used and using internal
25 comparison groups. The largest cohort study of industrial workers exposed to formaldehyde
26 (N=25,619), with the most extensive exposure assessment (Blair et al., 1986; Stewart et al.,
27 1986) and with the cohort followed for a median duration of 35 years (Hauptmann et al., 2003)
28 demonstrated that formaldehyde was a risk factor for lymphohematopoietic cancers, independent
29 of other risk factors, such as benzene exposure and smoking. This finding was reconfirmed with
30 an additional 10 years of follow-up (Beane Freeman et al., 2009). Another industrial cohort
31 study reported a significant increase in the risk of leukemia in garment workers 20 years after
32 their initial exposure and in workers with 10 or more years of exposure to formaldehyde
33 (Pinkerton et al. 2004). A third large occupational cohort study (Coggon et al., 2003) that did
34 not evaluate their findings with regard to latency reported somewhat lower mortality from
This document is a draft for review purposes only and does not constitute Agency policy.
6-19 DRAFT—DO NOT CITE OR QUOTE
-------
1 leukemia and other lymphatic and hematopoietic cancers than expected compared to national
2 rates.
3 The associations between myeloid leukemia and formaldehyde exposure are strong and
4 consistent (see Table 4-92). Of the four studies which formally assess myeloid leukemia
5 mortality, all are positive, including cohorts of both professional and industrial workers (Beane
6 Freeman et al., 2009; Hayes et al., 1990; Pinkerton at al., 2003; Stroup et al., 1986). Although
7 few cases exist for further subtype analysis, the available data indicate either no differences in
8 SMRs for acute myeloid leukemia (AML) versus chronic myeloid leukemia (CML) (Hayes et al.,
9 1990; Pinkerton et al., 2003) or suggest CML is more prominent (Blair et al., 2000; Stroup et al.,
10 1986). The association between formaldehyde exposure and myeloid leukemia in embalmers has
11 recently been confirmed in a large nested case control study by Hauptman et al (2009) which
12 includes cases identified from the previous studies of Hayes et al. (1990) and Walrath and
13 Fraumeni (1983 and 1984). Exposure estimates were based on interviews with next-of kin for
14 duration of job actively embalming and total number of embalmings performed. Strong and
15 statistically significant exposure-response relationships are demonstrated for duration of
16 exposure, total number of embalmings performed and estimated cumulative exposure to
17 formaldehyde with odds ratios of 13.6(1.6-119.7), 12.7(1.4-112.8) and 13.2(1.5-115.4)
18 respectively (Hauptmann et al., 2009).
19 The reported associations between formaldehyde exposure and lymphohematopoietic
20 cancers in general, and leukemia (especially myeloid leukemia) in particular, were in workers
21 exposed in very different environments (i.e., mortuary, chemical industry and garment industry).
22 Since coexposures to other agents are considerably different between these work environments,
23 it is unlikely that influence of confounding exposures plays a role in the observed associations.
24 There is no evidence of bias in the published reports, and the consistency across numerous
25 studies over time is sufficient to conclude that the results are not due to chance. Additionally,
26 where data are available for analysis, increased myeloid leukemia is not the sole driver of
27 increased leukemia and all lymphohematopoietic cancers (see Table 4-91). An evaluation of the
28 epidemiologic evidence for solid tumors of lymphoid origin indicates an association between
29 formaldehyde exposure and both Hodgkins lymphoma and multiple myeloma, but not
30 non-Hodkins lymphoma in general (see Section 4.1.2.2.1.4 and Section 4.5.2.6).
31 It has been argued that it is biologically implausible for a highly reactive agent such as
32 formaldehyde, whose primary action is expected to be at the portal of entry, to cause acute
33 lymphoid or myeloid leukemias (ALL and AML, respectively), which are both commonly
34 believed to arise from transformation of stem cells in the bone marrow. The modes of action
35 (MOAs) by which formaldehyde may induce these observed cancers are unknown, although it
This document is a draft for review purposes only and does not constitute Agency policy.
6-20 DRAFT—DO NOT CITE OR QUOTE
-------
1 has been postulated that circulating stem cells (Hauptmann et al., 2003) (e.g., early progenitor
2 cells in circulating blood or pluripotent cells in nasal/oral passages) may travel to bone marrow
3 where they become leukemic stem cells (Zhang et al., 2010a, b). In contrast, the mechanism for
4 the chronic lymphatic leukemia, lymphomas, multiple myelomas (from plasma B-cells) and
5 unspecified lymphohematopoietic cancers may involve an etiology in peripheral tissues, such as
6 cells, cell aggregates, germinal centers and lymph nodes. An association of these cancers to a
7 reactive exogenous agent primarily acting at the point of entry is biologically plausible.
8
9 6AA.2.Carcinogenicity in Laboratory Animals
10 The carcinogenic potential of formaldehyde is well documented in numerous animal
11 bioassays, especially for sites of first contact. Inhalation exposure of formaldehyde induced
12 primarily squamous cell carcinomas (SCC) in nasal passages of rats (Feron et al., 1988;
13 Holmstrom et al., 1989a; Woutersen et al., 1989; Tobe et al., 1985; Kamata et al., 1997; Albert
14 et al., 1982; Sellakumar, 1985; Kerns et al., 1983; Monticello et al., 1996) and mice (Battelle
15 Columbus Laboratories, 1981; Swenberg et al., 1980; Kerns et al., 1983; CUT, 1982).
16 Formaldehyde given as 0.5% formalin orally in drinking water to adult rats induced higher
17 incidences of papillomas in the forestomach, adenomatous hyperplasia in the fundus, and
18 adenocarcinomas in the pylorus in a 40-week study using an initiation-promotion protocol in rats
19 (Takahashi et al., 1986). Formaldehyde is toxic at the portal of entry in rodents, causing
20 increased cell proliferation, DPX formation, and focal lesions in the GI tract or upper respiratory
21 tract (depending on the route of exposure). The portal of entry toxicity of formaldehyde further
22 supports a finding of formaldehyde induced POE cancer in animal bioassays.
23 Direct support for lymphohematopoietic cancers in animal bioassays is less convincing.
24 Although many of the available chronic studies did not examine lymphoma/leukemia incidence,
25 four studies allow for some evaluation of the leukemic potential of formaldehyde. Inhalation
26 exposure of formaldehyde increased lymphoma in female mice and leukemia in female F344
27 rats, but not male rats (Battelle Laboratories, 1981). No increases in leukemia or lymphoma
28 were seen in male Wistar rats when exposed to formaldehyde in drinking water (Til et al., 1989)
29 or male rats after chronic inhalation exposures (Sellakumar et al., 1985).
30
31 6 A A.3.Carcinogenic Mode(s) of Action
32 Multiple plausible modes of action (MOAs) are presented in the document so as to
33 explore ways in which a combination of factors may contribute to cancer incidence in a
34 population exposed to formaldehyde. Multiple MO As for formaldehyde-induced cancer can be
35 reasonably supported based on various known biological actions of formaldehyde (e.g.,
This document is a draft for review purposes only and does not constitute Agency policy.
6-21 DRAFT—DO NOT CITE OR QUOTE
-------
1 mutation, cell proliferation, cytotoxicity and regenerative cell proliferation). Additionally,
2 alternative actions, such as immunosuppression or viral reactivation, are possible, although few
3 data exist to evaluate their potential relevance. Rather than a single MO A, it is plausible that a
4 combination of these factors contribute to cancer incidence in an exposed population.
5 Considering multiple factors may help to better understand the biological and mechanistic basis
6 for the increases in cancer incidence observed in exposed human populations. Unlike animal
7 bioassays, results in human epidemiological studies reflect not only the effects of the agent of
8 concern but also numerous other risk factors (e.g., viral status, diet, smoking, etc.). Additionally,
9 human studies may be impacted by biological human variability across individuals, cancer
10 biology (subtypes) and wide variability in exposure regimens in human populations.
11 The overall weight of evidence supports a role of mutagenic activity in formaldehyde's
12 carcinogenic MO A both for respiratory tract cancer and lymphohematopoietic cancers. As
13 reviewed in Section 4.3 and summarized in Section 4.5.3.1, numerous studies provide evidence
14 of formaldehyde's direct mutagenic activity and supports the relevance these data to
15 formaldehyde's carcinogenicity. It can be shown that:
16
17 1) Formaldehyde directly interacts with DNA, generating DNA-protein cross-links and
18 DNA adducts (in vitro, in vivo) in multiple species,
19 2) DNA-protein cross-links exhibit a dose-response relationship to formaldehyde exposure
20 in respiratory tract of laboratory animals and are observed at exposure concentrations of
21 relevance to some people (0.3 ppm, 0.7 ppm),
22 3) Formaldehyde-induced DNA-protein cross-links have been associated with
23 formaldehyde-induced micronuclei and chromosomal aberrations (in vitro),
24 4) Mutations induced by formaldehyde due to small deletions and rearrangements in DNA
25 in various experimental systems are consistent with formaldehyde's observed clastogenic
26 effects (micronuclei and chromosomal aberrations) (in vitro, in vivo),
27 5) Formaldehyde-induced mutations and clastogenic effects occur at levels below where
28 significant cytotoxicity is detected (in vitro),
29 6) Formaldehyde exposure has been correlated to similar increased micronuclei and
30 chromosomal aberrations in human buccal and oral cells corresponding to sites where
31 formaldehyde-induced tumors arise, and
32 7) Chromosomal damage in blood-borne immune cells, relevant to agent-induced
33 lymphohematopoietic cancers has been documented in formaldehyde exposed workers
34 including increased micronuclei and chromosomal aberrations, increased incidence and
35 aneuploidy in hematopoietic stem cells.
36
This document is a draft for review purposes only and does not constitute Agency policy.
6-22 DRAFT—DO NOT CITE OR QUOTE
-------
1 In addition, mutations may arise indirectly from formaldehyde-induced DNA damage
2 during cell proliferation or due to errors in DNA repair mechanisms. Therefore, formaldehyde's
3 DNA reactivity on a population of proliferating cells strengthens the role of formaldehyde-
4 induced mutagenicity in its carcinogenic MOA. The nasal and gut mucosa are tissues which are
5 continually sloughing and regenerating cells (Junqueira et al., 1992). Mucosal cells proliferate
6 in response to environmental challenges in order to repair cell damage, increase adaptive
7 response and remodel tissue. Additionally, since the pseudostratified epithelium of the
8 respiratory tract is only 1-2 cells in depth, cells with proliferative capacity would be directly
9 impacted by formaldehyde during exposure. Formaldehyde-induced clastogenic effects have
10 been demonstrated in these tissues (e.g., nasal) in humans, as well as in tissues which possess
11 stratified epithelium (e.g., buccal). Therefore, formaldehyde would not need to transport beyond
12 the portal of entry to directly impact and induce DNA mutations in routinely proliferating cells.
13 In regards to generating the observed clastogenic effects (micronuclei and chromosomal
14 aberrations in peripheral blood lymphocytes, aneuploidy in circulating hematopoietic stem cells),
15 it is less clear as to where formaldehyde is making contact with components of the immune
16 system. Mature lymphocytes present in nasal and gut tissues, and would be vulnerable to the
17 direct toxic actions of formaldehyde including genotoxicity. Since mature lymphocytes
18 routinely traffic through the body and clonally respond in response to an immune challenge, the
19 observed effects in peripheral blood lymphocytes (micronuclei and chromosomal aberrations)
20 are consistent with direct action on these cells. Lymphohematopoietic cancers are known to
21 arise from mature lymphocytes including: Hodgkin lymphoma, multiple myeloma some
22 leukemia and non-Hodgkin lymphoma (Greaves 2004, Harris et al., 2000).
23 Formaldehyde may also be directly acting upon circulating stem cells or more mature
24 progenitor cell in the peripheral blood (Zhang et al., 2010a). Any genetic damage sustained by
25 circulating cells could contribute to a broad spectrum of lymphohematopoietic cancers if those
26 cells returned to the bone marrow and contributed to hematopoiesis. Evidence of bone marrow
27 toxicity and stem cell aneuploidy has been reported in formaldehyde exposed workers (Zhang
28 et al., 2010b). Finally, formaldehyde is readily hydrated in aqueous systems, existing in
29 equilibrium with its hydrated form methylene glycol, which is able to transport through the
30 blood. It has been hypothesized that this hydration reaction may allow formaldehyde to act
31 systemically and therefore on the bone marrow directly (Zhang et al., 2010a.) Formaldehyde-
32 induced DNA damage, and resulting mutation in the bone marrow and circulating stem cells
33 could contribute to any of the lymphohematopoietic cancers including leukemia (both lymphoid
34 and myeloid) as well as myeloproliferative disorders.
This document is a draft for review purposes only and does not constitute Agency policy.
6-23 DRAFT—DO NOT CITE OR QUOTE
-------
1 Cell replication allows unrepaired DNA damage to be "fixed" into heritable changes to
2 the genome. Therefore, increased cell proliferation could serve not only to increase the
3 mutagenic effects of formaldehyde on a given tissue but also to enhance the mutagenic effects of
4 other agents in the diet or in the environment. Since epidemiological studies include humans
5 exposed to a range of agents in the environment, increased cell proliferation could contribute to
6 increased cancer incidence. The promotion studies in animal bioassays, though limited in
7 number, support the relevance of formaldehyde's ability to enhance the actions of other agents
8 (initiators) on tumor formation.
9 Although the other biologic effects discussed above have not been explicitly tested in
10 animal systems, the available data are consistent with these actions contributing to the
11 carcinogenic potential of formaldehyde. For example, localized immunosuppression by
12 formaldehyde may serve to increase viral reactivation (e.g., EBV, HPV etc.) or decrease tissue
13 surveillance and immune activity against preneoplastic cells. Both these actions could contribute
14 to increased cancer risk in a human population, which may not be evident in animal bioassays,
15 where the animals are not subject to the many risk factors for human cancer. Even the simple
16 action of the breakdown of the mucociliary apparatus could increase cancer incidence by
17 increasing toxic insult to the URT and increasing URT infections. Again, these actions may be
18 relevant to human populations, but they have not been adequately tested in animal bioassays.
19 Animal bioassays suggest a role for regenerative proliferation in contributing to
20 formaldehyde's carcinogenicity. However, these data are not evidence against a role of direct
21 mutagenic action either in the observed tumorigenicity or in the potential low-dose
22 carcinogenicity of formaldehyde. As reviewed, a role for mutagenic action is also consistent
23 with the results of the animal bioassays (Crump et al, 2008; Subramaniam et al., 2007, USEPA
24 2008). The mutagenic effects of formaldehyde are well-documented to occur below levels of
25 significant cytotoxicity. This observation is important for the relevance of formaldehyde-
26 induced mutagenicity to human health risk. Given the above sequence of evidence—from the
27 nature of formaldehyde's DNA reactivity through clastogenic effects observed in human cells
28 from the various tumor sites—there is an adequate weight of evidence (WOE) to consider
29 formaldehyde-induced mutations relevant to human carcinogenic risk. Although occupational
30 exposures may have resulted in high episodic exposures (especially historically), it is unlikely
31 that any worker would have endured repeated exposures which resulted in gross focal lesions to
32 the upper respiratory tract (URT) or oro-digestive tract as seen in the animal bioassays. It is
33 noteworthy that even without these gross formaldehyde-induced lesions, cancer incidence is
34 increased from occupational (and perhaps nonoccupational) exposures to formaldehyde.
This document is a draft for review purposes only and does not constitute Agency policy.
6-24 DRAFT—DO NOT CITE OR QUOTE
-------
1 Therefore, we believe formaldehyde carcinogenicity can be attributed, at least in part, to a
2 mutagenic MOA.
3
4 6.1.5. Cancer Hazard Characterization
5 Formaldehyde is Carcinogenic to Humans by the Inhalation Route of Exposure.
6 Human epidemiological evidence is sufficient to conclude a causal association between
7 formaldehyde exposure and nasopharyngeal cancer, nasal and paranasal cancer, all leukemias,
8 myeloid leukemia and lymphohematopoietic cancers as a group. Epidemiological evidence is
9 also strongly supportive of, but in itself not sufficient for, a conclusion of causal association for
10 other upper-respiratory tract cancers, Hodgkins lymphoma, or multiple myeloma. Animal
11 bioassays consistently demonstrate formaldehyde-induced nasal cancers in rodents which
12 provide strong support for the observed upper respiratory tract cancers in humans. Limited
13 evidence from animal bioassays is available to support the conclusion from human
14 epidemiologic data that formaldehyde causes some types of lymphohematopoietic cancers.
15
16 6.2. DOSE-RESPONSE CHARACTERIZATION
17 6.2.1. Noncancer Toxicity: Reference Concentration (RfC)
18 The portals of entry are major targets for formaldehyde, as can be seen in many studies,
19 because formaldehyde is highly reactive and water soluble. Human and laboratory animal
20 studies demonstrate that formaldehyde also causes systemic effects, including neurotoxicity,
21 reproductive toxicity, developmental toxicity, and immunotoxicity, although the data are less
22 extensive than those supporting the sensory irritation and respiratory tract effects. Critical data
23 gaps have been identified and uncertainties associated with data deficiencies are more fully
24 discussed in Chapter 5 and summarized below.
25
26 6.2.1.1. Assessment Approach Employed
27 RfC values for noncancer effects are derived using EPA's RfC methodologies (U.S. EPA,
28 1994, 1993, 2002b). EPA reviewed the existing literature and identified health effects associated
29 with formaldehyde exposure, defining health effect categories where evidence was sufficient:
30 sensory irritation, respiratory tract pathology, pulmonary effects, asthma, increased allergic
31 sensitization, immune function, neurological and behavioral effects and reproductive and
32 developmental effects. Specific key studies were identified within each health effects category
33 which provided adequate exposure-response information to support RfC derivation (see
34 Table 5-4). Although not all identified endpoints are represented by these studies, at least one
35 study was identified for each category. A screening process (described in Section 5.1.3.1) was
This document is a draft for review purposes only and does not constitute Agency policy.
6-25 DRAFT—DO NOT CITE OR QUOTE
-------
1 used to identify key studies for a variety of health effects that would best inform the derivation
2 oftheRfC. For each selected key study, a candidate RfC (cRfC) was derived. In several cases
3 more than one alternative was considered for application of the uncertainty factor (UF)
4 addressing human variability (see Table 5-6).
5
6 6.2.1.2. Derivation of Candidate Reference Concentrations
1 Seven studies were selected as key studies for further consideration in RfC derivation
8 (see Section 5.3.1, Table 5-4). Candidate RfCs from these studies address various health effects
9 including: sensory irritation, respiratory effects, asthma, increased allergic sensitization, and
10 decreased fecundity (see Table 5-6). From these studies three cocritical studies were selected
11 which provide similar cRfCs for related health effects (Rumchev et al., 2002; Garrett et al., 1999
12 a,b; Krzyzanowski et al., 1999). These three studies identify serious health effects in residential
13 populations including children: increased asthma incidence, decreased pulmonary function,
14 increase in respiratory symptoms, and increased allergic sensitization (Rumchev et al., 2002;
15 Garrett et al., 1999 a,b; Krzyzanowski et al., 1999). Asthma, allergic sensitization, altered
16 pulmonary function, and symptoms of respiratory disease are not only clinically related, but
17 etiologically related, and it is reasonable that they should be considered together. These health
18 effects are observed below the exposure levels that result in sensory irritation, and the resulting
19 cRfCs are correspondingly lower—ranging from 2.8 tol 1 ppb—depending on the study,
20 endpoint considered, and the application of alternative uncertainty factors for human variability
21 (see Table 6-1). Additionally, these cRfCs are considered protective of the decreased
22 fecundability density ratio (FDR) reported by Taskinen et al. (1999) which yielded a cRfC of
23 8.6 ppb. One of the uncertainties in the cRfC for decreased FDR is the use of a time-weighted
24 exposure metric which does not address possible contributions of peak exposure levels to the
25 observed health effect thus; it is possible that a cRfC of 8.6 ppb is lower than is needed for
26 protection against decreased FDR.
27 As discussed in Section 6.2.1.4, there are uncertainties in establishing an RfC which are
28 not fully captured in the quantitative process or the standard uncertainty factors. The range of
29 RfCs from the critical studies (even with various alternative considered for the human variability
30 uncertainty factor are in close agreement spanning only 1A order of magnitude.) Therefore EPA
31 is considering a simple mean of these cRfCs as adequately representative of the three cocritical
32 studies. Alternatives are to take the median as a different way to represent the three studies
33 together, or the lowest cRfC as most protective. There is little numerical difference in the result
34 of these decisions.
35
This document is a draft for review purposes only and does not constitute Agency policy.
6-26 DRAFT—DO NOT CITE OR QUOTE
-------
1 6.2.1.3 Adequacy of Overall Data Base for RfC Derivation
2 The database of available laboratory animal studies, clinical and epidemiological studies,
3 and supporting mechanistic information for formaldehyde is substantial. Many of the health
4 effects are well studied in animals and humans, especially those endpoints related to sensory
5 irritation and respiratory effects at the portal of entry, such as impacts on respiratory tract
6 pathology, asthma and reduced pulmonary function. This is reflected in the number and high
7 quality of human studies presented in Table 5-4 and supporting data summarized in Chapter 4.
8 The data also indicate effects in other health effect categories, specifically neurotoxic
9 effects, reproductive toxicity, and developmental toxicity (see Section 5.1.2). These nonportal-
10 of-entry effects are areas where additional research may be warranted to reduce uncertainty and
11 better characterize the potential for health effects and the formaldehyde concentrations at which
12 they might occur in humans.
13 EPA guidance indicates that an uncertainty factor for database deficiencies should be
14 applied where there is an indication that the existing studies may not completely characterize the
15 hazard of a specific agent. This may be the result of lacking studies to assess toxicity to key
16 functional areas or organ systems, or where "... a review of existing data may also suggest that a
17 lower reference value might result if additional data were available." (U.S. EPA 2002b)
This document is a draft for review purposes only and does not constitute Agency policy.
6-27 DRAFT—DO NOT CITE OR QUOTE
-------
a 5
o S.
*•* TO
rs i
KO
S3
Table 6-1. Summary of candidate reference concentrations (RfC) for cocritical studies
Endpoint
Study
Study
size
Homes
Children
POD (ppb)
Application of
study-specific UF
UFL UFS
UFH
cRfC1
(PPb)
Respiratory effects/asthma and sensitization
Reduction of
PEFR in children
(10%)
Asthma
prevalence
Asthma, atopy
and severity of
allergic
sensitization
Krzyzanowski
etal. (1990)
Rumchev et al.
(2002)
Garrett et al.
(1999 a,b)
208
192
148
Yes
Yes
Yes
Yes
Yes
Yes
BMCLio= 17
NOAEL = 33
LOAEL = 28
1
1
3
1
3
1
3
5.6
Alternative A
3
3.3
Alternative B
l
11
Alternative A
3
2.8
Alternative B
l
9.3
§
to
oo
Notes: 1: The final RfC will be rounded to one significant digit per EPA policy. Since the Candidate RfC is an interim calculation, two-significant digits are
retained as is common practice in mathematics {i.e., one significant diget more that the final result, to avoid rounding errors compounding across multiple
mathematical manipulations.
fe
H
o
o
-------
1 Application of an uncertainty factor of 3 was considered by EPA based on the lack of a
2 satisfactory two-generation study to fully evaluate the effects of formaldehyde exposure on
3 reproductive and developmental endpoints and limitations of the available studies evaluating
4 neurotoxic effects. An uncertainty factor of 3 rather than 10 was considered given the relative
5 completeness of the database across all major health effect categories such that it is believed all
6 major health effects have been identified at least qualitatively. The observed adverse health
7 effect levels (LOAELs) for those endpoints where the database is not adequate for alternative
8 RfC derivation are above the range of candidate RfCs; however, it is unclear if the candidate
9 RfCs would be protective of these other health effects (neurotoxic, reproductive and
10 developmental) since NOAELs were not identified for several observed health effects.
11 Therefore EPA is considering several options to address database deficiencies in the final
12 RfC.
13
Approaches to the application of a database uncertainty factor:
Options EPA is considering include:
(1) Provide an RfC derived from studies of respiratory and allergenic responses and
protective of sensory irritation effects with a database uncertainty factor of one given
significant data on formaldehyde, but noting that further research reproductive,
developmental and neurotoxic effects would be valuable.
(2) Provide an RfC with a database uncertainty factor of one, with this RfC explicitly
identified as being protective of the well-studied effects.
(3) Apply a database UF of 3 to the RfC derived from studies of respiratory and allergenic
responses to reflect the potential that reproductive, developmental, or neurotoxic effects
might occur at lower doses:
(4) Provide both an RfC identified as protective of the better-studied effects and an RfC with
a database uncertainty factor of 3 incorporated to account for limits to the data on
reproductive, developmental and neurotoxic effects.
14
15
16 It is unclear what uncertainty factors are appropriate to account for human variability and
17 deficiencies in the overall database. For this reason, several alternatives have been presented.
18
19 6.2.1.4.Uncertainties in the Reference Concentration (RfC)
20 A number of uncertainties that underlie the RfC for formaldehyde are discussed in this
21 section. A fundamental uncertainty in an RfC is that the critical study(ies) and endpoint(s)
This document is a draft for review purposes only and does not constitute Agency policy.
6-29 DRAFT—DO NOT CITE OR QUOTE
-------
1 selected reflect an actual hazard, i.e., a chemically related effect. As summarized in Section
2 6.1.3, there is strong and consistent evidence, from both human and laboratory animal studies,
3 for the critical effects that form the basis of the RfC for formaldehyde. This section pertains to
4 uncertainties in the quantitative derivation of the RfC.
5
6 6.2.1.4.1. Point of departure (POD).
1 Most of the studies considered for RfC derivation did not provide enough data to support
8 benchmark dose modeling. Rather, the PODs for most studies were LOAELs or NOAELs,
9 which have a number of shortcomings relative to a POD obtained from benchmark dose-
10 response modeling (i.e., a benchmark concentration or dose):
11
12 • LOAELs and NOAELs are a reflection of the particular exposure/dose levels used in a
13 study, contributing some inaccuracy to the POD determination.
14 • LOAELs and NOAELs are often determined based on statistical significance and, thus,
15 reflect the number of study subjects or test animals. Studies are typically dissimilar in
16 detection ability and statistical power, with smaller studies tending to identify higher
17 exposure levels as NOAELs relative to larger, but otherwise similarly designed, studies.
18 • Different LOAELs and NOAELs represent different response rates, so direct qualitative
19 and quantitative comparisons are not possible.
20
21 PODs identified from benchmark dose models overcome some of the deficiencies
22 associated with LOAELs and NOAELs. Benchmark models were used for two inhalation data
23 sets—Hanrahan et al. (1984) and Krzyzanowski et al. (1990).
24 It should also be noted, however, that even for benchmark concentrations/doses there is
25 often uncertainty, in particular for continuous responses, about what response level to select as
26 the benchmark response, i.e., where to define the cut-point between a level of change that is not
27 adverse and one that is adverse. In addition, benchmark dose models currently in use are purely
28 mathematical models and are not intended to accurately reflect the biology of the effect being
29 modeled.
30 Another source of uncertainty in the POD is the adjustment for continuous exposure.
31 RfCs are meant to apply to continuous (24 hour/day) exposures. Exposure patterns in human
32 and laboratory animal inhalation studies are typically not for continuous exposures, and
33 assumptions must be made in converting reported exposure levels to equivalent continuous
34 exposures. Similarly, there are uncertainties about potential dose rate effects, in particular the
35 effect of peak exposures in occupational studies.
36 6.2.1.4.2. Extrapolation from laboratory animal data to humans.
This document is a draft for review purposes only and does not constitute Agency policy.
6-30 DRAFT—DO NOT CITE OR QUOTE
-------
1 Because the inhalation database for formaldehyde contains many human studies for a
2 variety of health effects, it was not necessary to rely on animal data for the endpoints from which
3 the RfC was derived. Thus, unlike for most RfCs, this is not a source of uncertainty in the RfC
4 for formaldehyde.
5
6 6.2.1.4.3. Human variation.
1 Heterogeneity among humans is another uncertainty associated with extending results
8 observed in a limited human study population or laboratory animal experiment to a larger, more
9 diverse human population.
10 For three of the studies used to derive the RfC, a value of 3 was used for the human
11 variability UF (rather than the default value of 10) because the studies had an apparent over-
12 representation of populations expected to have increased susceptibility (see Section 5.5.3.1):
13
14 • The residential study by Ritchie and Lehnen (1987) evaluated eye, nose, and throat
15 irritation in a large number of subjects, including children and the elderly. As a result of
16 the study's participation criteria, individuals with greater sensitivity were potentially
17 over-represented.
18 • Thirty percent of the subjects in the residential study by Krzyzanowski et al. (1990) are
19 children, who are more sensitive to formaldehyde-associated decreases in peak expiratory
20 flow rates (PEFR) than adults. The candidate RfC determination for this study focused
21 on the results in the children, among which asthmatics were over-represented (roughly
22 3-times) compared to the national average.
23 • Garrett et al. (1999 a,b) conducted a cross-sectional survey of allergy and asthma-like
24 symptoms in children with or without a doctor's diagnosis of asthma. The study was
25 designed to include a high proportion of asthmatic children, a sensitive population for the
26 effects being studied.
27
28 EPA notes, however, that, while a human variability UF of 3 rather than 10 was used to
29 attempt to account for certain special attributes of these studies/effects, there is still uncertainty
30 about how much of the overall population heterogeneity is actually reflected even in these
31 relatively diverse residential studies.
32
33 6.2.1.4.4. Subchronic-to-chronic extrapolation.
34 RfCs are intended to apply to chronic lifetime exposures. If a study is subchronic
35 (typically less than 10% of lifetime), an UF for subchronic-to-chronic extrapolation is generally
36 applied to the candidate RfC for that study. For the human residential and occupational studies
37 comprising the key studies for the RfC in this assessment, the average durations of exposure in
This document is a draft for review purposes only and does not constitute Agency policy.
6-31 DRAFT—DO NOT CITE OR QUOTE
-------
1 the households or workplaces under study is unknown. In this assessment, these studies were
2 considered chronic in nature and no subchronic-to-chronic UF was applied. However, there is
3 uncertainty about whether or not the responses observed fully reflected the potential effects of
4 chronic exposure, especially for effects in children, where effects on the developing respiratory
5 and immune systems, for example, could be predisposing the children to further health effects
6 later in life.
7
8 6.2.l.5.Conclusions
9 Seven different noncancer health effects were identified from formaldehyde inhalation
10 exposure studies, including: 1) sensory irritation of the eyes, nose, and throat, 2) upper
11 respiratory tract pathology, 3) pulmonary function, 4) asthma and atopy, 5) neurologic and
12 behavioral toxicity, 6) reproductive and developmental toxicity, and 7) immunological toxicity.
13 Of note, epidemiological evidence is available for most of these noncancer effects. EPA has
14 derived candidate RfCs for critical effects based on seven key studies. Three cocritical studies
15 were selected which provide similar cRfCs for related adverse health effects observed in
16 residential populations including children i.e., increased asthma incidence, decreased pulmonary
17 function, increase in respiratory symptoms, and increased allergic sensitization (Rumchev et al.,
18 2002; Garrett et al., 1999 a,b; Krzyzanowski et al., 1999). The resulting cRfCs fall in a range
19 between 2.8 and 11 ppb, depending on the study, or endpoints considered, and the application of
20 alternative uncertainty factors for human variability (see Table 6-1). The RfC is taken as the
21 average of the cRfCs from the three cocritical studies (See Section 6.2.1.2).
22 EPA has assessed the adequacy of the overall database for RfC derivation, and although
23 the database is quite large, and provides significant information on well studied POE effects.
24 There are remaining uncertainties in the database. Most notably, there is a need for additional
25 exposure-response information for observed neurotoxic effects, reproductive and developmental
26 effects as well as a two-generation study to evaluate the effects of formaldehyde exposure on
27 reproductive and developmental endpoints. EPA is considering 4 options to address database
28 uncertainties in the final RfC (see Section 6.2.1.3). It is unclear what uncertainty factors are
29 appropriate to account for human variability and deficiencies in the overall database. For this
30 reason, several alternatives have been presented. EPA is seeking advice from the NAS and the
31 public on this matter.
32
This document is a draft for review purposes only and does not constitute Agency policy.
6-32 DRAFT—DO NOT CITE OR QUOTE
-------
1 6.2.2. Cancer Risk Estimates
2 6.2.2.l.Choice of Data
3 As explained above, the human epidemiologic data and the animal bioassay data indicate
4 multiple sites of concern, remote as well as at the portal of entry. The quantitative cancer risk
5 derivations in this document consider the risks of lymphohematopoietic cancers and solid
6 cancers of the respiratory tract. When adequate human data are available, as is the case with
7 formaldehyde, it is generally preferable to base cancer risk estimates on the human data rather
8 than on data from experimental animals because of the inherent uncertainties associated with
9 interspecies extrapolation. Sufficient exposure-response data from a large, high-quality
10 epidemiologic study for the quantitative estimation of risk were available for some
11 lymphohematopoietic cancers and for nasopharyngeal cancer.15 Risk estimates based on nasal
12 tumors in rats were also derived for comparison with the estimates based on human data. The
13 data used for the quantitative risk assessment are as follows:
14
15 1. Nasopharyngeal cancer (NPC): The dose-response modeling of NPCs is based on results
16 from a large NCI cohort study of over 25,000 workers in 10 U.S. plants producing or
17 using formaldehyde (Hauptmann et al., 2004).
18 2. Lymphohematopoietic cancers: The dose-response modeling of select
19 lymphohematopoietic cancers is based on results from a more recent follow-up study (of
20 lymphohematopoietic malignancies only) of the same NCI cohort (Beane Freeman et al.,
21 2009).
22 3. Squamous cell carcinoma (SCC) in the upper and lower respiratory tract: An increased
23 incidence of nasal SCC was seen in two large long-term bioassays using F344 rats (Kerns
24 et al., 1983; Monticello et al., 1996). Although other studies in laboratory animals exist,
25 these two studies, when combined, provided the most robust data for analyses. The nasal
26 tumor incidence data from these rat bioassays is used for extrapolating the risk of SCC to
27 the entire human respiratory tract.16
28
15 Only two other epidemiological studies were available with quantitative exposure estimates for the individual
workers. One was a much smaller study (it focused on one of the ten plants covered in the selected study), and it
evaluated only pharyngeal cancers. The second was a study of lymphohematopoietic and brain cancers in funeral
industry workers which, as discussed in detail in Section 5.1.1, had serious limitations in the exposure assessment,
precluding its use for quantitative risk assessment.
16 That is, we do not assume site concordance between rat and human. This is reasonable because the respiratory
and transitional cell types considered to be at risk of SCC in the upper respiratory tract are also prevalent in the
lower human respiratory tract. Greater fractional penetration of formaldehyde is thought to occur posteriorly in the
human respiratory tract compared to the rat (Kimbell et al. 2001, Overton et al. 2001). Furthermore, some
epidemiological studies reported an increase in lung cancer with formaldehyde exposure (Gardner et al. 1993, Blair
et al. 1990, 1986), and lesions were seen in the lower respiratory tract of rhesus monkeys exposed to formaldehyde
(Monticello et al. 1989).
This document is a draft for review purposes only and does not constitute Agency policy.
6-33 DRAFT—DO NOT CITE OR QUOTE
-------
1 6.2.2.2 Analysis of Epidemiologic Data
2 The NCI cohort consisted of 25,619 workers employed in 10 plants prior to 1966. A
3 follow-up through 1994 presented exposure-response analyses for 9 NPC deaths, as well as
4 analyses of deaths from other solid cancers (Hauptmann et al., 2004). The most recent follow-up
5 (through 2004; lymphohematopoietic cancers only) analyzed 319 deaths attributed to
6 lymphohematopoietic malignancy from a total of 13,951 deaths (Beane Freeman et al., 2009). A
7 detailed exposure assessment was conducted for each worker, based on exposure estimates for
8 different jobs held and tasks performed (Stewart et al., 1986). Exposure estimates were made
9 using several different metrics—peak exposure, average intensity, cumulative exposure, and
10 duration of exposure. Respirator use and exposures to formaldehyde-containing particulates and
11 other chemicals were also considered. Relative Risks (RRs) were estimated using log-linear
12 Poisson regression models stratified by calendar year, age, sex, and race and adjusted for pay
13 category (salary/wage/unknown). The NCI investigators used the low-exposure category as the
14 reference category to "minimize the impact of any unmeasured confounding variables since
15 nonexposed workers may differ from exposed workers with respect to socioeconomic
16 characteristics."
17 Although other upper respiratory tract cancers were also identified as being causally
18 associated with formaldehyde exposure in the weight-of-evidence analysis in Section 4.5, NPC
19 was the only upper respiratory tract cancer with exposure-response data adequate for the
20 derivation of unit risk estimates in the Hauptmann et al. (2004) follow-up study of solid tumors.
21 Similarly, the weight-of evidence analysis in Section 4.5 concluded that there were causal
22 relationships between formaldehyde exposure and all lymphohematopoietic cancers as a group as
23 well as leukemias as a group (with the strongest evidence for myeloid leukemia). However,
24 from the Beane Freeman et al. (2009) follow-up study of lymphohematopoietic malignancies,
25 only all leukemias combined and Hodgkin lymphoma were judged to have exposure-response
26 data adequate for the derivation of unit risk estimates.
27 For the NPCs, significant trends were observed for the cumulative and peak exposure
28 metrics. The cumulative exposure metric provides a good fit to the data (p trend = 0.029 for all
29 person-years). Since this is generally the preferred metric for quantitative risk assessment for
30 environmental exposure to carcinogens, cumulative exposure is chosen as the exposure metric
31 for the risk estimate calculations for NPC in this assessment. For the latency of solid cancers,
32 including nasopharyngeal tumors, a 15-year lag interval was used by Hauptmann et al. (2004).
33 For the lymphohematopoietic cancers, using the peak exposure metric, statistically
34 significant log-linear trends were observed for all lymphohematopoietic cancers, Hodgkin
35 lymphoma, and leukemia (the latter only when the unexposed person-years were included)
This document is a draft for review purposes only and does not constitute Agency policy.
6-34 DRAFT—DO NOT CITE OR QUOTE
-------
1 (Beane Freeman et al., 2009). Using the average exposure metric, there was a significant trend
2 for Hodgkin lymphoma. Similar results were seen with the cumulative exposure metric,
3 although the trends were not statistically significant, with ^-values slightly greater than 0.05
4 (Hodgkin lymphoma/' trends = 0.06 and 0.08 with and without the unexposed person-years,
5 respectively; leukemia/? trends = 0.08 and 0.12 with and without the unexposed person-years,
6 respectively). For the latency of lymphohematopoietic cancers, a 2-year lag interval was used by
7 Beane Freeman et al. (2009).
8 Although the peak exposure metric provides the most statistically robust dose-response
9 relationship, it is not clear how to extrapolate RR estimates based on the peak exposure estimates
10 to meaningful estimates of lifetime extra risk of cancer from environmental exposures. The
11 average exposure metric is also problematic because it suggests that duration of exposure is not
12 important, i.e., exposure to a given exposure level for one year conveys the same amount of risk
13 as exposure to the same level for 70 years.
14 Cumulative exposure is generally the preferred metric for quantitative risk assessment for
15 environmental exposure to carcinogens. Given the consistency of increased mortality from
16 Hodgkin lymphoma and leukemia overall (exposed versus unexposed) and for each exposure
17 metric (see Table 5-12), indicating risk from these cancers is more than chance, a determination
18 was made that the cumulative exposure results for these two cancer types constituted the best
19 data sets from which to calculate unit risk estimates for lymphohematopoietic cancers from the
20 NCI cohort.
21 Regression coefficients from the NCI log-linear trend test models for the NPCs
22 (Hauptmann et al., 2004) and the various lymphohematopoietic cancers (Beane Freeman et al.,
23 2009) were provided by Drs. Hauptmann and Beane Freeman, respectively. These trend tests
24 were of the form RR = eP*exP°sure xhe coefficients (i.e., P) were used in lifetable analyses to
25 calculate lifetime extra cancer risks from formaldehyde exposure (see Section 5.2). Extra risk
26 estimates for cancer incidence for the three cancer types were approximated by assuming that
27 cancer incidence and cancer mortality have the same dose-response relationships and then using
28 background cause-specific incidence rates instead of mortality rates in the lifetable analysis.
29 Points of departure (PODs) based on the dose-response modeling of these cancers were
30 calculated as the exposure concentration at which the 95% upper confidence bound on extra risk
31 was 0.0005 (i.e., 0.05%) for NPC and for Hodgkin lymphoma and 0.005 (i.e., 0.5%) for
32 leukemia (see Sections 5.2.2 and 5.2.3). These values approximate the lower confidence bounds
33 on dose at these extra risk levels. The values for these extra risk levels, 0.0005 and 0.005, were
34 chosen because they are near the lower end of the observable range of the data. Having such low
35 response levels associated with the points of departure is warranted because of the low
This document is a draft for review purposes only and does not constitute Agency policy.
6-3 5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
background lifetime risks for these cancer types (e.g., 0.00022 for NPC mortality). Higher extra
risk levels would entail extrapolation above the range of the bulk of the observable data to obtain
PODs. The resulting effective concentration values for the selected extra risk values for cancer
incidence are presented in Table 6-2.
Table 6-2. Effective concentrations (lifetime continuous exposure levels) predicted
for specified extra cancer risk levels for selected formaldehyde-related cancers3
Cancer type
Extra risk level
ECb(ppm)
LECC (ppm)
NPC
Hodgkin
lymphoma
Leukemias
0.0005
0.0005
0.005
0.074
0.052
0.16
0.046
0.030
0.088
"calculated including all person-years (see Section 5.2)
beffective concentration.
S95% lower confidence bound on the EC; this value is the POD.
Linear low-dose extrapolation from the PODs was used to derive unit risk estimates for
NPC, Hodgkin lymphoma, and leukemia, as discussed in Section 6.2.2.4. To obtain an
approximate (upper bound) unit risk estimate of the total cancer risk from formaldehyde
exposure, risk estimates for these three cancer types (NPC, Hodgkin lymphoma, and leukemia)
were combined assuming a normal distribution (see Section 5.2.4). This was considered the
most reasonable approach for estimating total cancer risk from the available data; however, it
should be noted that this estimate may not reflect all of the cancer types associated with
formaldehyde exposure.
6.2.2.3 Analysis of Laboratory Animal Data
Various bioassays have been conducted studying the effects of formaldehyde on rats,
mice, and rhesus monkeys and have been discussed at length earlier in this document. Of these,
two inhalation bioassays of rats, when combined, allow for the most robust characterization of
the long-term dose-response relationship in a laboratory species. These long-term bioassays
found an increased incidence of nasal SCCs in rats exposed to formaldehyde by the inhalation
route (Monticello et al., 1996; Kerns et al., 1983). In the combined data, rats were exposed to 0,
0.7, 2.0, 6.0, 9.93, and 14.96 ppm (0, 0.86, 2.5, 7.4, 12.2, and 18.4 mg/m3) exposure
This document is a draft for review purposes only and does not constitute Agency policy.
6-36 DRAFT—DO NOT CITE OR QUOTE
-------
1 concentrations of formaldehyde (Monticello et al. 1996; Kerns et al. 1983). SCCs were observed
2 only at 6 ppm and higher exposure concentrations.
3 A large amount of mechanistic information relevant to the dose-response relationship of
4 formaldehyde in the respiratory tract has been generated either following or in conjunction with
5 these two bioassays, as reviewed in Chapter 3, 4 and 5. This information includes the following:
6
7 1. Measurements of DNA-protein cross-links (DPXs) formed by formaldehyde in F344 rats
8 and rhesus monkeys (Casanova et al., 1989, 1994). Several PBPK models have been
9 developed in the literature based on these data. Some of these efforts integrated the data
10 in both species (Casanova et al., 1991; Conolly et al., 2000; Klein et al., 2010).
11 2. Measurements of cell proliferation in F344 rats and rhesus monkeys (Monticello et al.,
12 1989, 1990, 1991, 1996).
13 3. Simulations of airflow in anatomically realistic representations of the upper respiratory
14 tract of the F344 rat, rhesus monkey and human, and in an idealized representation of the
15 human lower respiratory tract, using computer and physical models (Kimbell et al., 1993,
16 1997a; Kepler et al., 1998; Subramaniam et al., 1998). These simulations were used to
17 predict regional formaldehyde dosimetry in the corresponding sections of the respiratory
18 tract of these three species (Kimbell et al., 2001a, b; Overton et al., 2001).
19
20 The combined nasal tumor incidence data in the two inhalation bioassays (Kerns et al.
21 1983, Monticello et al. 1996) were analyzed using a multistage-weibull time-to-tumor approach
22 as well as models derived from the biologically based dose-response (BBDR) modeling
23 approach in Conolly et al. (2003) [see Crump et al. (2005), Subramaniam et al. (2007), Section
24 5.3, and Appendix E for details]. The BBDR approach enabled integration of the mechanistic
25 information and the time-to-tumor incidence data within a single conceptual framework.
26
27 6.2.2.4.Extmpolation Approaches
28 An EPA inhalation unit risk is developed to estimate cancer risk from environmental
29 exposures or in order to determine exposure levels corresponding with cancer risks as low as
30 1 excess cancer in 10,000 or 1 excess cancer in 1 million. As neither data from animal studies,
31 nor human epidemiological studies, provide direct observation of these low level risks, the
32 observed exposure response relationship is extrapolated to estimate low dose risk. The model
33 used to extrapolate below the range of exposures clearly associated with increased risk of health
34 effects has a great influence on the inhalation unit risk, as there may be several orders of
35 magnitude difference between the observed risk and the target risk range. In the absence of
36 empirical data or a biologically-informed model, the EPA applies a simple straight line
37 extrapolation from the point of departure to zero exposure (U.S. EPA, 2005a). The Mode of
This document is a draft for review purposes only and does not constitute Agency policy.
6-37 DRAFT—DO NOT CITE OR QUOTE
-------
1 Action evaluation reviews available data and determines if an MOA can be sufficiently
2 established and whether it informs the shape of the exposure-response relationship.
O
4 6.2.2.4.1. Low-dose extrapolation for Lymphohematopoietic cancers.
5 Formaldehyde is a mutagen, and known to act directly on cells at the site of first contact.
6 Clastogenic effects have been documented in formaldehyde-exposed workers including
7 peripheral blood lymphocytes and circulating stem cells (Zhang et al., 2010a). Thus a mutagenic
8 MOA has been hypothesized for lymphohematopoeitic cancers, and supports a linear low-dose
9 extrapolation of human cancer risk. Additionally, formaldehyde may also induce some form of
10 bone marrow toxicity, as suggested by observed pancytopenia in exposed workers (Tang et al.,
11 2008, Zhang et al., 2010b). However, as the mechanism of transport to the bone marrow, and
12 biological activity leading to the observed toxicity are unknown, this information does not
13 inform the low-dose extrapolation. Although the mechanisms underlying formaldehyde-induced
14 leukemia and lymphoma are still largely speculative, there is little doubt of an association
15 between formaldehyde exposures and lymphohematopoeitic cancer mortality, especially for
16 myeloid leukemia. Therefore, without a known MOA which would justify an alternative
17 approach, and with a hypothesized mutagenic MOA under consideration which supports a simple
18 straight line extrapolation from the point of departure to zero risk at zero exposure, this is
19 applied when estimating human cancer risk from both leukemia and Hodgkin lymphoma from
20 the NCI cohort.
21
22 6.2.2.4.2. Low-dose extrapolation for cancer of the upper respiratory tract.
23 There are multiple plausible MO As for formaldehyde carcinogenesis regarding upper
24 respiratory tract cancers (see Section 4.5.3), however they may not be necessarily relevant to
25 describing the lower end of the exposure response curve. For example, although regenerative
26 cell proliferation associated with focal and gross tissue lesions due to cell death may contribute
27 to the high incidence of rat nasal tumor in F344 rats, these mechanisms may not be operative in
28 the low exposure region expected for human environmental exposure (e.g., less than Ippm) and
29 therefore may not inform low-dose extrapolation. There are MO As which are more appropriate
30 to the low-dose region. Specifically, formaldehyde is a known mutagen, may inhibit DNA repair
31 activity and may have additional activity as a tumor promoter. Finally, other effects such as
32 formaldehyde-induced cell proliferation, immunosuppression and disruption of the mucociliary
33 apparatus may influence both the level of tissue damage and ultimately cancer incidence.
34 EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) recommend using
35 biologically based dose-response (BBDR) models for extrapolation when data permit. Conolly
This document is a draft for review purposes only and does not constitute Agency policy.
6-3 8 DRAFT—DO NOT CITE OR QUOTE
-------
1 et al. (2003, 2004) developed BBDR models to predict squamous cell carcinoma risk in the rat
2 and human respiratory tract at exposures well below the range of the observed animal data. The
3 primary conclusion from their modeling effort was that human exposure standards protective of
4 effects of formaldehyde-induced cytotoxicity should be sufficient to protect from the potential
5 carcinogenic effects of formaldehyde. The authors assessed that such a conclusion was
6 conservative in the face of model uncertainties.17 The current assessment evaluated the
7 uncertainties and alternative parameterizations of the modeling in Conolly et al. (2003, 2004)
8 extensively, and concluded that the human cancer risk values computed by Conolly et al. (2004)
9 were not conservative estimates. Models resulting from alternative parametrizations were as
10 consistent with the experimental data as the original model but resulted in maximum likelihood
11 estimates of added human risk that ranged from negative to large positive values at
12 environmental exposure concentrations. Model uncertainty far exceeded statistical uncertainty
13 (see Table E-4 in Appendix E). Each of these models, including the modeling in Conolly et al.,
14
15 1. was judged to be just as biologically plausible given the available data,
16 2. described the rat tumor incidence data equally well,
17 3. was based on different characterizations of the same empirical cell kinetic data, and
18 4. was based on the same empirical data on DPX measurements.
19
20 This assessment's evaluation18 (detailed in Section 5.3) of the above models concluded
21 that these models, including alternative implementations of those in Conolly et al. (2003, 2004),
22 were too uncertain to be useful for low-dose extrapolation of risk. In particular:
23
24 • When used for the purpose of extrapolating risk, the BBDR models did not appear to
25 reasonably constrain either
26 > risk estimates extrapolated from the F344 rat to the human, regardless of whether the
27 extrapolation was carried out at low or comparable exposures, or
28 > risk estimates for the F344 rat when extrapolated outside the range of observable
29 data.
17 Based on their modeling, Conolly et al. (2003, 2004) concluded that the directly mutagenic action of formaldehyde
does not play a significant role in formaldehyde carcinogenicity. Respiratory cancer risks associated with inhaled
formaldehyde were predicted to be de minimis (10~6 or less) at relevant human exposure levels when an upper bound
on the model estimate for the directly mutagenic action of formaldehyde was used.
18 The scope of this evaluation was informed by views provided by several experts convened by EPA in October
2004. The participants were Drs. Rory Conolly, Kenny Crump, Linda Hanna, Dale Hattis, Julia Kimbell, George
Lucier, Christopher Portier and Fred Miller (guest participant). The meeting agenda and summary are provided in
Appendix H.
This document is a draft for review purposes only and does not constitute Agency policy.
6-39 DRAFT—DO NOT CITE OR QUOTE
-------
1 • Human risk calculated from these BBDR models was numerically unstable when certain
2 parameter conditions were realized (see Section 5.3.3 and Appendix F).
3
4 It may be noted that the sensitivity analyses on the basis of which these conclusions were
5 reached have been criticized as resulting in implausible risk estimates (given the epidemiologic
6 data) as a consequence of implementing model variations that are not biologically reasonable
7 (Conolly et al. 2009). This criticism was rebutted by Crump et al. (2009) on biological and
8 epidemiological grounds. These debates have been discussed fully in Appendix F.
9 However, using the BBDR model to characterize the dose-response in the range of the
10 available data has the advantage of utilizing the available biological and dosimetry data on
11 formaldehyde in an integrated manner as well as providing statistically sound descriptions of the
12 empirical tumor incidence data. Therefore, this assessment uses the BBDR modeling of the rat
13 data to derive multiple PODs (for SCC in the respiratory tract) in the range of the observed data
14 and uses model-derived internal dose estimates. For the reasons detailed above, the BBDR
15 modeling is not used to extrapolate far below the observed data.
16 The lowest observed incidence of SCC in the bioassays used in the dose-response
17 assessment was equal to 0.0087 (at 6 ppm exposure). In addition, the BBDR modeling of the
18 tumor data was informed by its use of data on cell proliferation and formation of DPXs at the
19 lower exposure concentrations of 0.7 and 2.0 ppm. Thus, the available data supported estimation
20 of response levels below the 10% response level commonly used in BMD analyses of tumor
21 data. Therefore, points of departure corresponding to 95% statistical upper bound levels of extra
22 risk of 0.005, 0.01 and 0.1 were estimated when the BBDR modeling was used.
23
24 6.2.2.4.3. Summary.
25 As discussed earlier in the hazard characterization, formaldehyde is a direct-acting
26 mutagen, and its genotoxic effects have been observed following human occupational
27 exposures.19 Furthermore, a low-dose nonlinear MOA for formaldehyde-induced
28 lymphohematopoietic cancers, NPCs, or cancers in other regions of the respiratory tract has not
29 been established. In particular, the formation of DPXs by formaldehyde, considered a dose
30 surrogate for the molecular dose associated with formaldehyde's mutagenic action, has been
31 observed at doses well below those considered cytotoxic. Therefore, linear low-dose
32 extrapolation from the suitably chosen PODs was considered most appropriate for all the cancers
19 While formaldehyde may also contribute to mutations indirectly, such an effect is likely to be relevant only at the
higher doses.
This document is a draft for review purposes only and does not constitute Agency policy.
6-40 DRAFT—DO NOT CITE OR QUOTE
-------
1 (whether the PODs were based on epidemiological data or rodent bioassay data), which is also in
2 accordance with EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a).
3
4 6.2.2.5. Inhalation Unit Risk Estimates for Cancer
5 The epidemiological and rodent inhalation data indicate multiple sites of concern. Unit
6 risk estimates calculated separately from these data are summarized in Table 6-3.
7 As can be seen in the Table 6-3, the unit risk estimate based on human data for NPC is in
8 the range of the estimates calculated for respiratory tract cancer from the rodent nasal cancer
9 data. Experimental animal data were inadequate for estimating risk of lymphohematopoietic
10 cancers. The unit risk estimate for Hodgkin lymphoma is also in the same range, while the unit
11 risk estimate for leukemia and the total cancer unit risk estimate are up to 4-fold higher.
12 As documented in EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
13 when high-quality human data are available, they are generally preferred over laboratory animal
14 data for quantitative risk assessment. Thus, the preferred (plausible upper bound) unit risk
15 estimate in this assessment is the value of 8.1 x 10~2 per ppm (6.6 x 10~5 per ug/m3) based on
16 (adult) human data for NPC, Hodgkin lymphoma, and leukemia. Note that, as discussed in
17 Section 6.2.2.6 below, if there is early-life exposure, the age-dependent adjustment factors
18 (AD AFs) should be applied, in accordance with EPA's Supplemental Guidance for Assessing
19 Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b).
20
21 6.2.2.6.Earfy-Life Susceptibility
22 There are no chemical-specific data for quantitatively addressing the susceptibility of different
23 life stages to carcinogenicity from inhalation exposure to formaldehyde. As documented in
24 Section 4.5, formaldehyde is a mutagenic carcinogen and the weight of evidence supports the
25 conclusion that formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic
26 MOA. Therefore, increased early-life susceptibility should be assumed and, if there is early-life
27 exposure, the AD AFs should be applied, in accordance with EPA's Supplemental Guidance for
28 Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b). See
29 Section 5.4.4 for details on the application of the AD AFs.
30 Accordingly, for full lifetime exposures, the overall (plausible upper bound) unit risk
31 estimate is 0.13 per ppm (1.1 x 10~4 per ug/m3) for the three cancer types (NPC, Hodgkin
32 lymphoma, and leukemia) combined (see Table 5-26 for calculations).
33
34 Table 6-3. Inhalation unit risk estimates based on epidemiological and
35 experimental animal data
This document is a draft for review purposes only and does not constitute Agency policy.
6-41 DRAFT—DO NOT CITE OR QUOTE
-------
Cancer type3
Dose metric
Unit Risk Estimate1"
(ppnT1)
Based on epidemiological data
Nasopharyngeal
Hodgkin lymphoma
Leukemia
Cumulative exposure
Cumulative exposure
Cumulative exposure
0.011
0.017
0.057
All three cancer sites combined: 0.081°
Based on Experimental Animal Data
Squamous cell carcinoma of the
respiratory tract
Local dose (flux) of formaldehyde
in pmol/mm2/hour
0.011-0.022d
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
athe unit risk estimates are all for cancer incidence.
b these unit risk estimates do not include ADAFs (see Section 6.2.2.6 below).
°this total cancer unit risk estimate is an estimate of the upper bound on the sum of risk estimates calculated for the 3
individual cancer types (nasopharyngeal cancer, Hodgkin lymphoma, and leukemia); it is not the sum of the
individual (upper bound) unit risk estimates (see Section 5.2.4).
"Values are similar to estimates from Schlosser et al. (2003). These authors determined their PODs based on tumor
and cell proliferation as endpoints, and extrapolated benchmark exposure concentrations to humans using
formaldehyde flux to the tissue and DPX concentrations as internal dose metrics.
6.2.2.1 .Uncertainties in the Quantitative Risk Estimates
Uncertainties in the risk estimates based on the human data are discussed in detail in
Sections 5.2.2.4 and 5.2.3.4. Major uncertainties inherent in the NPC, Hodgkin lymphoma, and
leukemia risk estimates are
• the retrospective exposure estimation,
• the appropriateness of the dose-response model and exposure metric, and
• the extrapolation from occupational exposures to lower environmental exposures.
In addition, the NPC and Hodgkin lymphoma estimates are limited by the sparse data for these
cancers in the NCI cohort study (estimates are based on the exposure-response modeling of only
9 NPC deaths and 27 Hodgkin lymphoma deaths).
Of note, Marsh et al. (2002, 1996) independently studied one of the 10 plants that was in
the NCI study, and there were large differences in the exposure estimates for that plant from the
two different studies. If the exposure estimates of Marsh et al. (2002) are closer to the true
This document is a draft for review purposes only and does not constitute Agency policy.
6-42 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposures, then the potency of formaldehyde could be greater than reflected in the risk estimates
2 derived from the NCI data.
3 The linear low-dose extrapolation (see Section 6.2.2.4) from the 95% lower bound on the
4 exposure level associated with the benchmark response is generally considered to provide a
5 plausible upper bound on the risk at lower exposure levels. Although the linear low-dose
6 extrapolation used here is supported by the mutagenicity of formaldehyde, nonlinearities in the
7 exposure-response relationship may be present below, as well as above, the POD. The strong
8 association with peak exposures for all 3 cancer types in the NCI study suggests that dose-rate
9 effects may be operative (i.e., the risk from peak occupational exposures may be greater than the
10 [linearly] proportional risks from lower exposures and, similarly, the risk from an occupational
11 cumulative exposure may be greater than the proportional risk from a lower environmental
12 cumulative exposure).20 Any such dose-rate effects would not be reflected in the cumulative
13 exposure metric used for the exposure-response modeling in the range of the occupational
14 exposures nor in the linear low-dose extrapolation approach used in this assessment. Actual
15 low-dose risks may be lower to an unknown extent.
16 Other significant uncertainties may also remain. For example, risk estimates could not be
17 derived from the NCI cohort study for rare upper respiratory tract cancers other than NPC. In
18 addition, although unit risk estimates were derived for Hodgkin lymphoma and leukemia because
19 they exhibited the strongest trend results of the lymphohematopoietic cancers using the
20 cumulative exposure metric, it is uncertain which specific lymphohematopoietic cancer subtypes
21 are associated with formaldehyde exposure. Furthermore, the potential role of particulates in the
22 NPC risk is unclear. Moreover, as for all occupational epidemiology studies, there is uncertainty
23 in extrapolating risk from an adult worker population (in this case predominantly white males) to
24 the more diverse general population.
25 Despite inevitable uncertainties, it is important not to lose sight of the strengths of the
26 estimates, which are based on human data from a high-quality NCI study. In addition to the use
27 of internal analyses and the extensive exposure assessment and consideration of potential
28 confounding or modifying variables, the NCI study has a large cohort that has been followed for
29 a long time. With the additional follow-up through 2004, reflected in the lymphohematopoietic
30 cancer results of Beane Freeman et al. (2009), the median duration of follow-up was 42 years,
31 and the 25,619 cohort members had accrued 998,106 person-years of follow-up.
20 Dose-rate effects are also suggested by the very steep, nonlinear exposure-response relationships observed in the
rodent cancer bioassays, although, in the rodents, this steep increase in tumor incidence at high exposures is thought
to be due to the contribution of cytotoxicity and regenerative proliferation, which is not apparent with the human
exposures (Section 4.5).
This document is a draft for review purposes only and does not constitute Agency policy.
6-43 DRAFT—DO NOT CITE OR QUOTE
-------
1 Significant uncertainties also exist in the risk estimates derived from the rodent bioassay
2 data. In general, the difficulties in extrapolating from experimental animal bioassays are
3 considerable, and the use of human data is preferred, while recognizing the different
4 uncertainties that are present in risk estimates based on epidemiological data.
5 In the case of formaldehyde, this general uncertainty associated with extrapolation from
6 rodent data is increased due to the highly curvilinear nature of the dose-response relationships
7 associated with DPX formation, labeling index data, and tumor responses. The mechanistic
8 interpretation of these observed data has provided grounds for arguments in the literature that
9 formaldehyde tumorigenicity (at exposures > 6 ppm) should be uncoupled from its potential
10 carcinogenicity in the low-dose region.
11 Quantitative models have been used in the literature to further argue that the observed
12 risk in animal experiments is entirely due to cell proliferation induced by regenerative
13 hyperplasia in response to cell injury at cytotoxic doses, i.e., without a relevant role for the direct
14 mutagenic action of formaldehyde. In the context of using these data for quantitative risk
15 assessment, this document notes that such an inference of the data has been found to be
16 extremely uncertain. A quantitative analysis of the uncertainties in interpreting the available
17 data has shown that the directly mutagenic action of formaldehyde could be very important in
18 explaining the high-dose effect (Subramaniam et al., 2007).
19 While acknowledging these substantial difficulties, the quantitative dose-response
20 modeling of the rat data does allow inference about upper bound risks for respiratory cancer,
21 consistent with the observed experimental tumorigenicity. These upper bound risk estimates are
22 consistent with those estimated from the epidemiological data; however, such a consistency may
23 be entirely artifactual. As noted earlier, the BBDR modeling helped characterize some of the
24 uncertainty associated with extrapolating from the rodent data to the environmental risk in
25 people. The actual risk may be substantially lower or higher than the reasonable upper bound
26 risk estimated from the animal data.
27
28 6.2.2.8.Conclusions
29 Cancer unit risk estimates for formaldehyde inhalation exposure were derived from both
30 human and laboratory animal data. As documented in EPA's Guidelines for Carcinogen Risk
31 Assessment (U.S. EPA, 2005a), when high-quality human data are available, they are generally
32 preferred over laboratory animal data for quantitative risk assessment. Thus, the preferred unit
33 risk estimate in this assessment is based on human data for NPC, Hodgkin lymphoma, and
34 leukemia from a high-quality NCI occupational cohort study (Hauptmann et al., 2004; Beane
35 Freeman et al., 2009). (The qualitative hazard assessment suggests causal associations between
This document is a draft for review purposes only and does not constitute Agency policy.
6-44 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde exposure and other cancer types as well [e.g., other upper respiratory tract cancers
2 and possibly other lymphohematopoietic cancers; see Section 4.5], but quantitative data from the
3 NCI cohort study were not amenable for deriving quantitative risk estimates for those cancer
4 types. Because there were not clear exposure-response data for these cancer types in that cohort
5 study [based on cumulative exposure], any contributions to the total cancer risk from
6 environmental formaldehyde exposure for these cancers are not expected to be large; however,
7 this is a source of uncertainty.)
8 The unit risk estimate for the total cancer incidence extra risk for these three cancer types
9 combined based on the (adult) human data is 8.1 x 1CT2 per ppm (6.6 x 1CT5 per ug/m3). As
10 documented in Section 4.5, formaldehyde is a mutagenic carcinogen and the weight of evidence
11 supports the conclusion that formaldehyde carcinogenicity can be attributed, at least in part, to a
12 mutagenic MO A. Therefore, as there are no chemical-specific inhalation data on cancer
13 susceptibility at different life-stages, increased early-life susceptibility is assumed and ADAFs
14 should be applied in accordance with EPA's Supplemental Guidance for Assessing Susceptibility
15 from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b). Applying the ADAFs, the overall
16 (upper bound) unit risk estimate for full lifetime exposure is 0.13 per ppm (1.1 x 1CT4 per |ig/m3)
17 for the three cancer types (NPC, Hodgkin lymphoma, and leukemia) combined. Using this
18 lifetime unit risk estimate, the upper bound estimate of the cancer risk at the RfC of 1 ppb is 1 x
19 1(T4.
20
21 6.3. SUMMARY AND CONCLUSIONS
22 Seven different noncancer health effects were identified from formaldehyde inhalation
23 exposure studies, including: 1) sensory irritation of the eyes, nose, and throat, 2) upper
24 respiratory tract pathology, 3) pulmonary function, 4) asthma and atopy, 5) neurologic and
25 behavioral toxicity, 6) reproductive and developmental toxicity, and 7) immunological toxicity.
26 Of note, epidemiological evidence is available for most of these noncancer effects. EPA has
27 derived candidate RfCs for critical effects based on seven key studies. Three cocritical studies
28 were selected which provide similar cRfCs for related adverse health effects observed in
29 residential populations including children i.e., increased asthma incidence, decreased pulmonary
30 function, increase in respiratory symptoms, and increased allergic sensitization (Rumchev et al.,
31 2002; Garrett et al., 1999 a,b; Krzyzanowski et al., 1999). The resulting cRfCs fall in a range
32 between 2.8 and 11 ppb, depending on the study, or endpoints considered, and the application of
33 alternative uncertainty factors for human variability (see Table 6-1). The representative RfC for
34 the cocritical studies is taken as the average of the cRfCs (see Section 6.2.1.2).
This document is a draft for review purposes only and does not constitute Agency policy.
6-45 DRAFT—DO NOT CITE OR QUOTE
-------
1 EPA has assessed the adequacy of the overall database for RfC derivation, and although
2 the database is quite large, and provides significant information on well studied POE effects.
3 There are remaining uncertainties in the database. Most notably, there is a need for additional
4 exposure-response information for observed neurotoxic effects, reproductive and developmental
5 effects as well as a two-generation study to evaluate the effects of formaldehyde exposure on
6 reproductive and developmental endpoints. EPA is considering 4 options to address database
7 uncertainties in the final RfC (see Section 6.2.1.3). It is unclear what uncertainty factors are
8 appropriate to account for human variability and deficiencies in the overall database. For this
9 reason, several alternatives have been presented. EPA is seeking advice from the NAS and the
10 public on this matter.
11 Formaldehyde is Carcinogenic to Humans by the Inhalation Route of Exposure. Human
12 epidemiological evidence is sufficient to conclude a causal association between formaldehyde
13 exposure and nasopharyngeal cancer, nasal and paranasal cancer, all leukemias, myeloid
14 leukemia and lymphohematopoietic cancers as a group. Epidemiological evidence is also
15 strongly supportive of, but in itself not sufficient for, a conclusion of causal association for other
16 upper-respiratory tract cancers, Hodgkins lymphoma, or multiple myeloma. Animal bioassays
17 consistently demonstrate formaldehyde-induced nasal cancers in rodents which provide strong
18 support for the observed upper respiratory tract cancers in humans. Limited evidence from
19 animal bioassays is available to support the conclusion from human epidemiologic data that
20 formaldehyde causes some types of lymphohematopoietic cancers.
21 The (upper bound) unit risk estimate for the total cancer incidence based on (adult)
22 human data is 8.1 x 1CT2 per ppm (6.6 x 1CT5 per ug/m3). Applying the age-dependent
23 adjustment factors for increased early-life susceptibility, the overall combined cancer unit risk
24 estimate for full lifetime exposure is 0.13 per ppm (1.1 x 10^ per |ig/m3).
25
26
This document is a draft for review purposes only and does not constitute Agency policy.
6-46 DRAFT—DO NOT CITE OR QUOTE
-------
REFERENCES
Abbas, AK; Fausto, N; Kumar, V. (2004) Robbins and Cotran pathologic basis of disease. W.B. Saunders Co.
Silver Spring, MD.
ACGIH (American Conference of Governmental Industrial Hygienists). (2002) Documentation of the TLVs® and
BEIs® with other worldwide occupational exposure values CD-ROM—2002. American Conference of
Governmental Industrial Hygienists, Cincinnati, OH.
Acheson, ED; Barnes, HR; Gardener, MJ; et al. (1984) Formaldehyde in the British chemical industry: an
occupational cohort study. Lancet 1:611-616.
Achkor, H; Diaz, M; Fernandez, MR; et al. (2003) Enhanced formaldehyde detoxification by overexpression of
glutathione-dependent formaldehyde dehydrogenase from Arabidopsis. Plant Physiol 132:2248-2255.
Adams, DO; Hamilton, TA; Lauer, LD; et al. (1987) The effect of formaldehyde exposure upon the mononuclear
phagocyte system of mice. Toxicol Appl Pharmacol 88:165-174.
Ahmad, I; Whitson, TC. (1973) Formaldehyde: how much of a hazard? Ind Med Surg 42:26-27.
Ahmed, S; Tsukahara, S; Tin-Tin-Tin-Shwe; et al. (2007) Effects of low-level formaldehyde exposure on synaptic
plasticity-related gene expression in the hippocampus of immunized mice. J. Neuroimmunol 186:104-111.
Akbar-Khanzadeh, F; Mlynek, JS. (1997) Changes in respiratory function after one and three hours of exposure to
formaldehyde in non-smoking subjects. Occup Environ Med 54:296-300.
Akbar-Khanzadeh, F; Vaquerano, MU; Akbar-Khanzadeh, M; et al. (1994) Formaldehyde exposure, acute
pulmonary response, and exposure control options in a gross anatomy laboratory. Am J Ind Med 26:61-75.
Alarie, Y. (1981) lexicological evaluation of airborne chemical irritants and allergens using respiratory reflex
reactions. In: Leong, B, ed. Proceedings of the inhalation toxicology and technology symposium. Ann Arbor, MI:
Ann Arbor Science Publishers, Inc.; pp. 207-231.
Albert, RE; Pasternack, BS; Shore, RE; et al. (1975) Mortality patterns among workers exposed to chloromethyl
ethers-a preliminary report. Environ Health Perspect 11:209-214.
Albert, RE; Sellakumar, AR; Laskin, S; et al. (1982) Gaseous formaldehyde and hydrogen chloride induction of
nasal cancer in the rat. J Natl Cancer Inst 68:597-603.
Alderson, T. (1985) Formaldehyde-induced mutagenesis: a novel mechanism for its action. Mutat Res 154:101-10.
Alexandersson, R; Hedenstierna, G. (1988) Respiratory hazards associated with exposure to formaldehyde and
solvents in acid-curing paints. Arch Environ Health 43:222-227'.
Alexandersson, R; Hedenstierna, G. (1989) Pulmonary function in wood workers exposed to formaldehyde: A
prospective study. Arch Environ Health 44:5-11.
Alexandersson, R; Hedenstierna, G; Kolmodin-Hedman, B. (1982) Exposure to formaldehyde effects on pulmonary
function. Arch Environ Health 37:279-284.
Almaguer, D; Blade, LM. (1990) Health hazard evaluation report: Buckeye Hills Career Center, Rio Grande, Ohio.
National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services, Cincinnati,
OH; HETA 88-153-2072
This document is a draft for review purposes only and does not constitute Agency policy.
R. i DRAFT—DO NOT CITE OR QUOTE
-------
Almaguer, D; Klein, M. (1991) Health hazard evaluation report: Northwest Vocational School, Cincinnati, Ohio.
National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services, Cincinnati,
OH; HETA 89-170-2100
Amdur, MO. (1960) The response of guinea pigs to inhalation of formaldehyde and formic acid alone and with a
sodium chloride aerosol. Int J Air Pollut 3:201-220.
Ames, BN; Shigenaga, MK; Gold, LS. (1993) DNA lesions, inducible DNA repair, and cell division: three key
factors in mutagenesis and carcinogenesis. Environ Health Perspect 101(Suppl 5):35^4.
Andersen, I; Lundqvist, GR. (1974) Design and performance of an environmental chamber. Int J Biometeor 14:402-
405.
Andersen, I; Melhave, L. (1983) Controlled human studies with formaldehyde. In: Gibson, JE; ed. Formaldehyde
toxicity. Washington, DC: Hemisphere Publishing Corporation; pp. 154-165.
Andersen, ME; Jarabek, AM. (2001) Nasal tissue dosimetry. Issues and approaches for "Category 1" gases: a report
on a meeting held in Research Triangle Park, NC, February 11-12, 1998. Inhal Toxicol 13:415-435.
Andersen, ME; Clewell III, HJ, Bermudez, E; et al. (2008) Genomic signatures and dose dependent transitions in
nasal epithelial responses to inhaled formaldehyde in the rat. Toxicol Sci 105:368-383.
Andersen, I (1979) Indoor climate: Effects on human comfort, performance, and health in residential, commercial,
and light-industry buildings. In: Fanger, P.O. and Valbjern; ed. Proceedings of the first international indoor climate
symposium in Copenhagen Aguust 30-September 1, 1978. Danish Building Research Institute; Copenhagen; pp. 65-
87.
Anderson, H; Dally, KA; Hanrahan, L; et al. (1983) The epidemiology of mobile home formaldehyde vapor
concentration and residents' health status. Chicago, IL: Region V. Available from NTIS, PB84-123504. EPA/905/1-
83-001.
Andersson, M; Agurell, E; Vaghef, H, et al. (2003) Extended-term cultures of human T-lymphocytes and the comet
assay: a useful combination when testing for genotoxicity in vitro? Mutat Res 540:43-55.
Anderson, S; Jensen, O; Oliva, D. (1982) Formaldehyde exposure and lung cancer in Danish doctors. Ugeskr Laeg
144:1571-1573.
Andjelkovich, DA; Janszen, DB; Brown, MH; et al. (1995) Mortality of iron foundry workers: IV. Analysis of a
subcohort exposed to formaldehyde. J Occup Environ Med 37:826-837.
Ang, HL; Deltour, L; Zgombic-Knight, M; et al. (1996a) Expression patterns of class I and class IV alcohol
dehydrogenase genes in developing epithelia suggest a role for alcohol dehydrogenase in local retinoic acid
synthesis. Alcohol Clin Exp Res 20:1050-1064.
Ang, HL; Deltour, L; Hayamizu, TF; et al. (1996b) Retinoic acid synthesis in mouse embryos during gastrulation
and craniofacial development inked to class IV alcohol dehydrogenase gene expression. J Biol Chem
271:9526-9534.
Apfelbach, R; Reibenspies, M. (1991) Olfactory learning is inhibited after low-level formaldehyde gas exposure in
the ferret. Chem Senses 16:498
Apfelbach, R; Weiler, E. (1991) Sensitivity to odors in Wistar rats is reduced after low-level formaldehyde-gas
exposure. Naturwissenschaften 78:221-223.
Apfelbach, R; Reibenspies, M; Schmidt, R. (1992) The effect of long-lasting formaldehyde gas exposure on the
olfactory epithelium and on the olfactory discrimination ability in the ferret. Chem Senses 17:590-591.
This document is a draft for review purposes only and does not constitute Agency policy.
R-2 DRAFT—DO NOT CITE OR QUOTE
-------
Appelman, LM; Woutersen, RA; Zwart, A; et al. (1988) One-year inhalation toxicity study of formaldehyde in male
rats with a damaged or undamaged nasal mucosa. J Appl Toxicol 8:85-90.
Armstrong, RW; Imrey, PB; Lye, MS; et al. (2000) Nasopharyngeal carcinoma in Malaysian Chinese: occupational
exposures to particles, formaldehyde and heat. Int J Epidemiol 29:991-8.
Arts, JH; Droge, SC; Spanhaak, S; et al. (1997) Local lymph node activation and IgE responses in brown Norway
and Wistar rats after dermal application of sensitizing and non-sensitizing chemicals. Toxicology 117:229-234.
Ashby, J; Callander, RD; and Rose, FL. (1985) Weak mutagenicity to Salmonella of the formaldehyde-releasing
anti-tumour agent hexamethylmelamine. MutatRes 142:121-5.
Asian, H; Songur, A; Tune, AT; et al. (2006) Effects of formaldehyde exposure on granule cell number and volume
of dentate gyrus: A histopathological and stereological study. Brain Res 1122:191-200.
ASTM (American Society for Testing and Materials). (2000) Standard test method for estimating sensory irritancy
of airborne chemicals. Annual book of ASTM standards. Designation E981 (reapproved 2000). Philadelphia, PA:
American Society for Testing of Materials.
ATC (American Thoracic Society). (2000) What constitutes an adverse health effect of air pollution? Am J Respir
Crit Care Med 161:665-673.
ATSDR (Agency for Toxic Substances and Disease Registry). (1999) Toxicological profile for formaldehyde.
Public Health Service, U.S. Department of Health and Human Services, Atlanta, GA. Available online at
http://www.atsdr.cdc.gov/toxprofiles/tpl 11.html.
ATSDR, 2007. An update and revision of ATSDR's February 2007 health consultation: Formaldehyde sampling of
FEMA temporary-housing trailers; Baton Rouge, Louisiana September-October, 2006. Public Health Service, U.S.
Department of Health and Human Services, Atlanta, GA. Available online at
www.atsdr.cdc.gov/substances/formaldehy de/public_assessment.html.
Auerbach, C; Moutschen-Dahmen, M; Moutschen, J. (1977) Genetic and cytogenetical effects of formaldehyde and
related compounds. MutatRes 39:317-361.
Axelsson, G; Liitz, C; Rylander, R. (1984) Exposure to solvents and outcome of pregnancy in university laboratory
employees. BrJInd Med 41:305-312.
Baba, K; Yagi, T; Niwa, S; et al. (2000) Measurement of formaldehyde-specific IgE antibodies in adult asthmatics.
Arerugi49(5):404-411.
Babiuk, C; Steinhagen, WH; Barrow, CS. (1985) Sensory irritation response to inhaled aldehydes after
formaldehyde pretreatment. Toxicol Appl Pharmacol 79:143-149.
Bach, B; Pedersen, OF; Melhave, L. (1990) Human performance during experimental formaldehyde exposure.
Environ Int 16:105-114.
Baird, DD; Wilcox, AJ. (1985) Cigarette smoking associated with delayed conception. J Am Med Assoc
253:2979-2983.
Baker, H; Kawano, T; Margolis, FL; et al. (1983) Transneuronal regulation of tyrosine hydroxylase expression in
olfactory bulb of mouse and rat. J Neurosci 3:69-78.
Ballarin, C; Sarto, F; Giacomelli, L; et al. (1992) Micronucleated cells in nasal mucosa of formaldehyde-exposed
workers. MutatRes 280:1-7.
Barker, S; Weinfeld, M; Murray, D. (2005) DNA-protein crosslinks: Their induction, repair, and biological
consequences. MutatRes 589:111-135.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 DRAFT—DO NOT CITE OR QUOTE
-------
Barrow, CS; Steinhagen, WH; Chang, JCF. (1983) Formaldehyde sensory irritation. In: Gibson, JE; ed.
Formaldehyde toxicity. Washington, DC: Hemisphere Publishing Corporation; pp. 16-25.
Bartnik, FG; Gloxhuber, C; Zimmermann, V. (1985) Percutaneous absorption of formaldehyde in rats. Toxicol Lett
25:167-172.
Basecke, J; Griesinger, F; Trumper, L; et al. (2002) Leukemia- and lymphoma-associated genetic aberrations in
healthy individuals. AnnHematol 81:64-75.
Easier, A; v d Hude, W; and Scheutwinkel-Reich, M. (1985) Formaldehyde-induced sister chromatid exchanges in
vitro and the influence of the exogenous metabolizing systems S9 mix and primary rat hepatocytes. Arch Toxicol
58:10-3.
Bateson, TF; Schwartz, J. (2008) Children's response to air pollutants. J Toxicol Environ Health A 71:23 8-243.
Battelle Columbus Laboratories. (1981) Final report on a chronic inhalation toxicology study in rats and mice
exposed to formaldehyde. Prepared by Battelle Columbus Laboratories, Columbus, OH for the Chemical Industry
Institute of Toxicology (CUT), Research Triangle Park, NC; CUT Docket # 10922.
Bauchinger, M; Schmid, E. (1985) Cytogenetic effects in lymphocytes of formaldehyde workers of a paper factory.
MutatRes 158:195-199.
Beane-Freeman, LE; Blair, A; Lubin, JH; et al. (2009) Mortality from lymphohematopoietic malignancies among
workers in formaldehyde industries: the National Cancer Institute cohort. J Natl Cancer Inst 101:751-761.
Beland, FA; Fullerton, NF; Heflich, RH. (1984) Rapid isolation, hydrolysis and chromatography of formaldehyde-
modified DNA. J Chromatogr 308:121-131.
Bender, JR; Mullin, LS; Graepel, GJ; et al. (1983) Eye irritation response of humans to formaldehyde. Am Ind Hyg
Assoc J 44:463-465.
Benedetti, MS; Whomsley, R; Canning, M. (2007) Drug metabolism in the paediatric population and in the elderly.
Drug Discov Today 12:599-610.
Berke, JH. (1987) Cytologic examination of the nasal mucosa in formaldehyde-exposed workers. J Occup Med
29:681-684.
Bermudez, E; Allen, PF. (1984) The assessment of DNA damage and repair in rat nasal epithelial cells.
Carcinogenesis 5:1453-1458.
Bermudez, E; Chen, Z; Gross, EA, et al. (1994) Characterization of cell lines derived from formaldehyde-induced
nasal tumors in rats. Mol Carcinog 9:193-199.
Bermudez, E; Delehanty, LL. (1986) The effects of in vitro formaldehyde treatment on the cells of the rat nasal
epithelium. Environ Mutagen 8 (Suppl.6): 11.
Bertazzi, PA; Pesatori, AC; Radice, L; et al. (1986) Exposure to formaldehyde and cancer mortality in a cohort of
workers producing resins. Scand J Work Environ Health 12:461-468.
Bertazzi, PA; Pesatori, A; Guercilena, S; et al. (1989) Carcinogenic risk for resin producers exposed to
formaldehyde: extension of follow-up. Med Lav 80:111-122.
Bhalla, DK; Mahavni, V; Nguyen, T; et al. (1991) Effects of acute exposure to formaldehyde on surface morphology
of nasal epithelia in rats. J Toxicol Environ Health 33:171-188.
Biagini, RE; Moorman, WJ; Knecht, EA; et al. (1989) Acute airway narrowing in monkeys from challenge with 2.5
ppm formaldehyde generated from formalin. Arch Environ Health 44:12-17.
This document is a draft for review purposes only and does not constitute Agency policy.
R-4 DRAFT—DO NOT CITE OR QUOTE
-------
Bickis, M; Krewski, D. (1989) Statistical issues in the analysis of the long-term carcinogenicity bioassay in small
rodents: An empirical evaluation of statistical decision rules. Fundam Appl Toxicol 12:202-221.
Binzak, BA; Vockley, JG; Jenkins, RB; et al. (2000) Structure and analysis of the human dimethylglycine
dehydrogenase gene. Mol Genet Metab 69:181-187.
Black, KA; Eells, JT; Noker, PE; et al. (1985) Role of hepatic tetrahydrofolate in the species difference in methanol
toxicity. Proc Natl Acad Sci USA 82:3854-3858.
Blackburn, GR; Dooley, JFI; Schreiner, CA; etal. (1991) Specific identification of formaldehyde-mediated
mutagenicity using the mouse lymphoma L5178Y TK +/- assay supplemented with formaldehyde dehydrogenase. In
Vitro Toxicol 4:121-132.
Blair, A; Stewart, P; O'Berg, M; et al. (1986) Mortality among industrial workers exposed to formaldehyde. J Natl
Cancer Inst 76:1071-1084.
Blair, A; Stewart, PA; Hoover, RN; et al. (1987) Cancers of the nasopharynx and oropharynx and formaldehyde
exposure. J Natl Cancer Inst 78:191-193.
Blair, A; Stewart, PA. (1990) Correlation between different measures of occupational exposure to formaldehyde,
Am J Epidemiol 131:510-516.
Blair, A; Stewart, PA; Hoover, RN. (1990) Mortality from lung cancer among workers employed in formaldehyde
industries. Am J Ind Med 17:683-699.
Blair, A; Linos, A; Stewart, A; et al. (1993) Evaluation of risks for Non-Hodgkin's Lymphoma by occupation and
industry exposures from a case-control study. Am J Ind Med 23:301-312.
Blasiak, J; Trzeciak, A; Malecka-Panas, E, et al. (2000) In vitro genotoxicity of ethanol and acetaldehyde in human
lymphocytes and the gastrointestinal tract mucosa cells. Toxicol In Vitro 14:287-295.
Blejer, HP; Miller, BH. (1966) Occupational health report of formaldehyde concentrations and effects on workers at
the Bayly Manufacturing Company, Visalia, California. Bureau of Occupational Health, California Department of
Public Health, Los Angeles, CA; Study No. S-1806.
Bloom, B; Cohen, RA; Freeman, G. (2009) Summary health statistics for U.S. children: National Health Interview
Survey, 2008. National Center for Health Statistics. Vital Health Stat 10(244).
Boeniger, MF. (1987) Formate in urine as a biological indicator of formaldehyde exposure: A review. Am Ind Hyg
Assoc J 48:900-908.
Boffetta, P; Stellman, SD; Garfinkel, L. (1989) A case-control study of multiple myeloma nested in the American
Cancer Society prospective study. Int J Cancer 43:554-559.
Bogdanffy, MS; Morgan, PH; Starr, TB; et al. (1987) Binding of formaldehyde to human and rat nasal mucus and
bovine serum albumin. Toxicol Lett 38:145-154.
Bogdanffy, MS; Sarangapani, R; Plowchalk, DR; et al. (1999) A biologically based risk assessment for vinyl
acetate-induced cancer and noncancer inhalation toxicity. Toxicol Sci 51:19-3 5.
Bogdanffy, MS; Plowchalk, DR; Sarangapani, R, et al. (2001) Mode of action-based dosimeters for interspecies
extrapolation of vinyl acetate inhalation risk. Inhal Toxicol 13:377-396.
Boja, JW; Nielsen, JA; Foldvary, E; et al. (1985) Acute low-level formaldehyde behavioural and neurochemical
toxicity in the rat. Prog Neuro-Psychopharm 9:671-674.
This document is a draft for review purposes only and does not constitute Agency policy.
R-5 DRAFT—DO NOT CITE OR QUOTE
-------
Bolumar, F; Olsen, J; Boldsen, J. (1996) Smoking reduces fecundity: A European multicenter study on fertility and
subfecundity. The European study group on infertility and subfecundity. Am J Epidemiol 143:578-587.
Bolumar, F; Olsen, J; Boldsen, J. (1997) Smoking reduces fecundity: A European multicenter study on infertility and
subfecundity. The European study group on infertility and subfecundity. Am J Epidemiol 145:81-82.
Bond, G; Flores, G; Shellenberger, R. (1986) Nested case-control study of lung cancer among chemical workers.
Am J Epidemiol 124:53-66.
Boreiko, CJ; and Ragan, DL (1983). Formaldehyde effects in the CSH/lOT'/i cell transformation assay. In: Gibson,
JE,ed, Formaldehyde toxicity, Washington, DC: Hemisphere Publishing Corporation, 63-71.
Bos, PMJ; Zwart, A; Reuzel, PGJ; et al. (1992) Evaluation of the sensory irritation test for the assessment of
occupational health risk. Crit Rev Toxicol 21:423^50.
Bosetti, C; McLaughlin, JK; Tarone, RE; et al. (2008) Formaldehyde and cancer risk: a quantitative review of cohort
studies through 2006. Annal Oncol 19:29-43.
Bosworth, D; Crofton-Sleigh, C; Venitt, S. (1987) A forward mutation assay using ampicillin-resistance in
Escherichia coli designed for investigating the mutagenicity of biological samples. Mutagenesis 2:455-467.
Boysen, M; Zadig, E; Digernes, V; et al. (1990) Nasal mucosa in workers exposed to formaldehyde: a pilot study.
BrJIndMed 47:116-121.
Brenneman, KA; James, RA; Gross, EA; et al. (2000) Olfactory neuron loss in adult male CD rats following
subchronic inhalation exposure to hydrogen sulfide. Toxicol Pathol 28:326-333.
Brinton, LA; Blot, W; Becker, J; et al. (1984) A case-control study of cancers of the nasal cavity and paranasal
sinuses. Am J Epidemiol 119:896-906.
Brooks, SM. (1985) Occupational asthma. In Bronchial Asthma; Mechanism and Therapeutics. Second edition.
Eds Weiss, EB; Segal, MS; Stein, M. Little, Brown and Co, Boston/Toronto.
Brutlag, D; Schlehuber, C; Bonner, J. (1969) Properties of formaldehyde-treated nucleohistone. Biochemistry
8:3214-3218.
Buckley, LA; Jiang, XZ; James, RA; et al. (1984) Respiratory tract lesions induced by sensory irritants at the RD50
concentration. Toxicol Appl Pharmacol 74:417-429.
Bull, PJ. (2000) Mode of action of liver tumor induction by trichloroethylene and its metabolites, trichloroacetate
and dichloroacetate. Environ Health Perspect 108(Suppl 2):241-259.
Burgaz, S; Cakmak, G; Erdem, O; et al. (2001) Micronuclei frequencies in exfoliated nasal mucosa cells from
pathology and anatomy laboratory workers exposed to formaldehyde. Neoplasma 48:144-147.
Burgaz, S; Erdem, O; Cakmak, G; et al. (2002) Cytogenetic analysis of buccal cells from shoe-workers and
pathology and anatomy laboratory workers exposed to n-hexane, toluene, methyl ethyl ketone and formaldehyde.
Biomarkers 7:151-161.
Burge, PS; Harries, MG; Lam, WK; et al. (1985) Occupational asthma due to formaldehyde. Thorax 40:255-260.
Butt, RD; Vaughan, TL; McKnight, B. (1992) Descriptive epidemiology and survival analysis of nasopharyngeal
carcinoma in the United States. Int J Cancer 52:549-556.
Buss, J; Kuschinsky, K; Kewitz, H; et al. (1964) Enteric resorption of formaldehyde. Naunyn Schmiedebergs Arch
Pharmacol 247:380-381.
This document is a draft for review purposes only and does not constitute Agency policy.
R-6 DRAFT—DO NOT CITE OR QUOTE
-------
Cain, H. (2001) Bronchoprovocation testing. Clin Chest Med 22:651-659.
Cain, W; See, L; Tosun, T. (1986) Irritation and odor from formaldehyde: chamber studies. In: IAQ '86: Managing
indoor air for health and energy conservation: proceedings of the American Society of Heating, Air Conditioning
Engineers Conference, Atlanta, GA; pp. 126-137.
CalEPA (2008) California OEHHA. Hot Spots Guidelines. Air Toxicology and Epidemiology. Acute, 8-hour and
Chronic Reference Exposure Levels. Appendix D. http://oehha.ca.gov/air/hot_spots/2008/AppendixDl_final.pdf
Callas, PW; Pastides, H; Hosmer, DW, Jr. (1996) Lung cancer mortality among workers in formaldehyde industries.
J Occup Environ Med 38:747-748.
Cap, P; Dryahina, K; Pehal, F; et al. (2008) Selected ion flow tube mass spectrometry of exhaled breath condensate
headspace. Rapid Commun Mass Spectrom. 22(18):2844-2850.
Carraro, E; Gasparini, S; Petrini, T; et al. (1997) Immune response prevalence to formaldehyde-human serum
albumin molecular adduct in a healthy population. J Environ Pathol Toxicol Oncol 16(2-3):215-218.
Carson, S; Goldhamer, R; Weinberg, MS. (1966) Characterization of physical, chemical and biological properties of
mucus in the intact animal. Ann NY Acad Sci 130:935-943.
Casanova, M; Heck, HD. (1987) Further studies of the metabolic incorporation and covalent binding of inhaled [3H]-
and [14C]formaldehyde inFischer-344 rats: effects of glutathione depletion. Toxicol Appl Pharmacol 89:105-121.
Casanova, M; Heck, HD; Everitt, JI; et al. (1988) Formaldehyde concentrations in the blood of rhesus monkeys after
inhalation exposure. Food Chem Toxicol 26:715-716.
Casanova, M; Devo, DF; Heck, HD. (1989) Covalent binding of inhaled formaldehyde to DNA in the nasal mucus
of Fischer 344 rats: analysis of formaldehyde and DNA by high performance liquid chromatography and provisional
pharmacokinetic interpretation. Fundam Appl Toxicol 12:397-417.
Casanova, M; Morgan, KT; Steinhagen, WH; et al. (1991) Covalent binding of inhaled formaldehyde to DNA in the
respiratory tract of rhesus monkeys: pharmacokinetics, rat-to-monkey interspecies scaling, and extrapolation to man.
Fundam Appl Toxicol 17:409^428.
Casanova, M; Morgan, KT; Gross, EA; et al. (1994) DNA-protein cross-links and cell replication at specific sites in
the nose of F344 rats exposed subchronically to formaldehyde. Fundam Appl Toxicol 23:525-536.
Casanova, M; Bell, DA; and Heck, HD. (1997) Dichloromethane metabolism to formaldehyde and reaction of
formaldehyde with nucleic acids in hepatocytes of rodents and humans with and without glutathione S-transferase
Tl and Ml genes. Fundam Appl Toxicol 37:168-180.
Casanova, M; and Heck, HA. (1997) Lack of evidence for the involvement of formaldehyde in the
hepatocarcinogenicity of methyl tertiary-butyl ether in CD-I mice. Chem Biol Interact 105:131-143.
Casanova, M; Cole, P; Collins, JJ; et al. (2004) Re: mortality from lymphohematopoietic malignancies among
workers in formaldehyde industries. J Natl Cancer Inst 96:966-967.
Casanova-Schmitz, M; Heck, HD. (1983) Effects of formaldehyde exposure on the extractability of DNA from
proteins in the rat nasal mucosa. Toxicol Appl Pharmacol 70:121-132.
Casanova-Schmitz, M; David, RM; Heck, HD. (1984a) Oxidation of formaldehyde and acetaldehyde by NAD+-
dependent dehydrogenases in rat nasal mucosal homogenates. Biochem Pharmacol 33:1137-1142.
Casanova-Schmitz, M; Starr, TB; Heck, HD. (1984b) Differentiation between metabolic incorporation and covalent
binding in the labeling of macromolecules in the rat nasal mucosa and bone marrow by inhaled [14C]- and [3H]-
formaldehyde. Toxicol Appl Pharmacol 76:26-44.
This document is a draft for review purposes only and does not constitute Agency policy.
R-7 DRAFT—DO NOT CITE OR QUOTE
-------
Cascieri, T; Clary, JJ. (1992) Formaldehyde-oral toxicity assessment. Comments Toxicol 4:295-304.
Cassee, FR; Arts, JH; Groten, JP; et al. (1996) Sensory irritation to mixtures of formaldehyde, acrolein, and
acetaldehyde in rats. Arch Toxicol 70:329-337.
Cassee, FR; Feron, VJ. (1994) Biochemical and histopathological changes in nasal epithelium of rats after 3-day
intermittent exposure to formaldehyde and ozone alone or in combination. Toxicol Lett 72:257-268.
Cassee, FR; Groten, JP; Feron, VJ. (1996) Changes in the nasal epithelium of rats exposed by inhalation to mixtures
of formaldehyde, acetaldehyde, and acrolein. Fundam Appl Toxicol 29:208-218.
Casset, A; Purohit, A; Marchand, C; et al. (2006) The bronchial response to inhaled formaldehyde. Rev Mai Respir
23(1 Suppl):3S25-3S34.
Cassidy, SL; Dix, KM; Jenkins, T. (1983) Evaluation of a testicular sperm head counting technique using rats
exposed to dimethoxyethyl phthalate (DMEP), glycerol alpha-monochlorohydrin (GMCH), epichlorohydrin (ECH),
formaldehyde (FA), or methyl methanesulphonate (MMS). Arch Toxicol 53:71-78.
Castonguay, A; Pepin, P; Briere, N. (1991) Modulation of 4-(methylnitrosamino)-l-(3-pyridyl)-! butanone
demethylation and denitrosation by rat liver microsomes. Cancer Lett 59:67-74.
CDC (U.S. Centers for Disease Control and Prevention). (2000) Measuring childhood asthma prevalence before and
after the 1997 redesign of the National Health Interview Survey - United States. MMWR 49:908-911.
CDC. (2008) Final report on formaldehyde levels in FEMA-supplied travel trailers, park models, and mobile homes.
Available from http://www.cdc.gov/nceh/ehhe/trailerstudy/residents.htmffinal.
Ceder, R; Merne, M; Staab, C; et al. (2007) The application of normal, SV40 T-antigen-immortalized and tumour-
derived oral keratinocytes, under serum-free conditions, to the study of the probability of cancer progression as a
result of environmental exposure to chemicals. Alternatives to Laboratory Animals 35:621-639.
Cederbaum, Al; Qureshi, A. (1982) Role of catalase and hydroxyl radicals in the oxidation of methanol by rat liver
microsomes. Biochem Pharmacol 31:329-335.
Chang, JC; Barrow, CS. (1984) Sensory irritation tolerance and cross-tolerance inF-344 rats exposed to chlorine or
formaldehyde gas. Toxicol Appl Pharmacol 76:319-327.
Chang, JC; Steinhagen, WH; Barrow, CS. (1981) Effect of single or repeated formaldehyde exposure on minute
volume of B6C3F1 mice and F-344 rats. Toxicol Appl Pharmacol 61:451^159.
Chang, JC; Gross, EA; Swenberg, JA; et al. (1983) Nasal cavity deposition, histopathology, and cell proliferation
after single or repeated formaldehyde exposures in B6C3F1 mice and F-344 rats. Toxicol Appl Pharmacol
68:161-176.
Chebotarev, AN; Titenko, NV; Selezneva, TG, et al. (1986) [Comparison of chromosome aberrations, sister
chromatid exchanges and unscheduled DNA synthesis in the evaluation of the mutagenicity of environmental
factors]. Tsitol Genet 20:109-115.
Cheng, YS; Hansen, GK; Su, FY; et al. (1990) Deposition of ultrafine aerosols in rat nasal molds. Toxicol Appl
Pharmacol 106:222-233.
Cheng, G; Shi, Y; Sturla, SJ; et al. (2003) Reactions of formaldehyde plus acetaldehyde with deoxyguanosine and
DNA: formation of cyclic deoxyguanosine adducts and formaldehyde cross-links. Chem Res Toxicol 16:145-152.
Cheng, XM; Zhao, J; Feng, B; et al. (2008) Relationship between susceptibility of formaldehyde metabolism and
genetic polymorphisms of ALD2 and cytochrome P4502E1. Zhonghua Yu Fang Yi Xue Za Zhi 42:582-587.
This document is a draft for review purposes only and does not constitute Agency policy.
R-8 DRAFT—DO NOT CITE OR QUOTE
-------
Cheng, G; Wang, M; Upadhyaya, P, et al. (2008) Formation of formaldehyde adducts in the reactions of DNA and
deoxyribonucleosides with alpha-acetates of 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone (NNK), 4-
(methylnitrosamino)-l-(3-pyridyl)-l-butanol (NNAL), and N-nitrosodimethylamine (NDMA). Chem Res Toxicol
21:746-751.
Chernoff, N; Kavlock, RJ. (1982) An in vivo teratology screen utilizing pregnant mice. J Toxicol Environ Health
10:541-550.
Chiu, WA, Euling SY, Scott CS, and Subramaniam RP. (2010) Approaches to advancing quantitative human health
risk assessment of environmental chemicals in the post-genomic era. Toxicology and Applied Pharmacology (Epub
ahead of print March 29). Doi:10.1016/j.taap.2010.03.019.
Cho, JY; Min, N; Franzen, L; et al. (1996) Rapid down-regulation of tyrosine hydroxylase expression in the
olfactory bulb of naris-occluded adult rats. J Comp Neurol 369:264-276.
Chowdhury, AR; Gautam, AK; Patel, KG; etal. (1992) Steroidogenic inhibition in testicular tissue of formaldehyde
exposed rats. Ind J Physiol Pharmacol 36:162-168.
CUT (Chemical Industry Institute of Toxicology). (1982) Summary of final report on formaldehyde study. CUT
Activities 2:1,9-10.
CUT. (1999) Formaldehyde: Hazard characterization and dose-response assessment for carcinogenicity by the route
of inhalation (Revised Edition). Research Triangle Park, NC: Chemical Industry Institute of Toxicology.
Claeys, S; Cuvelier, C; Quatacker, J; et al. (1996a) Ultrastructural investigation of M-cells and lymphoepithelial
contacts in naso-pharangeal associated lymphoid tissue (NALT). Acta Otolaryngeal Supple 523:40^12.
Claeys, S; Cuvelier, C; Van Cauwenberge, P. (1996b) Immunohistochemical analyses of the lymphoepithelium in
human nasopharyngeal associated lymphoid tissue (NALT) Acta Otolaryngol Suppl 523:38-39.
Coggon, D; Pannett, B; Acheson, ED. (1984) Use of job-exposure matrix in an occupational analysis of lung and
bladder cancers on the basis of death certificates. JNatl Cancer Inst72(l):61-65.
Coggon, D; Harris, EC; Poole, J; et al. (2003) Extended follow-up of a cohort of British chemical workers exposed
to formaldehyde. J Natl Cancer Inst 95:1608-1615.
Cohen-Hubal, EA; Schlosser, PM; Conolly, RB; et al. (1997) Comparison of inhaled formaldehyde dosimetry
predictions with DNA-protein cross-link measurements in the rat nasal passages. Toxicol Appl Pharmacol
143:47-55.
Cole, P; Axten, C. (2004) Formaldehyde and leukemia: an improbable causal relationship. Regul Toxicol Pharmacol
40:107-112.
Collins, JJ. (2004) Formaldehyde exposure and leukemia. Occup Environ Med 61:875-876.
Collins, JJ; Lineker, GA. (2004) A review and meta-analysis of formaldehyde exposure and leukemia. Regul
Toxicol Pharmacol 40:81-91.
Collins, JJ; Acquavella, JF; Esmen, NA. (1997) An updated analysis of formaldehyde exposure and upper
respiratory tract cancers. J Occup Environ Med 3 9: 63 9-651.
Collins, JJ; Esmen, NA; Hall, TA. (200 la) A review and meta-analysis of formaldehyde exposure and pancreatic
cancer. Am J Ind Med 39:336-345.
Collins, JJ; Ness, R; Tyl, RW; et al. (200 Ib) A review of adverse pregnancy outcomes and formaldehyde exposure
in human and animal studies. Regul Toxicol Pharmacol 34:17-34.
This document is a draft for review purposes only and does not constitute Agency policy.
R-9 DRAFT—DO NOT CITE OR QUOTE
-------
Cometto-Muniz, JE; Cain, WS. (1992) Sensory irritation. Relation to indoor air pollution. Ann NY Acad Sci
641:137-151.
Conaway, CC; Whysner, J; Verna, LK, et al. (1996) Formaldehyde mechanistic data and risk assessment:
endogenous protection from DNA adduct formation. Pharmacol Ther 71:29-55.
Connor, TH; Barrie, MD; Theiss, JC; et al. (1983) Mutagenicity of formalin in the Ames assay. Mutat Res
119:145-149.
Connor, TH; Theiss, JC; Hanna, HA; et al. (1985) Genotoxicity of organic chemicals frequently found in the air of
mobile homes. Toxicol Lett 25:33-40.
Conolly, RB, (1995) Cancer and non-cancer risk assessment: Not so different if you consider mechanisms.
Toxicology 102:179-188.
Conolly, RB. (2002) The use of biologically based modeling in risk assessment. Toxicology 181-182:275-279.
Conolly, RB. (2009) Commentary on "Toxicity testing in the 21st century: Implications for human health risk
assessment" by Krewski et al. Risk Anal 29:480-481.
Conolly, RB; Lutz, WK. (2004) Nonmonotonic dose-response relationships: Mechanistic basis, kinetic modeling,
and implications for risk assessment. Toxicol Sci 77:151-157.
Conolly, RB; Lilly, PD; Kimbell, JS. (2000) Simulation modeling of the tissue disposition of formaldehyde to
predict nasal DNA-protein cross-links in Fischer 344 rats, rhesus monkeys, and humans. Environ Health Perspect
108 Suppl 5:919-924.
Conolly, RB; Kimbell, JS; Janszen, D; et al. (2003) Biologically motivated computational modeling of formaldehyde
carcinogenicity in the F344 rat. Toxicol Sci 75:432^47.
Conolly, RB; Kimbell, JS; Janszen, D; et al. (2004) Human respiratory tract cancer risks of inhaled formaldehyde:
dose-response predictions derived from biologically-motivated computational modeling of a combined rodent and
human dataset. Toxicol Sci 82:279-296.
Conolly, RB; Miller, FJ; Kimbell, JS; et al. (2009) Formaldehyde risk assessment. Ann Hyg 53:181-184.
Cook, RJ; Champion, KM; Giometti, CS. (2001) Methanol toxicity and formate oxidation in NEUT2 mice. Arch
Biochem Biophys 393:192-198.
Coon, RA; Jones, RA; Jenkins, LJ, Jr; et al. (1970) Animal inhalation studies on ammonia, ethylene glycol,
formaldehyde, dimethylamine, and ethanol. Toxicol Appl Pharmacol 16:646-655.
Correa, A; Gray, RH; Cohen, R; et al. (1996) Ethylene glycol ethers and risks of spontaneous abortion and
subfertility. Am JEpidemiol 143:707-717.
Cosma, GN; Marchok, AC. (1988) Benzo[a]pyrene- and formaldehyde-induced DNA damage and repair in rat
tracheal epithelial cells. Toxicology 51:309-320.
Cosma, GN; Jamasbi, R; Marchok, AC. (1988) Growth inhibition and DNA damage induced by benzo[a]pyrene and
formaldehyde in primary cultures of rat tracheal epithelial cells. Mutat Res 201:161-168.
Cosma, GN; Wilhite, AS; and Marchok, AC. (1988) The detection of DNA-protein cross-links in rat tracheal
implants exposed in vivo tobenzo[a]pyrene and formaldehyde. Cancer Lett 42:13-21.
Costa, M; Zhitkovich, A; Harris, M; et al. (1997) DNA-protein cross-links produced by various chemicals in
cultured human lymphoma cells. J Toxicol Environ Health 50:433-449.
This document is a draft for review purposes only and does not constitute Agency policy.
R. i o DRAFT—DO NOT CITE OR QUOTE
-------
Costa, S; Coelho, P; Costa, C, et al. (2008) Genotoxic damage in pathology anatomy laboratory workers exposed to
formaldehyde. Toxicology 252:40-48.
Coste, A; Rateau, JG; Roudot-Thoraval, F; et al. (1996) Increased epithelial cell proliferation in nasal polyps. Arch
Otol Head Neck Surg 122:432^36.
Couch, DB; Allen, PF; and Bales, HC. (1982) The mutagenicity of formaldehyde to Salmonella typhimurium
[Abstract]. Environ Mutagen 4:336-337.
Cox, DR. (1972) Regression models and life tables (with discussion). J Roy Stat Soc (Series B) 34:187-220. (as
cited inBattelle Columbus Laboratories [1981]).
Cox, D; Hinkley, D. (1974) Theoretical Statistics. Chapman and Hall, London, UK.
Cox, D; Oakes, D. (1984) Analysis of Survival Data. Chapman and Hall, London, UK.
Craft, TR; Bermudez, E; Skopek, TR. (1987) Formaldehyde mutagenesis and formation of DNA-protein crosslinks
in human lymphoblasts in vitro. Mutat Res 176:147-155.
Cralley, LV. (1942) The effect of irritant gases upon the rate of ciliary activity. J Ind Hyg Toxicol 24:193-198.
Crosby, RM; Richardson, KK; Craft, TR; et al. (1988) Molecular analysis of formaldehyde-induced mutations in
human lymphoblasts and E. coli. Environ Mol Mutagen 12:155-166.
Crump, KS. (1994a) Limitations of biological models of carcinogenesis for low-dose extrapolation. Risk Anal
14:883-886.
Crump, KS. (1994b) Use of mechanistic models to estimate low-dose cancer risks. Risk Anal 14:1033-1038.
Crump, K. (2002) Benchmark analysis. In: Encyclopedia of Environmetrics. Eds: El-Shaarawi, A; Piegorsch, WW.
Wiley, West Sussex, UK.
Crump, KS; Subramaniam, RP; Van Landingham, C. (2005) A numerical solution to the non-homogenous two-stage
MVK model of cancer. Risk Anal 25:921-926.
Crump, K; Subramanian, RP; Van Landingham, C. (2006) Acknowledgment of prior solution to MVK cancer model.
Risk Anal 26:3-4.
Crump, KS; Chen, C; Fox, JF; et al. (2008) Sensitivity analysis of biologically motivated model for formaldehyde-
induced respiratory cancer in humans. Ann Occup Hyg 52:481^195.
Crump, KS; Chen, C; Fox, JF; et al. (2009) Reply to Conolly et al. (2009) Ann Hyg 53:184-189.
Dahl, AR; Hadley, WM. (1983) Formaldehyde production promoted by rat nasal cytochrome P-450-dependent
monooxygenases with nasal decongestants, essences, solvents, air pollutants, nicotine, and cocaine as substrates.
Toxicol Appl Pharmacol 67:200-205.
Dalbey, WE. (1982) Formaldehyde and tumors in hamster respiratory tract. Toxicology 24:9-14.
Dalhamn, T. (1956) Mucus flow and ciliary activity in the trachea of healthy rats and rats exposed to respiratory
irritant gases (SO2, H3N, HCHO): a functional and morphologic (light microscopy and electron microscopy) study,
with special reference to technique. Acta Physiol Scand 36(Suppl): 1-161.
Dallas, CE; Badeaux, P; Theiss, JC; et al. (1989) The influence of inhaled formaldehyde on rat lung cytochrome
P450. Environ Res 49:50-59.
Dallas, CE; Scott, MJ; Ward, JB, Jr.; et al. (1992) Cytogenetic analysis of pulmonary lavage and bone marrow cells
of rats after repeated formaldehyde inhalation. J Appl Toxicol 12:199-203.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 11 DRAFT—DO NOT CITE OR QUOTE
-------
Dally, K; Hanrahan, L; Woodbury, M; et al. (1981) Formaldehyde exposure in nonoccupational environments. Arch
Environ Health 36(6):277-284.
Dalton, P. (2003) Upper airway irritation, odor perception and health risk due to airborne chemicals. Toxicol Lett
140-141:239-248.
Dasgupta PK, J Li, G Zhang, WT Luke, WA McClenny, J Stutz and A Fried. Summertime ambient formaldyde in
five U.S. metropolitan areas: Nashville, Atlanta, Houston, Philadelphia and Tampa. Environ Sci Technol 39:4767-
4783.
Daston, GP. (2008) Gene expression, dose-response, and phenotypic anchoring: Applications for toxicogenomics in
risk assessment. Toxicol Sci 105:233-234.
Day, JH; Lees, RE; Clark, RH; et al. (1984) Respiratory response to formaldehyde and off-gas of urea formaldehyde
foam insulation. Can Med Assoc J 131:1061-1065.
De Ceaurriz, JC; Micillino, JC; Bonnet, P; et al. (1981) Sensory irritation caused by various industrial airborne
chemicals. Toxicol Lett 9:137-143.
De Flora, S. (1981) Study of 106 organic and inorganic compounds in the Salmonella/microsome test.
Carcinogenesis 2:283-98.
de Flora, S; Camoirano, A; Zanacchi, P, et al. (1984) Mutagenicity testing with TA97 and TA102 of 30 DNA-
damaging compounds, negative with other Salmonella strains. Mutat Res 134:159-165.
De Stefani, E; Boffetta, P; Brennan, P; et al. (2005) Occupational exposures and risk of adenocarcinoma of the lung
in Uruguay. Cancer Causes and Control 16:851-856.
Dean, JH; Lauer, LD; House, RV; et al. (1984) Studies of immune function and host resistance in B6C3F1 mice
exposed to formaldehyde. Toxicol Appl Pharmacol 72:519-529.
Dearman, RJ; Humphreys, N; Skinner, RA; et al. (2005) Allergen-induced cytokine phenotypes in mice: role of CD4
and CDS T cell populations. Clin Exp Allergy 35:498-505.
Degitz, SJ; Zucker, RM; Kawanishi, CY; et al. (2004a) Pathogenesis of methanol-induced cranofacial defects in
C57BL/6J mice. Birth Defects A Clin Mol Teratol 70:172-178.
Degitz, SJ; Rogers, JM; Zucker, RM; et al. (2004b) Developmental toxicity of methanol: Pathogenesis in CD-I and
C57BL/6J mice exposed in whole embryo culture. Birth Defects Res A Clin Mol Teratol. 70(4): 179-184.
Delfino, RJ. (2002) Epidemiologic evidence for asthma and exposure to air toxics: linkages between occupational,
indoor, and community air pollution research. Environ Health Perspect 110 Suppl 4:573-589.
Delfino, RJ; Gong, H Jr; Linn, WS; et al. (2003) Asthma symptoms in Hispanic children and daily ambient
exposures to toxic and criteria air pollutants. Environ Health Perspect 111 :647-656.
Dell, L; Teta, MJ. (1995) Mortality among workers at a plastics manufacturing and research and development
facility: 1946-1988. Am JInd Med 28:373-384.
Deltour, L; Foglio, MH; Duester, G. (1999) Metabolic deficiencies in alcohol dehydrogenase Adhl, Adh3, and Adh4
null mutant mice. Overlapping roles of Adhl and Adh4 in ethanol clearance and metabolism of retinol to retinoic
acid. JBiolChem 274:16796-16801.
Demkowicz-Dobrzanski, K; and Castonguay, A. (1992) Modulation by glutathione of DNA strand breaks induced
by 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone and its aldehyde metabolites in rat hepatocytes. Carcinogenesis
13:1447-1454.
This document is a draft for review purposes only and does not constitute Agency policy.
R-12 DRAFT—DO NOT CITE OR QUOTE
-------
Dicker, E; Cederbaum, AI. (1986) Inhibition of the low-Km mitochondria! aldehyde dehydrogenase by diethyl
maleate andphorone in vivo and in vitro. Implications for formaldehyde metabolism. Biochem 1240:821-827.
Dikalova, AE; Kadiiska, MB; Mason, RP. (2001) An in vivo ESR spin-trapping study: free radical generation in rats
from formate intoxication-role of the Fenton reaction. Proc Natl Acad Sci USA 98:13549-13553.
Dillon, D; Combes, R; Zeiger, E. (1998) The effectiveness of Salmonella strains TA100, TA102 and TA104 for
detecting mutagenicity of some aldehydes and peroxides. Mutagenesis 13:19-26.
Dinsdale, D; Riley, RA; Verschoyle, RD. (1993) Pulmonary cytochrome P450 in rats exposed to formaldehyde
vapor. Environ Res 62:19-27.
Doenecke, D. (1978) Digestion of chromosomal proteins in formaldehyde treated chromatin. Hoppe Seylers Z
Physiol Chem 359:1343-1352.
Doi, S; Suzuki, S; Morishita, M; et al. (2003) The prevalence of IgE sensitization to formaldehyde in asthmatic
children. Allergy 58(7):668-671.
Doll, R; Peto, R. (1978) Cigarette smoking and bronchial carcinoma: Dose and time relationships among regular
smoker and lifetime nonsmokers. J Epidem Commun Health 32:303-313.
Donovan, SM; Krahn, DF; Stewart, JA; et al. (1983) Mutagenic activities of formaldehyde (HCHO) and
hexamethylphosphoramide (HMPA) in reverse and forward Salmonella typhimurium mutation assays. Environ
Mutagen 5:476.
Doolittle, DJ; Furlong, JW; and Butterworth, BE. (1985) Assessment of chemically induced DNA repair in primary
cultures of human bronchial epithelial cells. Toxicol Appl Pharmacol 79:28-38.
Doty, RL; Cometto-Muniz, JE; Jalowayski, AA; et al. (2004) Assessment of upper respiratory tract and ocular
irritative effects of volatile chemicals in humans. Crit Rev Toxicol 34:85-142.
Dragan, YP; Hully, J; Baker, K; et al. (1995) Comparison of experimental and theoretical parameters of the
Moolgavkar-Venzon-Knudson incidence function for the stages of initiation and promotion in rat
hepatocarcinogenesis. Toxicology 102:161-175.
Dresp, J; and Bauchinger, M. (1988) Direct analysis of the clastogenic effect of formaldehyde in unstimulated
human lymphocytes by means of the premature chromosome condensation technique. Mutat Res 204:349-352.
Dubreuil, A; Bouley, G; Godin, J; et al. (1976) Continuous inhalation of small amounts of formaldehyde:
experimental study in the rat. Eur J Toxicol Environ Hyg 9:245-250.
Dulskiene, V; Grazuleviciene, R. (2005) Environmental risk factors and outdoor formaldehyde and risk of congenital
heart malformations. Medicina (Kaunas) 41:787-795.
Dykewicz, M; Patterson, R; Cugell, DW; et al. (1991) Serum IgE and IgG to formaldehyde-human serum albumin
lack of relation to gaseous formaldehyde exposure and symptoms. J Allergy Clin Immunol 87:48-57.
Edling, C; Jarvholm, B; Andersson, L; et al. (1987) Mortality and cancer incidence among workers in an abrasive
manufacturing industry. Br J Ind Med 44:57-59.
Edling, C; Hellquist, H; Odkvist, L. (1988) Occupational exposure to formaldehyde and histopathological changes in
the nasal mucosa. Br J Ind Med 45:761-765.
Egle, JL, Jr. (1972) Retention of inhaled formaldehyde, propionaldehyde, and acrolein in the dog. Arch Environ
Health 25:119-124.
Ehlers, CL. (2007) Variations in ADH and ALDH in Southwest California Indians. Alcohol Res Health 30:14-17.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 13 DRAFT—DO NOT CITE OR QUOTE
-------
Emri, G; Schaefer, D; Held, B; et al. (2004) Low concentrations of formaldehyde induce DNA damage and delay
DNA repair after UV irradiation in human skin cells. Exp Dermatol 13:305-315.
Environment California. (2008) New Report: Formaldehyde from baby nursery furniture contaminates indoor air
Californis's green chemistry initative can protect consumers. Available at
http://www.environmentcalifornia.org/newsroom/environmenM-health/environmental-health/new-report-
formaldehyde-from-baby-nursery-furniture-contaminates-indoor-air.
Epstein, S.S.; Shafner, H. (1968) Chemical mutagens in human environment. Nature 219: 385.
Epstein, S.S. etal (1972) Detection of chemical mutagens by the dominant lethal assay in the mosue. Toxicol. Appl.
Pharmacol. 23:288.
Erdei, E; Bobvos, J; Brozik, M; et al. (2003) Indoor air pollutants and immune biomarkers among Hungarian
asthmatic children. Arch Environ Health 58:33 7-3 47.
Estonius, M; Svensson, S; Hoog, JO. (1996) Alcohol dehydrogenase in human tissues: localization of transcript
coding for five classes of the enzyme. FEES Lett 397:338-342.
Ezratty, V; Bonay, M; Neukirch, C; et al. (2007) Effect of formaldehyde on asthmatic response to inhaled allergen
challenge. Environ Health Perspect 115(2):210-214.
Fanucchi, MV; Buckpitt, AR; Murphy, ME; et al. (1997) Naphthalene cytotoxicity of differentiating Clara cells in
neonatal mice. Toxicol Appl Pharmacol 144:96-104.
Fanucchi, MV; Buckpitt, AR; Murphy, ME; et al. (2000) Development of phase II xenobiotic metabolizing enzymes
in differentiating murine Clara cells. Toxicol Appl Pharmacol 168:253-267.
Fanucchi, MV; Day, KC; Clay, CC; et al. (2004) Increased vulnerability of neonatal rats and mice to 1-
nitronaphthalene-induced pulmonary injury. Toxicol Appl Pharmacol 201:53-65.
Farooqui, MY; Upreti, RK; Ahmed, AE; et al. (1986) Influence of intraperitoneally administered formaldehyde on
bile production and tissue glutathione levels in rats. Res Commun Chem Pathol Pharmacol 53:233-236.
Fayerweather, WE; Pell, S; Bender, J. (1983) Case-control study of cancer deaths in DuPont workers with potential
exposure to formaldehyde. In: Clary, J; Gibson, J; Waritz, RS; eds. Formaldehyde: toxicology, epidemiology, and
mechanisms. New York: Marcel Dekker; pp. 47-121.
Fennell, TR. (1994) Development of methods for measuring biological markers of formaldehyde exposure. Res Rep
Health Eff Inst: 1-20; discussion 21-26.
Fernandez, MR; Biosca, JA; Pares, X. (2003) S-nitrosoglutathione reductase activity of human and yeast
glutathione-dependent formaldehyde dehydrogenase and its nuclear and cytoplasmic localization. Cell Mol Life Sci
60:1013-1018.
Feron, VJ; Woutersen, RA; Appelman, LM. (1984) Epithelial damage and tumours of the nose after exposure to four
different aldehydes by inhalation. In: Grosdanoff, P; Bass, R; Hackenberg, U; et al.; eds. Problems of inhalation
toxicity studies. Munich: MMV Medizin Verlag; pp. 587-610.
Feron, VJ; Bruyntjes, JP; Woutersen, RA; et al. (1988) Nasal tumours in rats after short-term exposure to a cytotoxic
concentration of formaldehyde. Cancer Lett 39:101-111.
Feron, VJ; Wilmer, JW; Woutersen, RA; et al. (1989) Inhalation toxicity and carcinogenicity of formaldehyde in
animals: significance for assessment of human health risk. In: Bates, D; Dungworth, DL; Lee, P; et al.; eds.
Assessment of inhalation hazards: integration and extrapolation using diverse data. Berlin: Springer-Verlag; pp.
131-138.
This document is a draft for review purposes only and does not constitute Agency policy.
R-14 DRAFT—DO NOT CITE OR QUOTE
-------
Feron, VJ; Til, HP; de Vrijer, F; et al. (1991) Aldehydes: occurrence, carcinogenic potential, mechanism of action
and risk assessment. Mutat Res 259:363-385.
Feron, VJ; Arts, JH; Kuper, CF; et al. (2001) Health risks associated with inhaled nasal toxicants. Crit Rev Toxicol
31:313-347.
Fiddler, W; Miller, AJ; Pensabene, JW; et al. (1984) Investigation on the mutagenicity of N-nitrosothiazolidine using
the Ames Salmonella test. IARC Sci Publ:95-100.
Figuero, WG; Raszkowski, R; Weiss, W. (1973) Lung cancer in chloromethyl ether workers. N Engl J Med
288:1096-1097.
Fireman, P. (2003) Understanding asthma pathophysiology. Allergy Asthma Proc 24:79-83.
Fishbein, L. (1992) Exposure from occupational versus other sources. Scand J Work Environ Health
18(Suppl 1):5-16.
Fleig, I; Petri, N; Stacker, WG; et al. (1982) Cytogenetic analyses of blood lymphocytes of workers exposed to
formaldehyde in formaldehyde manufacturing and processing. J Occup Med 24:1009-1012.
Fleisher, J. (1987) Medical students' exposure to formaldehyde in gross anatomy laboratories. N Y State J Med
87(7):385-388.
F16-Neyret, C; Lorenzl-Filho, G; Macchione, M; et al. (2001) Effects of formaldehyde on the frog's mucociliary
epithelium as a surrogate to evaluate air pollution effects on the respiratory epithelium. J Med Bil Res 34:639-643.
Fontignie-Houbrechts, N. (1981) Genetic effects of formaldehyde in the mouse. Mutat Res 88:109-114.
Fontignie-Houbrechts, N; Moutschen-Dahmen, M; Degraeve, N; et al. (1982) Genetic effects in the mouse of
formaldehyde in combination with adenosine and hydrogen peroxide. Mutat Res 104:371-376.
Fornace, AJ, Jr. (1982) Detection of DNA single-strand breaks produced during the repair of damage by DNA-
protein cross-linking agents. Cancer Res 42:145-149.
Franklin, P; Dingle, P; Stick, S. (2000) Raised exhaled nitric oxide in healthy children is associated with domestic
formaldehyde levels. Am J Respir Crit Care Med 161:1757-1759.
Franks, SJ. (2005) A mathematical model for the absorption and metabolism of formaldehyde vapour in humans.
Toxicol Appl Pharmacol 206:309-320.
Fransman, W; McLean, D; Douwes, J; et al. (2003) Respiratory symptoms and occupational exposures in New
Zealand plywood mill workers. Ann Occup Hyg 47(4):287-295.
Frazelle, JH; Abernethy, DJ; and Boreiko, CJ. (1983) Weak promotion of C3H/10T1/2 cell transformation by
repeated treatments with formaldehyde. Cancer Res 43:3236-3239.
Frei, E; Pool, BL; Plesch, W; et al. (1984) Biochemical and biological properties of prospective N-nitrodialkylamine
metabolites and their derivatives. IARC Sci Publ:491-497.
Friedman, GD; Ury, HK. (1983) Screening for possible drug carcinogenicity: second report of findings. J Natl
Cancer Inst 71:1165-1175.
Frigas, E; Filley, WV; Reed, CE. (1984) Bronchial challenge with formaldehyde gas: lack of bronchoconstriction in
13 patients suspected of having formaldehyde-induced asthma. Mayo Clin Proc 59:295-299.
Fujimaki, H; Kurokawa, Y; Kunugita, N; et al. (2004a) Differential immunogenic and neurogenic inflammatory
responses in an allergic mouse model exposed to low levels of formaldehyde. Toxicology 197:1-13.
This document is a draft for review purposes only and does not constitute Agency policy.
R- 15 DRAFT—DO NOT CITE OR QUOTE
-------
Fujimaki, H; Kurokawa, Y; Kakeyama, M; et al. (2004b) Inhalation of low-level formaldehyde enhances nerve
growth factor production in the hippocampus of mice. Neuroimmunomodulation 11:373-375.
Fujimaki, H; Kurokawa, Y; Kunugita, N; et al. (2005) Erratum to "Differential immunogenic and neurogenic
inflammatory responses in an allergic mouse model exposed to low levels of formaldehyde" [Toxicology 197 2004
(1-13)]. Toxicology 209:305.
Fujimura, Y. (2000) Evidence of M cells as portals of entry for antigens in the nasopharyngeal lymphoid tissue of
humans. Virchows Arch 436:560-566.
Galloway, SM; Bloom, AD; Resnick, M, et al. (1985) Development of a standard protocol for in vitro cytogenetic
testing with Chinese hamster ovary cells: comparison of results for 22 compounds in two laboratories. Environ
Mutagen7:l-51.
Gaiter, D; Carmine, A; Buervenich, S; et al. (2003) Distribution of class I, III and IV alcohol dehydrogenase mRNAs
in the adult rat, mouse and human brain. Eur J Biochem 270:1316-1326.
Gandhi, M; Aweeka, F; Greenblatt, RM; et al. (2004) Sex differences in pharmacokinetics and pharmaco-dynamics
Ann Rev Pharmacol Toxicol 44:499-523.
Garcia, CL; Mechilli, M; De Santis, LP, et al. (2009) Relationship between DNA lesions, DNA repair and
chromosomal damage induced by acetaldehyde. Mutat Res 662:3-9.
Garcia, GJ; Schroeter, JD; Segal, RA; et al. (2009) Dosimetry of nasal uptake of water-soluble and reactive gases: a
first study of interhuman variability. Inhal Toxicol. 21(7):607-18.
Gardner, MJ; Pannett, B; Winter, PD; et al. (1993) A cohort study of workers exposed to formaldehyde in the British
chemical industry: an update. Br J Ind Med 50:827-834.
Gardner, RJ; Burgess, BA; Kennedy, GL, Jr. (1985) Sensory irritation potential of selected nasal tumorigens in the
rat. Food Chem Toxicol 23:87-92.
Garrett, MH; Hooper, MA; Hooper, BM; et al. (1999a) Increased risk of allergy in children due to formaldehyde
exposure in homes. Allergy 34:330-337.
Garrett, MH; Hooper, MA; Hooper, BM; et al. (1999b) Errata. Allergy 54(12): 1327.
Garry, VF; Kreiger, RA; and Wiencke, JK. (1981) The mutagenic and cytotoxic effects of formaldehyde in cultured
human lymphocytes [Abstract]. Environ Mutagen 3:341.
Gart, JJ; Krewski, D; Lee, PN; et al. (1986) Statistical methods in cancer research. Volume III~The design and
analysis of long-term animal experiments. IARC Sci Publ (79): 1-219.
Gaylor, DW; Zheng, O. (1996) Risk assessment of nongenotoxic carcinogens based upon cell proliferation/death
rates in rodents. Risk Anal 16:221-225.
Gaylor, DW; Lutz, WK; Conolly, RB. (2004) Statistical analysis of nonmonotonic dose-response relationships:
Research design and analysis of nasal cell proliferation in rats exposed to formaldehyde. Toxicol Sci 77:158-164.
Gee, IL; Watson, APR; Tavernier, G; et al. (2005) Indoor air quality, environmental tobacco smoke and asthma: A
case control study of asthma in a community population. Indoor Built Environ 14(3-4):215-219.
Georgieva, AV; Kimbell, JS; Schlosser, PM. (2003) A distributed-parameter model for formaldehyde uptake and
disposition in the rat nasal lining. Inhal Toxicol 15:1435-1463.
Gerard, C. (2005) Biomedicine. Asthmatics breathe easier when its SNO-ing. Science 308:1560-1561.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 16 DRAFT—DO NOT CITE OR QUOTE
-------
Gerberich, HR; Seaman, GC. (1994) Formaldehyde. In: Kroshwitz, JI; Howe-Grant, M; eds. Kirk-Othmer
encyclopedia of chemical technology. Vol. 11. 4th edition. New York, NY: John Wiley and Sons; pp. 929-951.
Gerde, P; Cheng, YS; Medinsky, MA. (1991) In vitro deposition of ultrafine aerosols in the nasal airways of the rat.
Fundam Appl Toxicol 16:330-336.
Gerin, M; Siemiatycki, J; Nadon, L; et al. (1989) Cancer risks due to occupational exposure to formaldehyde:
Results of a multi-site case-control study in Montreal Canada. Int J Cancer 44:53-58.
Ginsberg, G; Smolenski, S; Hattis, D; et al. (2002) Population distribution of aldehyde dehydrogenase-2 genetic
polymorphism: implications for risk assessment. Regul Toxicol Pharmacol 36:297-309.
Ginsberg, G; Hattis, D; Sonawane, B. (2004) Incorporating pharmacokinetic differences between children and adults
in assessing children's risks to environmental toxicants. Toxicol Appl Pharmacol 198:164-183.
Ginsberg, G; Foos, B; Firestone, M. (2005) Review and Analysis of Inhalation Dosimetry Methods for Application
to Children's Risk Assessment. J. Tox. & Env. Health 68:573-615.
Ginsberg,G; Asgharian, B; Kimbell, J; et al. (2008) Modeling approaches for estimating the dosimetry of inhaled
toxicants in children. J. Tox. & Env. Health 71:166-195.
Global Insight, (2006) Economic primer on formaldehyde. Prepared for Formaldehyde Council, Inc.
http://www.formaldehyde.org/pdfs/GIEconomicPrimer-04-20-06.pdf.
Gochfeld, M. (2007) Framework for gender differences in human and animal toxicology. Environ Res 104:4-21.
Gofmekler, V. (1968) Embryotropic action of benzene and formaldehyde inhalation. Hyg Sanit 33:327-332.
Gofmekler, V; Bonashevskaya, T. (1969) Experimental studies of teratogenic properties of formaldehyde, based on
pathological investigations. Hyg Sanit 34:266-268.
GolaliourMJ; AzarhoushR; Ghafari S; Gharravi AM; Fazeli SA; Bavarian A. 2007. Formaldehyde exposure
induces histopathological and morphometric changes in the rat testis. Folia Morphol. 66(3): 167-171.
Golden, R; Pyatt, D; Shields, PG. (2006) Formaldehyde as a potential human leukemogen; An assessment of
biological plausibility. Crit Rev Toxicol 36:135-153.
Goldmacher, VS; and Thilly, WG. (1983) Formaldehyde is mutagenic for cultured human cells. Mutat Res
116:417-422.
Goldstein, BD. (1973) Textiles and the chemical industry: A marriage. J Am Assoc Textile Chemists and Colorists
10:209-214.
Gordon, MY; Marley, SB; Apperley, JF; et al. (2003) Clinical heterogeneity in chronic myeloid leukemia reflecting
biological diversity in normal persons. Brit J Haematol 122:424-429.
Gordon, CJ; Spencer, PJ; Hotchkiss, J; et al. (2008) Thermoregulation and its influence on risk assessment.
Toxicology 244:87-97.
Gorski, P; Tarkowski, M; Krakowiak, A; et al. (1992) Neutrophil chemiluminescence following exposure to
formaldehyde in healthy subjects and in patients with contact dermatitis. Immunologia Clinica 11:53-58.
Gottschling, LM; Beaulieu, HJ; Melvin, WW. (1984) Monitoring of formic acid in urine of humans exposed to low
levels of formaldehyde. Am Ind Hyg Assoc J 45:19-23.
Grafstrom, RC; Fornace, AJ, Jr.; Autrup, H; et al. (1983) Formaldehyde damage to DNA and inhibition of DNA
repair in human bronchial cells. Science 220:216-218.
This document is a draft for review purposes only and does not constitute Agency policy.
R- 17 DRAFT—DO NOT CITE OR QUOTE
-------
Grafstrom, RC; Fornace, A, Jr.; Harris, CC. (1984) Repair of DNA damage caused by formaldehyde in human cells.
Cancer Res 44:4323-4327.
Grafstrom, RC; Curren, RD; Yang, LL; et al. (1985) Genotoxicity of formaldehyde in cultured human bronchial
fibroblasts. Science 228:89-91.
Grafstrom, RC. (1990) In vitro studies of aldehyde effects related to human respiratory carcinogenesis. Mutat Res
238:175-84.
Grafstrom, RC; Hsu, 1C; and Harris, CC. (1993) Mutagenicity of formaldehyde in Chinese hamster lung fibroblasts:
synergy with ionizing radiation and N-nitroso-N-methylurea. Chem Biol Interact 86:41-49.
Grammer, LC; Harris, KE; Shaughnessy, M; et al. (1990) Clinical and immunologic evaluation of 37 workers
exposed to gaseous formaldehyde. J Allergy Clin Immunol 86:177-181.
Grammer, LC; Harris, KE; Cugell, DW; et al. (1993) Evaluation of a worker with possible formaldehyde-induced
asthma. J Allergy Clin Immunol 92:29-33.
Grasl-Kraupp, B; Luebeck, G; Wagner, A; et al. (2000) Quantitative analysis of tumor initiation in rat liver: Role of
cell replication and cell death (apoptosis). Carcinogenesis 21:1411-1421.
Graves, RJ; Callander, RD; and Green, T. (1994) The role of formaldehyde and S-chloromethylglutathione in the
bacterial mutagenicity of methylene chloride. Mutat Res 320:235-243.
Graves, RJ; Trueman, P; Jones, S; et al. (1996) DNA sequence analysis of methylene chloride-induced HPRT
mutations in Chinese hamster ovary cells: comparison with the mutation spectrum obtained for 1,2-dibromoethane
and formaldehyde. Mutagenesis 11:229-233.
Gray, RH; Wu, LY. (2000) Subfertility and risk of spontaneous abortion. Am J Public Health 90:1452-1454.
Grazuleviciene, R; Dulskiene, V; Vencloviene, J. (1998) Formaldehyde exposure and low birth weight incidence. J
Occup Health 40:61-67.
Greaves, MA. (2004) Biological models for leukemia and lymphoma. IARC Sci Publ (157):351-372.
Green, DJ; Sauder, LR; Kulle, TJ; et al. (1987) Acute response to 3.0 ppm formaldehyde in exercising healthy
nonsmokers and asthmatics. Am Rev RespirDis 135:1261-1266.
Green, DJ; Bascom, R; Healey, EM; et al. (1989) Acute pulmonary response in healthy, nonsmoking adults to
inhalation of formaldehyde and carbon. Toxicol Environ Health 28:261-275.
Griesemer, RA; Boreiko, CJ; de Serres, FJ; et al. (1985) Report on the consensus workshop on formaldehyde:
Carcinogenicity/Histopathology/Genotoxicity Panel report. Environ Health Perspect 58:340-343.
Grosjcan D. RD. Ellis C. (1983) Carbonyls in Los Angeles air: of direct
photochemistry. Sci Total Environ 29:65-85,
Guess, H; Crump, K; Peto, R. (1977) Uncertainty estimates for low dose-rate extrapolations of animal
carcinogenicity data. Cancer Res 37:3475-3483.
Guseva, VA. (1972) Study of the gonadotropic effect in male rats under the effect of formaldehyde administered
simultaneously with air and water [translation]. Gig Sanit 37:101-102.
Gustavsson, P; Jakobsson, R; Johannson, H; et al. (1998) Occupational exposures and squamous cell carcinoma of
the oral cavity, pharynx, larynx, and oesophagus: a case-control study in Sweden. Occup Environ Med 55:393-400.
This document is a draft for review purposes only and does not constitute Agency policy.
R. i g DRAFT—DO NOT CITE OR QUOTE
-------
Hagiwara, M; Watanabe, E; Barrett, JC, et al. (2006) Assessment of genotoxicity of 14 chemical agents used in
dental practice: ability to induce chromosome aberrations in Syrian hamster embryo cells. Mutat Res 603:111-120.
Hakim, RB; Gray, RH; Zacur, H. (1995) Infertility and early pregnancy loss. Am J Obstet Gynecol 172:1510-1517.
Hall, A; Harrington, J; Aw, TC. (1991) Mortality study of British pathologists. Am J Ind Med 20:83-89.
Hamaguchi, F; and Tsutsui, T. (2000) Assessment of genotoxicity of dental antiseptics: ability of phenol, guaiacol,
p-phenolsulfonic acid, sodium hypochlorite, p-chlorophenol, m-cresol or formaldehyde to induce unscheduled DNA
synthesis in cultured Syrian hamster embryo cells. Jpn J Pharmacol 83:273-276.
Hanahan, D; Weinberg, RA. (2000) The hallmarks of cancer. Cell 100:57-70.
Hanrahan, LP; Dally, KA; Anderson, HA; et al. (1984) Formaldehyde vapor in mobile homes: a cross sectional
survey of concentrations and irritant effects. Am J Public Health 74(9): 1026-1027.
Hansel, A; Jordan, A; Holzinger, R; et al. (1995) Proton transfer reaction-mass spectrometry: on-
line trace gas analysis at the ppb level. Int. J. Mass Spectrom. Ion Process 149-150:609-619.
Hansen, J; Olsen, JH. (1995) Formaldehyde and cancer morbidity among male employees in Denmark. Cancer
Causes Control 6:354-360.
Hanzlik, RP; Fowler, SC; Eells, JT. (2005) Absorption and elimination of formate following oral administration of
calcium formate in female human subjects. Drug Metab Dispos 33:282-286.
Harrington, JM; Shannon, H. (1975) Mortality study of pathologists and medical laboratory technicians. BrMed J
4:329-332.
Harrington, JM; Oakes, D. (1984) Mortality study of British pathologists 1974-80. Br J Ind Med 41:188-191.
Harris, NL; Jaffe, ES; Diebold, J; et al. (2000a) The World Health Organization classification of neoplasms of the
hematopoietic and lymphoid tissues: report of the Clinical Advisory Committee meeting—Airlie House, Virginia,
November, 1997. Hematol J 1:53-66.
Harris, NL; Jaffe, ES; Diebold, J; et al. (2000b) The World Health Organization classification of hematological
malignancies report of the Clinical Advisory Committee meeting, Airlie House, Virginia, November 1997. Mod
Pathol 13:193-207.
Harris, C; Wand, S-W; Lauchu, JJ; et al. (2003) Methanol metabolism and embryotoxicity in rat and mouse
conceptuses: comparison of alcohol dehydrogenase (ADH1), formaldehyde dehydrogenase (ADH3), and catalase.
Reprod Toxicol 17:349-357.
Harving, H; Korsgaard, J; Pedersen, OF; et al. (1990) Pulmonary function and bronchial reactivity in asthmatics
during low-level formaldehyde exposure. Lung 168:15-21.
Haseman, JK. (1995) Data analysis: Statistical analysis and use of historical control data. Regul Toxicol Pharmacol
21:52-59.
Hauptmann, M; Lubin, JH; Stewart, PA; et al. (2003) Mortality from lymphohematopoietic malignancies among
workers in formaldehyde industries. J Natl Cancer Inst 95:1615-1623.
Hauptmann, M; Lubin, JH; Stewart, PA; et al. (2004a) Mortality from solid cancers among workers in formaldehyde
industries. AmJEpidemiol 159:1117-1130.
Hauptmann, M; Lubin, JH; Stewart, PA; et al. (2004b) Response. J Natl Cancer Inst 96:967-968.
This document is a draft for review purposes only and does not constitute Agency policy.
R- 19 DRAFT—DO NOT CITE OR QUOTE
-------
Hauptmann, M; Stewart, P; Lubin, J; et al. (2009) Mortality from lymphohematopoietic malignancies and brain
cancer among embalmers exposed to formaldehyde. JNCI101:1-13
Haworth, S; Lawlor, T; Mortelmans, K; et al. (1983) Salmonella mutagenicity test results for 250 chemicals.
Environ Mutagen 5 Suppl 1:1-142.
Hayashi, H; Kunugita, N; Arashidani, K; et al. (2004) Long-term exposure to low levels of formaldehyde increases
the number of tyrosine hydroxylase-immunopositive periglomerular cells in mouse main olfactory bulb. Brain Res
1007:192-197.
Hayes, RB; Raatgever, JW; de Bruyn, A; et al. (1986) Cancer of the nasal cavity and paranasal sinuses, and
formaldehyde exposure. Int J Cancer 37:487^192.
Hayes, RB; Blair, A; Stewart, PA; et al. (1990) Mortality of U.S. embalmers and funeral directors. Am J Ind Med
18:641-652.
Hayes, RB; Klein, S; Suruda, A, et al. (1997) O6-alkylguanine DNA alkyltransferase activity in student embalmers.
AmJIndMed31:361-365.
He, JL; Jin, LF; Jin, HY. (1998) Detection of cytogenetic effects in peripheral lymphocytes of students exposed to
formaldehyde with cytokinesis-blocked micronucleus assay. Biomed Environ Sci 11:87-92.
Health Canada and Environment Canada. (2001) Priority substances list assessment report: formaldehyde. Hull,
Quebec, Canada: Environment Canada and Health Canada.
Heck, HD, Casanova-Schmitz, M. (1984) The relevance of disposition studies to the toxicology of formaldehyde.
CUT Activities 4:1-5.
Heck HD; Casanova, M. (1987) Isotope effects and their implications for the covalent binding of inhaled [3H]- and
[14C]formaldehyde in the rat nasal mucosa. Toxicol Appl Pharmacol 89:122-134.
Heck, HD; Casanova, M. (1999) Pharmacodynamics of formaldehyde: applications of a model forthe arrest of DNA
replication by DNA-protein cross-links. Toxicol Appl Pharmacol 160:86-100.
Heck, HD, Casanova, M. (2004) The implausibility of leukemia induction by formaldehyde: A critical review of the
biological evidence on distant-site toxicity. Regul Toxicol Pharmacol 40:92-106.
Heck, HD; White, EL; Casanova-Schmitz, M. (1982) Determination of formaldehyde in biological tissues by gas
chromatography/mass spectrometry. Biomed Mass Spectrom 9:347-353.
Heck, HD; Chin, TY; Schmitz, MC. (1983) Distribution of [14C] formaldehyde in rats after inhalation exposure. In:
Gibson, J E, eds. Formaldehyde Toxicity. Washington, DC: Hemisphere Publishing Corporation; pp. 26-37.
Heck Hd, A; Casanova, M; L., C-W, et al. (1986). The formation of DNA-protein crosslinks by aldehydes present in
tobacco smoke. In: Hoffman, D and CC Harris,ed, Mechanisms in Tobacco Carcinogenesis: Cold Spring Harbor,
New York: CSH Press, 215-230.
Heck, HD; Casanova-Schmitz, M; Dodd, PB; et al. (1985) Formaldehyde (CH2O) concentrations in the blood of
humans and Fischer-344 rats exposed to CH2O under controlled conditions. Am Ind Hyg Assoc J 46:1-3.
Heck, HD; Casanova, M; Starr, TB. (1990) Formaldehyde toxicity~new understanding. Crit Rev Toxicol
20:397-426.
Hedberg, JJ; Heeg, JO; Nilsson, JA; et al. (2000) Expression of alcohol dehydrogenase 3 in tissue and cultured cells
from human oral mucosa. Am J Pathol 157:1745-1755.
This document is a draft for review purposes only and does not constitute Agency policy.
R-20 DRAFT—DO NOT CITE OR QUOTE
-------
Hedberg, JJ; Grafstrom, RC; Vondracek, M; et al. (200la) Micro-array chip analysis of carbonyl-metabolizing
enzymes in normal, immortalized and malignant human oral keratinocytes. Cell Mol Life Sci 58:1719-1726.
Hedberg, JJ; Backlund, M; Stromberg, P; et al. (200 Ib) Functional polymorphism in the alcohol dehydrogenase 3
(ADH3) promoter. Pharmacogenetics 11:815-824.
Hedberg, JJ; Griffiths, WJ; Nilsson, SJ; et al. (2003) Reduction of S-nitrosoglutathione by human alcohol
dehydrogenase 3 is an irreversible reaction as analyzed by electrospray mass spectrometry. Eur J Biochem
270:1249-1256.
Heikkila, P; Priha, E; Savola, A. (1991) Exposure at work. 14. Formaldehyde. Institute of Occupational Health.
Finnish Work Environmental Fund.
Heineman, EF; Olsen, JH; Pottern, LM; et al. (1992) Occupational risk factors for multiple myeloma among Danish
men. Cancer Causes Control 3:555-568.
Hemminki, K; Mutanen, P; Salonienmi, I; et al. (1982) Spontaneous abortions in hospital staff engaged in sterilizing
instruments with chemical agents. Br Med J 285:1461-1463.
Hemminki, K; Kyyronen, P; Lindbohm, ML. (1985) Spontaneous abortions and malformations in the offspring of
nurses exposed to anaesthetic gases, cytostatic drugs, and other potential hazards in hospitals, based on registered
information of outcome. JEpidemiol Community Health 39:141-147.
Hendrick, DJ; Lane, DJ. (1975) Formalin asthma in hospital staff. Br Med J 1(5958):607-608.
Hendrick, DJ; Lane, DJ. (1977) Occupational formalin asthma. Br J Ind Med 34:11-18.
Hendrick, DJ; Rando, RJ; Lane, DJ; et al. (1982) Formaldehyde asthma: challenge exposure levels and fate after five
years. JOccup Med 24:893-897.
Herbert, FA; Hessel, PA; Melenka, LS; et al. (1994) Respiratory consequences of exposure to wood dust and
formaldehyde of workers manufacturing oriented strand board. Arch Environ Health 49(6):465-470.
Hernandez, O; Rhomberg, L; Hogan, K; et al. (1994) Risk assessment of formaldehyde. J Hazard Mater 39:161-172.
Hess, DT; Matsumoto, A; Kim, SO; et al. (2005) Protein S-nitrosylation: purview and parameters. Nat Rev Mol Cell
Biol 6:150-166.
Hester, SD; Benavides, GB; Yoon, L; et al. (2003) Formaldehyde-induced gene expression in F344 rat nasal
respiratory epithelium. Toxicology 187:13-24.
Hester, SD; Barry, WT; Zou, F; et al. (2005) Transcriptomic analysis of F344 rat nasal epithelium suggests that the
lack of carcinogenic response to glutaraldehyde is due to its greater toxicity compared to formaldehyde. Toxicol
Pathol 33:415^24.
Hikiba, H; Watanabe, E; Barrett, JC; et al. (2005) Ability of fourteen chemical agents used in dental practice to
induce chromosome aberrations in Syrian hamster embryo cells. J Pharmacol Sci 97:146-152.
Hildesheim, A; Dosemeci, M; Chan, CC; et al. (2001) Occupational exposure to wood, formaldehyde, and solvents
and risk of nasopharyngeal carcinoma. Cancer Epidemiol Biomarkers Prev 10:1145-1153.
Hileman, B. (1984) Formaldehyde: assessing the risk. Environ Sci Technol 18:216A-221A.
Hilton, J; Dearman, RJ; Harvey, P; et al. (1998) Estimation of relative skin sensitizing potency using the local lymph
node assay: a comparison of formaldehyde with glutaraldehyde. Am J Contact Dermat 9:29-33.
This document is a draft for review purposes only and does not constitute Agency policy.
R-21 DRAFT—DO NOT CITE OR QUOTE
-------
Hines, RN; McCarver, DG. (2002) The ontogeny of human drag-metabolizing enzymes: phase I oxidative enzymes.
J Pharmacol Exp Ther 300:361-366.
Holly, EA; Aston, DA; Ahn, DK; et al. (1996) Intraocular melanoma linked to occupations and chemical exposures.
Epidemiology 7:55-61.
Holmstrom, M; Wilhelmsson, B. (1988) Respiratory symptoms and pathophysiological effects of occupational
exposure to formaldehyde and wood dust. Scand J Work Environ Health 14:306-311.
Holmstrom, M; Wilhelmsson, B; Hellquist, H. (1989a) Histological changes in the nasal mucosa in rats after long-
term exposure to formaldehyde and wood dust. Acta Otolaryngol 108:274-283.
Holmstrom, M; Rynnel-Dagoo, B; Wilhelmsson, B. (1989b) Antibody production in rats after long-term exposure to
formaldehyde. Toxicol Appl Pharmacol 100:328-333.
Holmstrom, M; Wilhelmsson, B; Hellquist, H; et al. (1989c) Histological changes in the nasal mucosa in persons
occupationally exposed to formaldehyde alone and in combination with wood dust. Acta Otolaryngol 107:120-129.
Holness, DL; Nethercott, JR. (1989) Health status of funeral service workers exposed to formaldehyde. Arch
Environ Health 44:222-228.
Hoog, JO; Hedberg, JJ; Stromberg, P; et al. (2001) Mammalian alcohol dehydrogenase -functional and structural
implications. J Biomed Sci 8:71-76.
Hoog, JO; Stromberg, P; Hedberg, JJ; et al. (2003) The mammalian alcohol dehydrogenases interact in several
metabolic pathways. Chem Biol Interact 143-144:175-181.
Hoogenveen, R; Clewell, H; Andersen, M; et al. (1999) An alternative exact solution of the two-stage clonal growth
model of cancer. Risk Anal 19:9-14.
Horton, A; Tye, R; Stemmer, K. (1963) Experimental carcinogenesis of the lung. Inhalation of gaseous
formaldehyde or an aerosol of coal tar by C3H mice. J Natl Cancer Inst 30:31^3.
Horvath, E, Jr; Anderson, J; Pierce, WE; et al. (1988) Effects of formaldehyde on the mucous membranes and lungs:
A study of an industrial population. JAMA 259:701-707.
Hou, Y; Hu, WG; Hirano, T; et al. (2002) A new intra-NALT route elicits mucosal and systemic immunity against
Moraxella catarrhalis in a mouse challenge model. Vaccine 20:2375-2381.
HSDB (Hazardous Substances Data Bank). (2001) Formaldehyde. National Library of Medicine, Bethesda, MD.
Available from .
Hurni, H; Ohder, H. (1973) Reproduction study with formaldehyde and hexamethylenetetramine in beagle dogs.
Food Cosmet Toxicol 11:459-462.
IARC (International Agency for Research on Cancer). (1995) IARC monographs on the evaluation of carcinogenic
risks to humans. Vol.62. Wood dust and formaldehyde. Lyon, France: International Agency for Research on
Cancer. Available from and
.
IARC (2006) IARC Monographs on the evaluation of carcinogenic risks to humans. Formaldehyde, 2-
Butoxyethanol and l-tert-Butoxy-2-propanol, vol. 88. IARC, Lyon, France.
http://monographs.iarc.fr/ENG/Monographs/vol88/mono88-6A.pdf
larmarcovai, G; Botta, A; Orsiere, T. (2006a) Number of centromeric signals in micronuclei and mechanisms of
aneuploidy. Toxicol Lett 166:1-10.
This document is a draft for review purposes only and does not constitute Agency policy.
R-22 DRAFT—DO NOT CITE OR QUOTE
-------
larmarcovai, G; Orsiere, T; Sari-Minodier, I; et al. (2006b) Different pathways of centromeric micronuclei formation
in peripheral blood lymphocytes of cancer patients and workers occupationally exposed to mutagens/carcinogens?
Advances in Molecular and Clinical Aspects of Cancer Prevention,4th International Conference on Cancer
Prevention 2006, 4:54-55, Switzerland.
larmarcovai, G; Bonassi, S; Sari-Minodier, I; et al. (2007) Exposure to genotoxic agents, host factors, and lifestyle
influence the number of centromeric signals in micronuclei: a pooled re-analysis. Mutat Res 615:18-27.
Ibanes, JD; Leininger, JR; Jarabek, AM; et al. (1996) Reexamination of respiratory tract responses in rats, mice, and
rhesus monkeys chronically exposed to inhaled chlorine. Inhal Tox 8:859-876.
ICRP (International Commission on Radiological Protection). (1994) Human respiratory tract model for radiological
protection. Publication 66 Annals of the ICRP Vol 24; Pergamon Press, Oxford, UK.
Ihrig, A; Hoffmann, J; Triebig, G. (2006) Examination of the influence of personal traits and habituation on the
reporting of complaints at experimental exposure to ammonia. Int Arch Occup Environ Health 79:332-338.
ILSI (International Life Sciences Institute). (2003) Final report of a workshop to develop a framework for assessing
risks to children from exposures to environmental agents. ILSI Risk Science Institute, Washington, DC.
Im, H; Oh, E; Mun, J; et al. (2006) Evaluation of lexicological monitoring markers using proteomic analysis in rats
exposed to formaldehyde. J Proteome Res 5:1354-1366.
lonescu, J; Marinescu, D; Tapu, V; et al. (1978) Experimental chronic obstructive lung disease. I.
Bronchopulmonary changes induced in rabbits by prolonged exposure to formaldehyde. Morphol Embryol
24:233-242.
IPCS (International Programme on Chemical Safety). (1989) Environmental Health Criteria. Vol. 89.
Formaldehyde. Geneva, Switzerland: World Health Organization. Available from
.
IPCS (International Programme on Chemical Safety). (1997) Environmental Health Criteria 196 ~ Methanol 180.
IPCS. (2002) Formaldehyde. Concise international chemical assessment document. Vol. 40. World Health
Organization, Geneva, Switzerland. Available online at
http://www.inchem.org/documents/cicads/cicads/cicad40.htm.
Iravani, J. (1974) Clearance function of the respiratory ciliated epithelium in normal and bronchitic rats. Inhaled
Part 1:143-148.
Iravani, J; Melville, GN. (1974) Long-term effect of cigarette smoke on mucociliary function in animals.
Respiration 31:358-366.
Ishidate, JM; Sofuni, T; and Yoshikawa, K. (1981) Chromosomal aberration tests in vitro as a primary screening
tool for environmental mutagens and/or carcinogens. GANN Monogr Cancer Res 27:95-108.
Ito, K; Sakamoto, T; Hayashi, Y; et al. (1996) Role of tachykinin and bradykinin receptors and mast cells in gaseous
formaldehyde-induced airway microvascular leakage in rats. Eur J Pharmacol 307:291-298.
Jaeger, RJ; Gearhart, JM. (1982) Respiratory and metabolic response of rats and mice to formalin vapor. Toxicology
25:299-309.
Jaiswal, S; Traver, D; Miyamoto, T; et al. (2003) Expression of BCR/ABL and BCL-2 in myeloid progenitors leads
to myeloid leukemias. Proc Natl Acad Sci USA 100:10002-10007.
Jakab, GJ. (1992) Relationship between carbon black particulate-bound formaldehyde, pulmonary antibacterial
defenses, and alveolar macrophage phagocytosis. Inhal Toxicol 4:325-342.
This document is a draft for review purposes only and does not constitute Agency policy.
R-23 DRAFT—DO NOT CITE OR QUOTE
-------
James, JT; Limero, TF; Beck, SW; et al. (2002) Environmental Monitoring Air Quality. Report of the Lunar-Mars
Support Test Project (LMLSTP4.1). In: Isolation: NASA Experiments in Closed Environment Living. Eds: Lane,
HW; Sauer, RL; Feeback, DL. Univeit, Inc, San Diego, CA. 104:177-191. Data also available at National
Aeronautics and Space Administration Life Sciences Data Archive, Houston, TX. Available at
http://lsda.jsc.nasa. gov/docs/research/research_detail.cfm?experiment_type_code=24&researchtype=.
Jarabek, AM; Fisher, JW; Rubenstein, R; et al. (1994) Mechanistic insights aid the search for CFC substitutes: risk
assessment of HCHC-123 as an example. Risk Anal 14:231-250.
Javdan, M; Taher, TE. (2000) The short and long term effects of formaldehyde vapour (2 ppm and 5 ppm) on rat
nasal epithelium. Pathol Int 50:A77.
Jeffcoat, AR; Chasalow, F; Feldman, DB; et al. (1983) Disposition of [14C] formaldehyde after topical exposure to
rats, guinea pigs, and monkeys. In: Gibson, J E , eds. Formaldehyde Toxicity. Washington, DC: Hemisphere; pp.
38-50.
Jensen, DE; Belka, GK; Du Bois, GC. (1998) S-Nitrosoglutathione is a substrate for rat alcohol dehydrogenase class
III isoenzyme. Biochem J 331 (Pt 2):659-668.
Johannsen, FR; Levinskas, GJ; Tegeris, AS. (1986) Effects of formaldehyde in the rat and dog following oral
exposure. Toxicol Lett 30:1-6.
John, EM; Savitz, DA; Shy, CM. (1994) Spontaneous abortions among cosmetologists. Epidemiology 5:147-155.
Joos, GF. (2003) Bronchial hyperresponsiveness: too complex to be useful? Curr Opin Pharmacol 3:233-238.
Jornvall, H; Hoog, JO; Persson, B; et al. (2000) Pharmacogenetics of the alcohol dehydrogenase system.
Pharmacology 61:184-191.
Junqueira, LC; Carneiro, J; Kelley, RO. (1992) Basic Histology. Appleton and Lange, Norwalk, CN.
Jung, R; Engelhart, G; Herbolt, B; et al. (1992) Collaborative study of mutagenicity with Salmonella typhimurium
TA102. Mutat Res 278:265-270.
Jurvelin, J; Vartiainen, M; Jantunen, M; et al. (2001) Personal exposure levels and microenvironmental
concentrations of formaldehyde and acetaldehyde in the Helsinki metropolitan area. J Air Waste Manag Assoc
51:17-24.
K S Crump Group. (2001) TOX_RISK dose-response software, version 5.2 (2000-2001). KS Crump Group.
Kakeyama, M; Kurokawa, Y; Kunugita, N; et al. (2004) Low dose exposure to formaldehyde alters glutamate
NMD A receptor mRNA expressions in the rat neocortex. Neurotoxicology 25:717.
Kalasz, H. (2003) Biological role of formaldehyde, and cycles related to methylation, demethylation, and
formaldehyde production. Mini Rev Med Chem 3:175-192.
Kallen, RG; Jencks, WP. (1966a) The dissociation constants of tetrahydrofolic acid. J Biol Chem 241:5845-5850.
Kallen, RG; Jencks, WP. (1966b) The mechanism of the condensation of formaldehyde with tetrahydrofolic acid. J
Biol Chem 241:5851-5863.
Kamata, E; Nakadate, M; Ogawa, Y; et al. (1996a) Acute inhalation toxicity study of formaldehyde in rats: effect of
vapor on the pulmonary surfactant. Oyo Yakuri 51:33-37.
Kamata, E; Nakadate, M; Uchida, O; et al. (1996b) Effects of formaldehyde vapor on the nasal cavity and lungs of
F-344 rats. J Environ Pathol Toxicol Oncol 15:1-8.
This document is a draft for review purposes only and does not constitute Agency policy.
R-24 DRAFT—DO NOT CITE OR QUOTE
-------
Kamata, E; Nakadate, M; Uchida, O; et al. (1997) Results of a 28-month chronic inhalation toxicity study of
formaldehyde in male Fisher-344 rats. J Toxicol Sci 22:239-254.
Kane, LE; Alarie, Y. (1977) Sensory irritation to formaldehyde and acrolein during single and repeated exposures in
mice. Am Ind Hyg Assoc J 38:509-522.
Kane, LE; Alarie, Y. (1978) Evaluation of sensory irritation from acrolein-formaldehyde mixtures. Am Ind Hyg
Assoc J 39:270-274.
Kane, LE; Barrow, CS; Alarie, Y. (1979) A short-term test to predict acceptable levels of exposure to airborne
sensory irritants. Am Ind Hyg Assoc J 40:207-229.
Kedderis, GL; Hollenberg, PF. (1983) Steady state kinetics of chloroperoxidase-catalyzed N-demethylation
reactions. J Biol Chem 258:12413-12419.
Kelly, JT; Prasad, AK; Wexler, AS. (2000) Detailed flow patterns in the nasal cavity. J Appl Physiol 89:323-337.
Kennedy, G; Slaich, PK; Golding, BT; et al. (1996) Structure and mechanism of formation of a new adduct from
formaldehyde and guanosine. Chem Biol Interact 102:93-100.
Kepler, GM; Richardson, RB; Morgan, KT; et al. (1998) Computer simulation of inspiratory nasal airflow and
inhaled gas uptake in a rhesus monkey. Toxicol Appl Pharmacol 150:1-11.
Kernan, GJ; Ji, BT; Dosemeci, M; et al. (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.
Kerns, WD; Pavkov, KL; Donofrio, DJ; et al. (1983) Carcinogenicity of formaldehyde in rats and mice after long-
term inhalation exposure. Cancer Res 43:4382-4392.
Khamgaonkar, MB; Fulare, MB. (1991) Pulmonary effects of formaldehyde exposure—an environmental-
epidemiological study. Ind J Chest Dis Allied Sci 33:9-13.
Kiernan, JA. (2000) Formaldehyde, formalin, paraformaldehyde and glutaraldehyde: What they are and what they
do. Microscopy Today 00-1 pp. 8-12.
Kilburn, KH. (1994) Neurobehavioral impairment and seizures from formaldehyde. Arch Environ Health 49:37-44.
Kilburn, KH; Moro, A. (1985) Reproductive and maternal effects of formaldehyde (HCHO) in rats. Fed Proc 44:535.
Kilburn, KH; Warshaw, R. (1992) Neurobehavioral effects of formaldehyde and solvents on histology technicians:
repeated testing across time. Environ Res 58:134-146.
Kilburn, KH; Seidman, BC; Warshaw, R. (1985) Neurobehavioral and respiratory symptoms of formaldehyde and
xylene exposure in histology technicians. Arch Environ Health 40:229-233.
Kilburn, KH; Warshaw, R; Thornton, JC. (1987) Formaldehyde impairs memory equilibrium and dexterity in
histology technicians effects which persist for days after exposure. Arch Environ Health 42:117-120.
Kim, CW; Song, JS; Ahn, YS; et al. (2001) Occupational asthma due to formaldehyde. Yonsei Med J 42:440-445.
Kim, KC; Hisatsune, A; Kim, DJ; et al. (2003) Pharmacology of airway goblet cell mucin release. J. Pharmacol Sci
92:301-307.
Kimbell, JS; Gross, EA; Joyner, DR; et al. (1993) Application of computational fluid dynamics to regional dosimetry
of inhaled chemicals in the upper respiratory tract of the rat. Toxicol Appl Pharmacol 121:253-263.
Kimbell, JS; Godo, MN; Gross, EA; et al. (1997a) Computer simulation of inspiratory airflow in all regions of the
F344 rat nasal passages. Toxicol Appl Pharmacol 145:388-398.
This document is a draft for review purposes only and does not constitute Agency policy.
R-25 DRAFT—DO NOT CITE OR QUOTE
-------
Kimbell, JS; Gross, EA; Richardson RB; et al. (1997b) Correlation of regional formaldehyde flux predictions with
the distribution of formaldehyde-induced squamous metaplasia in F344 rat nasal passages. Mutat Res 380:143-154.
Kimbell, JS; Subramaniam, RP; Gross, EA; et al. (2001a) Dosimetry modeling of inhaled formaldehyde:
comparisons of local flux predictions in the rat, monkey, and human nasal passages. Toxicol Sci 64:100-110.
Kimbell, JS; Overton, JH; Subramaniam, RP; et al. (200 Ib) Dosimetry modeling of inhaled formaldehyde: binning
nasal flux predictions for quantitative risk assessment. Toxicol Sci 64:111-121.
Kimbell, JS. (2006) Nasal dosimetry of inhaled gases and particles: Where do inhaled agents go in the nose?
Toxicol Pathol 34:270-273.
Kinzler, KW; Vogelstein, B. (1997) Cancer-susceptibility genes. Gatekeepers and caretakers. Nature 386:761-763.
Kitaev, EM; Savchenko, O; Lovchikov, VA; et al. (1984) Embryonic development and various indicators of
reproductive function in rats after inhalation of formaldehyde in the preimplantation phase [translation]. Akush
Ginekol 10:49-52.
Kitaeva, LV; Kitaev, E; Pimenova, MN. (1990) Cytopathic and cytogenetic effects of chronic inhalation of
formaldehyde on the female rat's germ marrow cells [translation]. Tsitologiia 32:1212-1216.
Kitaeva, LV; Mikheeva, EA; Shelomova, LF; et al. (1996) Genotoxic effect of formaldehyde in somatic human cells
in vivo [translation]. Genetika 32(9): 1287-1290. Translated from Russian by Ted Crump, Translator. NIH Library
Translation. National Institutes of Health, Public Health Service, U.S. Department of Health and Human Services;
NIH-03-640.
Klaassen, CD. (2001) Casarett and Doull's Toxicology: The Basic Science of Poisons. 6th Edition. McGraw-Hill,
New York, NY.
Klein, M; Sinha, B; Subramaniam, RP. (2010) Statistical inferences from formaldehyde DNA-protein cross-link
data: Improving methods for characterization of uncertainty. In Press, J. Biopharmaceutical Statistics. [To appear in
Vol 21(1)]
Kligerman, AD; Phelps, MC; Erexson, GL. (1984) Cytogenetic analysis of lymphocytes from rats following
formaldehyde inhalation. Toxicol Lett 21:241-246.
Klonisch, T; Steger, K; Kehlen, A; et al. (2004) INSL-3 ligand-receptor system in the equine testis. Biol Reprod
68:1975-1981.
Koppel, C; Tenczer, J; Peixoto-Menezes, KM. (1991) Formation of formaldehyde adducts from various drugs by use
of methanol in a lexicological screening procedure with gas chromatography-mass spectrometry. J Chromatogr
563:73-81.
Korczynski, RE. (1994) Formaldehyde exposure in the funeral industry. Appl Occup Environ Hyg 9:575-579.
Krakowiak, A; Gorski, P; Pazdrak, K; et al. (1998) Airway response to formaldehyde inhalation in asthmatic
subjects with suspected respiratory formaldehyde sensitization. Am J Ind Med 33:274-281.
Kramps, JA; Peltenburg, LT; Kerklaan, PR; et al. (1989) Measurement of specific IgE antibodies in individuals
exposed to formaldehyde. Clin Exp Allergy 19:509-514.
Krewski, D; Crump, KS; Farmer, J; et al. (1983) A comparison of statistical methods for low dose extrapolation
utilizing time-to-tumor data. Fundam Appl Toxicol 3:140-160.
Kriebel, D; Sama, SR; Cocanour, B. (1993) Reversible pulmonary responses to formaldehyde. A study of clinical
anatomy students. Am Rev Respir Dis 148:1509-1515.
This document is a draft for review purposes only and does not constitute Agency policy.
R-26 DRAFT—DO NOT CITE OR QUOTE
-------
Kriebel, D; Myers, D; Cheng, M; et al. (2001) Short-term effects of formaldehyde on peak expiratory flow and
irritant symptoms. Arch Environ Health 56(1): 11-18.
Kreiger, RA; and Garry, VF. (1983) Formaldehyde-induced cytotoxicity and sister-chromatid exchanges in human
lymphocyte cultures. MutatRes 120:51-55.
Krimsky, MM. (1986) Letter to the editor regarding Sallakumar et al., 1985. Toxicol Appl Pharmacol 85:295.
Krzyzanowski, M; Quackenboss, JJ; Lebowitz, MD. (1990) Chronic respiratory effects of indoor formaldehyde
exposure. Environ Res 52:117-125.
Kuby, J. (1991) Immunology. 2nd edition. New York, NY: W.H. Freeman and Company.
Kulle, TJ. (1993) Acute odor and irritation response in healthy nonsmokers with formaldehyde exposure. Inhal
Toxicol 5:323-332.
Kulle, TJ; Cooper, GP. (1975) Effects of formaldehyde and ozone on the trigeminal nasal sensory system. Arch
Environ Health 30:237-243.
Kulle, TJ; Sauder, LR; Hebel, R; et al. (1987) Formaldehyde dose-response in healthy nonsmokers. JAPCA 37:919-
924.
Kum, C; Kiral, F; Sekkin, S; et al. (2007) Effects of xylene and formaldehyde inhalations on oxidative stress in adult
and developing rats livers. Exp Anim 56:35-42.
Kuper, CF; Hameleers, DMH; Bruijntjes, JP; et al. (1990) Lymphoid and non-lymphoid cells in nasal-associated
lymphoid tissue (NALT) in the rat. Cell Tissue Res 259:371-377.
Kuppers, R; Klein, U; Hansmann, M-L; et al. (1999) Cellular origin of human B-cell lymphomas. New Eng J Med
341:1520-1529.
Kuschner, M; Laskin, S; Drew, RT; et al. (1975) Inhalation carcinogenicity of alpha halo ethers. III. Lifetime and
limited period inhalation studies with bis(chloromethyl)ether at 0.1 ppm. Arch Environ Health 30:73-77.
Kushch, I; Schwarz, K; Schwentner, L; et al. (2008) Compounds enhanced in a mass spectrometric profile of
smokers' exhaled breath versus non-smokers as determined in a pilot study using PTR-MS. J Breath Res 2:1-26.
Laforest, L; Luce, D; Goldberg, P; et al. (2000) Laryngeal and hypopharyngeal cancers and occupational exposure to
formaldehyde and various dusts: a case-control study in France. Occup Environ Med 57:767-773.
Lam, CW; Casanova, M; Heck, HD. (1985) Depletion of nasal mucosal glutathione by acrolein and enhancement of
formaldehyde-induced DNA-protein cross-linking by simultaneous exposure to acrolein. Arch Toxicol 58:67-71.
Lang, I; Bruckner, T; Triebig, G. (2008) Formaldehyde and chemosensory irritation in humans: A controlled human
exposure study. Regul Toxicol Pharmacol 50:23-36.
Lash, LH; Parker, JC. (2001) Hepatic and renal toxicities associated with perchloroethylene. Pharmacol Rev
53:177-208.
Laskin, S; Kuschner, M; Drew, RT; et al. (1971) Tumors of the respiratory tract induced by inhalation of
bis(chloromethyl)ether. Arch Environ Health 23:135-136.
Lazutka, JR; Lekevicius, R; Dedonyte, V, et al. (1999) Chromosomal aberrations and sister-chromatid exchanges in
Lithuanian populations: effects of occupational and environmental exposures. MutatRes 445:225-239.
Le Curieux, F; Marzin, D; Erb, F. (1993) Comparison of three short-term assays: results on seven chemicals.
Potential contribution to the control of water genotoxicity. Mutat Res 319:223-236.
This document is a draft for review purposes only and does not constitute Agency policy.
R-27 DRAFT—DO NOT CITE OR QUOTE
-------
Leach, C; Sharma, R; Oberg, S. (1983) Immunotoxic and pathologic effects of inhalation of various levels of
formaldehyde vapor in F-344 rats. Toxicologist 3:73.
Lee, H; Alarie, Y; Karol, M. (1984) Induction of formaldehyde sensitivity in guinea pigs. Toxicol Appl Pharmacol
75:147-155.
Lee, MH; Kim, YA; Na, TY; et al. (2008) Identification of formaldehyde-responsive genes by suppression
subtractive hybridization. Toxicology 243:224-235.
Lemiere, C; Desjardins, A; Cloutier, Y; et al. (1995) Occupational asthma due to formaldehyde resin dust with and
without reaction to formaldehyde gas. Eur Respir J 8:861-865.
Levine, RJ; Andjelkovich, DA; Shaw, L. (1984) The mortality of Ontario undertakers and a review of formaldehyde-
related mortality studies. J Occup Med 26:740-746.
Levy, S; Nocentini, S; and Billardon, C. (1983) Induction of cytogenetic effects in human fibroblast cultures after
exposure to formaldehyde or X-rays. Mutat Res 119:309-317.
Li. S.-M.: Anlauf. K, G.: Wicbc. H, A,; Bottcnhciin. J.W. (1994) Estimating primary secondary production of
HCHO in eastern North America based on gas-pliasc measurements component analysis. Gcopliys.
Res. Lett 21: 669-672.
Li, G-Y; Lee, H-Y; Shin, H-S; et al. (2007) Identification of gene markers for formaldehyde exposure in humans.
Environ Health Perspect 115:1460-1466.
Li, R; Lu, ZS; Qiao, Y; et al. (2004) Study on the formaldehyde-induced DNA damage with comet assay. Shi Yan
Sheng Wu Xue Bao 37:262-268.
Liber, HL; Benforado, K; Crosby, RM; et al. (1989) Formaldehyde-induced and spontaneous alterations in human
hprt DNA sequence and mRNA expression. Mutat Res 226:31-37.
Liden, S; Scheynius, A; Fischer, T; et al. (1993) Absence of specific IgE antibodies in allergic contact sensitivity to
formaldehyde. Allergy 48(7):525-529.
Liebling, T; Rosenman, KD; Griffith, RG; et al. (1984) Cancer mortality among workers exposed to formaldehyde.
Am J Ind Med 5:423-428.
Liesivuori, J; Savolainen, H. (1987) Urinary formic acid as an indicator of occupational exposure to formic acid and
methanol. Am Ind Hyg Assoc J 48:32-34.
Lin, Z; Luo, W; Li, H; et al. (2005) The effect of endogenous formaldehyde on the rat aorta endothelial cells.
Toxicol Lett 159:134-143.
Lindbohm, ML; Hemminki, K; Bonhomme, MG; et al. (1991) Effects of paternal occupational exposure on
spontaneous abortions. Am J Public Health 81:1029-1033.
Lino dos Santos Franco, A; Damazo, AS; Beraldo de Souza, HR; et al. (2006) Pulmonary neutrophil recruitment and
bronchial reactivity in formaldehyde-exposed rats are modulated by mast cells and differentially by neuropeptides
and nitric oxide. Toxicol Appl Pharmacol 214:35-42.
Lino dos Santos Franco, A; Domingos, HV; Damazo, AS; et al. (2009) Reduced allergic lung inflammation in rats
following formaldehyde exposure: long term effects on multiple effector systems. Toxicology. 256: 157-163.
Liu, KS; Huang, FY; Hayward, SB; et al. (1991) Irritant effects of formaldehyde exposure in mobile homes.
Environ Health Perspect 94:91-94.
This document is a draft for review purposes only and does not constitute Agency policy.
R-28 DRAFT—DO NOT CITE OR QUOTE
-------
Liu, L; Hausladen, A; Zeng, M; et al. (2001) A metabolic enzyme for S-nitrosothiol conserved from bacteria to
humans. Nature 410:490^94.
Liu, L; Yan, Y; Zeng, M; et al. (2004) Essential roles of S-nitrosothiols in vascular homeostasis and endotoxic
shock. Cell 116:617-628.
Liu, Y; Li, CM; Lu, Z; et al. (2006) Studies on formation and repair of formaldehyde-damaged DNA by detection of
DNA-protein crosslinks and DNA breaks. Front Biosci 11:991-997.
Logue, J; Barrick, M; Jessup, G. (1986) Mortality of radiologists and pathologists in the radiation registry of
physicians. JOccupMed28:91-99.
Lu, Z; Li, CM; Qiao, Y; et al. (2008) Effect of inhaled formaldehyde on learning and memory in mice. Indoor Air
18:77-83.
Luce, D; Gerin, M; Leclerc, A; et al. (1993) Sinonasal cancer and occupational exposure to formaldehyde and other
substances. Int J Cancer 53:224-231.
Luce, D; Leclerc, A; Begin, D; et al. (2002) Sinonasal cancer and occupational exposures: a pooled analysis of 12
case-control studies. Cancer Causes Control 13:147-157.
Luebeck, EG; Grasl-Kraupp, B; Timmermann-Trosiener, I; et al. (1995) Growth kinetics of enzyme-altered liver foci
in rats treated with phenobarbital or alpha-hexachlorocyclohexane. Toxicol Appl Pharmacol 130:304-315.
Luebeck, EG; Buchmann, A; Stinchcombe, S; et al. (2000) Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on
initiation and promotion of GST-P-positive foci in rat liver: A quantitative analysis of experimental data using a
stochastic model. Toxicol Appl Pharmacol 167:63-73.
Luo, W; Li, H; Zhang, Y; etal. (2001) Determination of formaldehyde in blood plasma by high-performance liquid
chromatography with fluorescence detection. J Chromatogr B Biomed Sci Appl 753:253-257.
Lyapina, M; Zhelezova, G; Petrova, E; et al. (2004) Flow cytometric determination of neutrophil burst activity in
workers exposed to formaldehyde. Int Arch Occup Environ Health 77:335-340.
Lindinger, W; Hansel, A; Jordan, A. (1998) On-line monitoring of volatile organic compounds at pptv levels by
means of Proton-Transfer-Reaction Mass-Spectrometry (PTR-MS): Medical applications, food control and
environmental research, Review paper. IntJMass Spectrom. lonProc 173:191-241.
Ma, TH; Harris, MM. (1988) Review of the genotoxicity of formaldehyde. Mutat Res 196:37-59.
Mackerer, CR; Angelosanto, FA; Blackburn, GR, et al. (1996) Identification of formaldehyde as the metabolite
responsible for the mutagenicity of methyl tertiary-butyl ether in the activiated mouse lymphoma assay. Proc. Soc.
exp. Biol. Med 212:338-341.
Magana-Schwencke, N; and Ekert, B. (1978) Biochemical analysis of damage induced in yeast by formaldehyde. II.
Induction of cross-links between DNA and protein. Mutat Res 51:11-19.
Main, D; Hogan, TJ. (1983) Health effects of low level exposure to formaldehyde. J Occup Med 25(12):896-900.
Majumder, PK; Kumar, VL. (1995) Inhibitory effects of formaldehyde on the reproductive system of male rats.
Indian J Physiol Pharmacol 39:80-82.
Malaka, T; Kodama, AM. (1990) Respiratory health of plywood workers occupationally exposed to formaldehyde.
Arch Environ Health 45:288-294.
Malek, FA; Moritz, KU; Fanghanel, J. (2003a) A study on the effect of inhalative formaldehyde exposure on water
labyrinth test performance in rats. Ann Anat 185:277-285.
This document is a draft for review purposes only and does not constitute Agency policy.
R-29 DRAFT—DO NOT CITE OR QUOTE
-------
Malek, FA; Moritz, KU; Fanghanel, J. (2003b) Formaldehyde inhalation and open field behaviour in rats. Indian J
Med Res 118:90-96.
Malek, FA; Moritz, KU; Fanghanel, J. (2003c) A study on specific behavioral effects of formaldehyde in the rat. J
Exp Anim Sci 42:160-170.
Malek, FA; Moritz, KU; Fanghanel, J. (2004) Effects of a single inhalative exposure to formaldehyde on the open
field behavior of mice. Int J Hyg Environ Health 207:151-158.
Mannix, RC; Phalen, RF; Walters, RF; et al. (1983) Effects of sulfur dioxide and formaldehyde on particle clearance
in the rat. J Toxicol Environ Health 12:429-440.
Marinari, UM; Ferro, M; Sciaba, L, et al. (1984) DNA-damaging activity of biotic and xenobiotic aldehydes in
Chinese hamster ovary cells. Cell Biochem Funct 2:243-248.
Marks, TA; Worthy, WC; Staples, RE. (1980) Influence of formaldehyde and sonacide (potentiated acid
glutaraldehyde) on embryo and fetal development in mice. Teratology 22:51-58.
Marnett, LJ; Kurd, HK; Hollstein, MC; et al. (1985) Naturally occurring carbonyl compounds are mutagens in
Salmonella tester strain TA104. MutatRes 148:25-34.
Maronpot, RR; Miller, RA; Clarke, WJ; et al. (1986) Toxicity of formaldehyde vapor in B6C3F1 mice exposed for
13 weeks. Toxicology 41:253-266.
Marsh, GM. (1982) Proportional mortality patterns among chemical plant workers exposed to formaldehyde. Br J
Ind Med 39:313-322.
Marsh, GM; Youk, AO. (2005) Reevaluation of mortality risks for nasopharyngeal cancer in the formaldehyde
cohort study of the National Cancer Institute. Regul Pharmacol Toxicol 42:275-283.
Marsh, GM; Stone, RA; Henderson, VL. (1992a) A reanalysis of the National Cancer Institute study on lung cancer
mortality among industrial workers exposed to formaldehyde. J Occup Med 34:42-44.
Marsh, GM; Stone, RA; Henderson, VL. (1992b) Lung cancer mortality among industrial workers exposed to
formaldehyde: a Poisson regression analysis of the National Cancer Institute Study. Am Ind Hyg Assoc J
53:681-691.
Marsh, GM; Stone, RA; Esmen, NA; et al. (1994) Mortality patterns among chemical plant workers exposed to
formaldehyde and other substances. J Natl Cancer Inst 86:384-386.
Marsh, GM; Stone, RA; Esmen, NA; et al. (1996) Mortality among chemical workers in a factory where
formaldehyde was used. Occup EnvironMed 53:613-627.
Marsh, GM; Youk, AO; Buchanich, JM; et al. (2002) Pharyngeal cancer mortality among chemical plant workers
exposed to formaldehyde. Toxicol Ind Health 18:257-268.
Marsh, GM; Youk, AO; Buchanich, JM; et al. (2007a) Work in the metal industry and nasopharyngeal cancer
mortality among formaldehyde-exposed workers. Regul Toxicol Pharmacol 48:308-319.
Marsh, GM; Youk, AO; Morfeld, P. (2007b) Mis-specified and non-robust mortality risk models for nasopharyngeal
cancer in the National Cancer Institute formaldehyde worker cohort study. Regul Toxicol Pharmacol 47:59-67.
Martin, CN; McDermid, AC; and Garner, RC. (1978) Testing of known carcinogens and noncarcinogens for their
ability to induce unscheduled DNA synthesis in HeLa cells. Cancer Res 38:2621-2627.
Martin, WJ. (1990) A teratology study of inhaled formaldehyde in the rat. Reprod Toxicol 4:237-239.
This document is a draft for review purposes only and does not constitute Agency policy.
R-30 DRAFT—DO NOT CITE OR QUOTE
-------
Mashford, PM; Jones, AR. (1982) Formaldehyde metabolism by the rat: a re-appraisal. Xenobiotica 12:119-124.
Matanoski, GM. (1989) Risks of pathologists exposed to formaldehyde. National Institute for Occupational Safety
and Health, Cincinnati, OH. Available from the National Technical Information Service, Springfield, VA; PB91-
173682.
Matsunaga, I; Miyake, Y; Yoshida, T; et al. (2008) Ambient formaldehyde levels and allergic disorders among
Japanese women: baseline data from the Osaka maternal and child health study. Ann Epidemiol 18:78-84.
Mauderly, JL. (1986) Respiration of F344 rats in nose-only inhalation exposure tubes. J Appl Toxicol 6:25-30.
Maurice, F; Rivory, JP; Larsson, PH; et al. (1986) Anaphylactic shock caused by formaldehyde in a patient
undergoing long-term hemodialysis. J Allergy Clin Immunol 77:594-597.
McGhee, JD; von Hippel, PH. (1975a) Formaldehyde as a probe of DNA structure. I. Reaction with exocyclic
amino groups of DNA bases. Biochemistry 14:1281-1296.
McGhee, JD; von Hippel, PH. (1975b) Formaldehyde as a probe of DNA structure. II. Reaction with endocyclic
imino groups of DNA bases. Biochemistry 14:1297-1303.
McGhee, JD; von Hippel, PH. (1977a) Formaldehyde as a probe of DNA structure. 3. Equilibrium denaturation of
DNA and synthetic polynucleotides. Biochemistry 16:3267-3276.
McGhee, JD; von Hippel, PH. (1977b) Formaldehyde as a probe of DNA structure, r. Mechanism of the initial
reaction of Formaldehyde with DNA. Biochemistry 16:3276-3293.
McGwin, G; Lienert, J; Kennedy, JL (2010) Formaldehyde Exposure and Asthma in Children: A Systematic Review.
Environ Health Perspect 118:313-317.
McKone, TE; Daniels, JL (1991) Estimating human exposure through multiple pathways from air, water, and soil.
Regul Toxicol Pharmacol 13:36-61.
McLean, JH; Shipley, MT. (1988) Postmitotic, postmigrational expression of tyrosine hydroxylase in olfactory bulb
dopaminergic neurons. J Neurosci 8:3658-3669.
McMartin, KE; Martin-Amat, G; Noker, PE; et al. (1979) Lack of a role for formaldehyde in methanol poisoning in
the monkey. Biochem Pharmacol 28:645-649.
Meggs, WJ. (1993) Neurogenic inflammation and sensitivity to environmental chemicals. Environ Health Perspect
101:234-238.
Meggs, WJ. (1995) Neurogenic switching: a hypothesis for a mechanism for shifting the site of inflammation in
allergy and chemical sensitivity. Environ Health Perspect 103:54-56.
Meng F, Bermudez E, McKinzie PB, Andersen ME, Clewell III HJ, Parsons BL. (2010). Measurement of tumor-
associated mutations in the nasal mucosa of rats exposed to varying doses of formaldehyde (2010). Regul Toxicol
Pharmacol. Corrected Proof, Available Online, DOL 10.1016/j.yrtph.2010.03.007.
Merk, O; Speit, G. (1998) Significance of formaldehyde-induced DNA-protein crosslinks for mutagenesis. Environ
Mol Mutagen 32:260-268.
Merk, O; and Speit, G. (1999) Detection of crosslinks with the comet assay in relationship to genotoxicity and
cytotoxicity. Environ Mol Mutagen 33:167-172.
Merletti, F; Boffetta, P; Ferro, G; et al. (1991) Occupation and cancer of the oral cavity or oropharynx in Turin,
Italy. Scand J Work Environ Health 17:248-254.
This document is a draft for review purposes only and does not constitute Agency policy.
R-31 DRAFT—DO NOT CITE OR QUOTE
-------
Mery, S; Gross, EA; Joyner, DR; et al. (1994) Nasal diagrams: a tool for recording the distribution of nasal lesions in
rats and mice. Toxicol Pathol 22:353-372.
Metz, B; Kersten, GF; Hoogerhout, P; et al. (2004) Identification of formaldehyde-induced modifications in proteins:
reactions with model peptides. J Biol Chem 279:6235-6243.
Metz, B; Kersten, GF; Baart, GJ; et al. (2006) Identification of formaldehyde-induced modifications in proteins:
reactions with insulin. Bioconjug Chem 17:815-822.
Migliore, L; Ventura, L; Barale, R; et al. (1989) Micronuclei and nuclear anomalies induced in the gastro-intestinal
epithelium of rats treated with formaldehyde. Mutagenesis 4:327-334.
Miretskaia, LV; and Shvartsman, P. (1982) [Chromosome damages in human lymphocytes as affected by
formaldehyde. I. Formaldehyde treatment of lymphocytes in culture]. Tsitologiia 24:1056-1060.
Miyachi, T; Tsutsui, T. (2005) Ability of 13 chemical agents used in dental practice to induce sister-chromatid
exchanges in Syrian hamster embryo cells. Odontology 93:24-29.
Molotkov, A; Fan, X; Deltour, L; et al. (2002) Stimulation of retinoic acid production and growth by ubiquitously
expressed alcohol dehydrogenase Adh3. Proc Natl Acad Sci USA 99:5337-5342.
Monteiro-Riviere, NA; Popp, JA. (1986) Ultrastructural evaluation of acute nasal toxicity in the rat respiratory
epithelium in response to formaldehyde gas. Fundam Appl Toxicol 6:251-262.
Monticello, TM; Morgan, KT. (1994) Cell proliferation and formaldehyde-induced respiratory carcinogenesis. Risk
Anal 14:313-319.
Monticello, TM; Morgan, KT. (1997) Chemically induced nasal carcinogenesis and epithelial cell proliferation: a
brief review. Mutat Res 380:33-41.
Monticello, TM; Morgan, KT; Everitt, JI; et al. (1989) Effects of formaldehyde gas on the respiratory tract of rhesus
monkeys. Pathology and cell proliferation. Am J Pathol 134:515-527.
Monticello, TM; Morgan, KT; Hurtt, ME. (1990) Unit length as the denominator for quantitation of cell proliferation
in nasal epithelia. Toxicol Pathol 18:24-31.
Monticello, TM; Miller, FJ; Morgan, KT. (1991) Regional increases in rat nasal epithelial cell proliferation
following acute and subchronic inhalation of formaldehyde. Toxicol Appl Pharmacol 111 :409^21.
Monticello, TM; Swenberg, JA; Gross, EA; et al. (1996) Correlation of regional and nonlinear formaldehyde-
induced nasal cancer with proliferating populations of cells. Cancer Res 56:1012-1022.
Moolgavkar, SH. (1994) Biological models of carcinogenesis and quantitative cancer risk assessment. Risk Anal
14:879-882.
Moolgavkar, SH; Verizon, DJ. (1979) Two-event models for carcinogenesis: Incidence curves for childhood and
adult tumors. Math Biosci 47:55-77.
Moolgavkar, SH; Knudson, AJ. (1981) Mutation and cancer: A model for human carcinogenesis. J Natl Cancer Inst
66:1037-1052.
Moolgavkar, SH; Luebeck, EG. (1992) Interpretation of labeling indices in the presence of cell death.
Carcinogenesis 13:1007-1010.
Moolgavkar, SH; Dewanji, A; Venzon, D. (1988) A stochastic two-stage model for cancer risk assessment. I. The
hazard function and the probability of tumor. Risk Anal 8:383-392.
This document is a draft for review purposes only and does not constitute Agency policy.
R-32 DRAFT—DO NOT CITE OR QUOTE
-------
Moolgavkar, SH; Luebeck, EG; Krewski, D; et al. (1993) Radon, cigarette smoke, and lung cancer: A reanalysis of
the Colorado plateau uranium miner's data. Epidemiology 4:204-217.
Moore, L; Ogrodnik, EC. (1986) Occupational exposure to formaldehyde in mortuaries. J Environ Health 49:32-35.
Moore, S; Montane-Jaime, K; Shafe, S; et al. (2007) Association of ALDH1 promoter polymorphisms with alcohol-
related phenotypes in Trinidad and Tobago. J Stud Alcohol Drugs 68:192-196.
Morgan, KT. (1997) A brief review of formaldehyde carcinogenesis in relation to rat nasal pathology and human
health risk assessment. Toxicol Pathol 25:291-307.
Morgan, KT; Patterson, DL; Gross, EA (1983) Localization of areas of inhibition of nasal mucociliary function in
rats following in vivo exposure to formaldehyde. Am Rev Respir Dis 127:166-166.
Morgan, KT; Patterson, DL; Gross, EA. (1984a) Frog palate mucociliary apparatus: structure, function, and response
to formaldehyde gas. Fundam Appl Toxicol 4:58-68.
Morgan, KT; Jiang, X-Z; Patterson, DL; et al. (1984b) The nasal mucociliary apparatus. Am Rev Respir Dis
130:275-281.
Morgan, KT; Gross, EA; Patterson, DL. (1986a) Distribution, progression, and recovery of acute formaldehyde-
induced inhibition of nasal mucociliary function in F-344 rats. Toxicol Appl Pharmacol. 86(3):448-456.
Morgan, KT; Jiang, XZ; Starr, TB; et al. (1986b) More precise localization of nasal tumors associated with chronic
exposure of F-344 rats to formaldehyde gas. Toxicol Appl Pharmacol 82:264-271.
Morgan, KT; Patterson, DL; Gross, EA. (1986c) Responses of the nasal mucociliary apparatus of F-344 rats to
formaldehyde gas. Toxicol Appl Pharmacol 82:1-13.
Morgan, K; Patterson, DL; Gross, E. (1986d) Responses of the nasal mucociliary apparatus to airborne irritants. In:
Barrow, CS; ed. Toxicology of the nasal passages. Washington, DC: Hemisphere Publishing Corporation; pp. 123-
133.
Morgan, KT; Kimbell, JS; Monticello, TM; et al. (1991) Studies of inspiratory airflow patterns in the nasal passages
of the F344 rat and rhesus monkey using nasal molds: relevance to formaldehyde toxicity. Toxicol Appl Pharmacol
110:223-240.
Moser, B; Bodrogi, F; Eibl, G; et al. (2005) Mass spectrometric profile of exhaled breath-field study by PTR-MS.
Respir Physiol Neurobiol 145:295-300.
Muller, W; Engelhart, G; Herbold, B; et al. (1993) Evaluation of mutagenicity testing with Salmonella typhimurium
TA102 in three different laboratories. Environ Health Perspect 101 Suppl 3:33-36.
Natarajan, AT; Darroudi, F; Bussman, CJ; et al. (1983) Evaluation of the mutagenicity of formaldehyde in
mammalian cytogenetic assays in vivo and vitro. Mutat Res 122:355-360.
NCHS (National Center for Health Statistics). (2002) United States Life Tables, 1999. National Vital Statistics
Reports, vol. 50, no. 6, March 21, pp. 7-8.
NCI (National Cancer Institute). (2003) SEER Cancer Statistics Review, 1975-2000. Bethesda, MD.
http://seer.cancer.gov/csr/1975_2000/.
NEG (Nordic Expert Group for Criteria Documentation of Health Risks from Chemicals); DECOS (Dutch Expert
Committee on Occupational Standards). (2003) 132. Formaldehyde. Stockholm, Sweden: National Institute for
Working Life. Available from. 2003: 11.
This document is a draft for review purposes only and does not constitute Agency policy.
R-33 DRAFT—DO NOT CITE OR QUOTE
-------
Neri, M; Bonassi, S; Knudsen, LE, et al. (2006) Children's exposure to environmental pollutants and biomarkers of
genetic damage. I. Overview and critical issues. Mutat Res 612:1-13.
Nielsen, GD. (1991) Mechanisms of activation of the sensory irritant receptor by airborne chemicals. Crit Rev
Toxicol 21:183-208.
Nielsen, GD; Hougaard, KS; Larsen, ST; et al. (1999) Acute airway effects of formaldehyde and ozone inBalb/c
mice. Hum Exp Toxicol 18:400^109.
Niinimaa, V; Cole, P; Mintz, S; etal. (1981) Oronasal distribution of respiratory airflow. Respir Physiol 43:69-75.
Nilsson, JA; Hedberg, JJ; Vondracek, M; et al. (2004) Alcohol dehydrogenase 3 transcription associates with
proliferation of human oral keratinocytes. Cell Mol Life Sci 61:610-617.
Norback, D; Bjornsson, E; Janson, C; et al. (1995) Asthmatic symptoms and volatile organic compounds,
formaldehyde, and carbon dioxide in dwellings. Occup Environ Med 52(6):388-395.
NRC (National Research Council). (1983) Risk assessment in the federal government: managing the process.
Washington, DC: National Academy Press.
NRC (1988) Health risks of radon and other internally deposited alpha-emitters: BEIRIV. Committee on the
Biological Effects of Ionizing Radiations. Washington, DC: National Academy Press.
NRC. (1993) Understanding child abuse and neglect. Panel on Research on Child Abuse and Neglect. National
Research Council, National academy Press, Washington, DC.
NTP (National Toxicology Program). (2002) Report on Carcinogens. Tenth Edition. Research Triangle Park, NC:
National Institute of Environmental Health Services, Public Health Service, U.S. Department of Health and Human
Services. Available from: .
NTP. (2005) Control data from NTP Technical Report Series Nos. 210, 267, 272, 306, 311, 314, 329, 346, 363, 371,
375, 377, 379, 380, 386, 421, 437, 440, 447, 448, 450, 451, 453, 454, 461, 462, 466, 467, 468, 471, 472, 475, 482,
486, 487, 490, 492, 499, 400, 507, 513, 515, 519 and control data from the wollastonite calcium silicates study. U.S.
Department of Health and Human Services, NTP, Research Triangle Park, NC. Available at http://ntp-
apps.niehs.nih.gov/ntp.tox/index.cfm.
Neuss, S; and Speit, G. (2008) Further characterization of the genotoxicity of formaldehyde in vitro by the sister
chromatid exchange test and co-cultivation experiments. Mutagenesis 23:355-7.
Nunn, A; Craigen, A; Darbyshire, J; et al. (1990) Six year follow up of lung function in men occupationally exposed
to formaldehyde. Br J Ind Med 47:747-752.
Obe, G; and Beek, B. (1979) Mutagenic activity of aldehydes. Drug Alcohol Depend 4:91-94.
O'Connor, PM; and Fox, BW. (1987) Comparative studies of DNA cross-linking reactions following methylene
dimethanesulphonate and its hydrolytic product, formaldehyde. Cancer Chemother Pharmacol 19:11-5.
Odeigah, PGC. (1997) sperm head abnormalities and dominant lethal effects of formaldehyde in albino rats. Mutat
Res 389:141-148.
O'Donovan, MR; Mee, CD. (1993) Formaldehyde is a bacterial mutagen in a range of Salmonella and Escherichia
indicator strains. Mutagenesis 8:577-581.
Oerstavik, D; andHongslo, JK. (1985) Mutagenicity of endodontic sealers. Biomaterials 6:129-132.
Ohba, Y; Morimitsu, Y; Watarai, A. (1979) Reaction of formaldehyde with calf-thymus nucleohistone. Eur J
Biochem 100:285-293.
This document is a draft for review purposes only and does not constitute Agency policy.
R-34 DRAFT—DO NOT CITE OR QUOTE
-------
Ohmichi, K; Komiyama, M; Matsuno, Y; et al. (2006) Relationship between exposure to formaldehyde and
immunoglobulin E (IgE) production during the gross anatomy laboratory. J Health Sci 52/5:642-647.
Ohta, T; Watanabe-Akanuma, M; Tokishita, S; et al. (1999) Mutation spectra of chemical mutagens determined by
Lac+ reversion assay with Escherichia coli WPS 101P-WP3106P tester strains. Mutat Res 440:59-74.
Ohta, T; Watanabe-Akanuma, M; Yamagata, H. (2000) A comparison of mutation spectra detected by the
Escherichia coli lac(+) reversion assay and the Salmonella typhimurium his(+) reversion assay. Mutagenesis
15:317-23.
Ohtani, T; Mizuashi, M; Nakagawa, S; et al. (2004a) Diesel exhaust particles and formaldehyde deviate human
immune system to Th2-dominant pattern by affecting both T cells and dendritic cells. J Invest Dermatol 122:A121.
Ohtani, T; Kurosawa, W; Mizuashi, M; et al. (2004b) Effects of diesel exhaust particles (DEP) and formaldehyde on
human monocyte-derived dendritic cells (MoDCs) and T cells. J Dermatol Sci 34:134.
Ohtsuka, R; Shuto, Y; Fujie, H; et al. (1997) Response of respiratory epithelium of BN and F344 rats to
formaldehyde inhalation. Exp Anim 46:279-286.
Ohtsuka, R; Shutoh, Y; Fujie, H; et al. (2003) Rat strain difference in histology and expression of Thl- and Th2-
related cytokines in nasal mucosa after short-term formaldehyde inhalation. Exp Toxicol Pathol 54:287-291.
Olin, KL; Cherr, GN; Rifkin, E, et al. (1996) The effects of some redox-active metals and reactive aldehydes on
DNA-protein cross-links in vitro. Toxicology 110:1-8.
Olsen, JH; Dossing, M. (1982) Formaldehyde induced symptoms in day care centers. Am Ind Hyg Assoc J
43:366-370.
Olsen, JH; Asnaes, S. (1986) Formaldehyde and the risk of squamous cell carcinoma of the sinonasal cavities. Br J
Ind Med 43:769-774.
Olsen, JH; Jensen, SP; Hink, M; et al. (1984) Occupational formaldehyde exposure and increased nasal cancer risk
in man. IntJ Cancer 34:639-644.
Olsen, GW; Lacy, SE; Bodner, KM; et al. (1997) Mortality from pancreatic and lymphopoietic cancer among
workers in ethylene and propylene chlorohydrin production. Occup Environ Med 54:592-598.
Orsiere, T; Sari-Minodier, I; larmarcovai, G; et al. (2006) Genotoxic risk assessment of pathology and anatomy
laboratory workers exposed to formaldehyde by use of personal air sampling and analysis of DNA damage of
peripheral lymphocytes. Mutat Res 605:30-41.
Orstavik, D; Hongslo, JK. (1985) Mutagenicity of endodontic sealers. Biomaterials 6:129-132.
Oshino, N; Chance, B; Sies, H; et al. (1973) The role of H2O2 generation in perfused rat liver and the reaction of
catalase compound I and hydrogen donors. Arch Biochem Biophys 154:117-131.
Ostojic, L; Bradaric, A; Mise, K; et al. (2006) Pulmonary function in persons who are professionally exposed to
formaldehyde fumes. Coll Antropol 30(3):507-511.
Overman, DO. (1985) Absence of embryotoxic effects of formaldehyde after percutaneous exposure in hamsters.
Toxicol Lett 24:107-110.
Overton, JH; Kimbell, JS; Miller, FJ. (2001) Dosimetry modeling of inhaled formaldehyde: the human respiratory
tract. Toxicol Sci 64:122-134.
Ovrebe, S; Haugen, A; Skaug, V. (2002) Biotransformation of formaldehyde in cultured human bronchus. Environ
Res 89:38-42.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 5 DRAFT—DO NOT CITE OR QUOTE
-------
Owen, BA; Dudney, CS; Tan, E; et al. (1990a) Comparative Hazard Evaluation, An Approach To
Regulation:Formaldehyde in Drinking Water. Health and Safety Research Division, Oak Ridge National Laboratory
Owen, BA; Dudney, CS; Tan, EL; et al. (1990b) Formaldehyde in drinking water: comparative hazard evaluation
and an approach to regulation. Regul Toxicol Pharmacol 11:220-236.
Ozen, OA; Yaman, M; Sarsilmaz, M; et al. (2002) Testicular zinc, copper and iron concentrations in male rats
exposed to subacute and subchronic formaldehyde gas inhalation. J Trace Elem Med Biol 16:119-122.
Ozen, OA; Songur, A; Sarsilmaz, M; et al. (2003a) Changes of zinc, copper, and iron levels in the lung of male rats
after subacute (4-week) and subchronic (13-week) exposure to formaldehyde. J Trace Elem Exp Med 16:67-74.
Ozen, OA; Songur, A; Sarsilmaz, M; et al. (2003b) Zinc, copper and iron concentrations in cerebral cortex of male
rats exposed to formaldehyde inhalation. J Trace Elem Med Biol 17:207-209.
Ozen, OA; Akpolat, N; Songur, A; et al. (2005) Effect of formaldehyde inhalation on Hsp70 in seminiferous tubules
of rat testes: An immunohistochemical study. Toxicol Ind Health 21:249-254.
Palczynski, C; Krakowiak, A; Hanke, W; et al. (1999) Indoor formaldehyde exposure and airway allergic diseases.
Int Rev Allergol Clin Immunol 5:65-69.
Park, H-S; Francis, KP; Yu, J; et al. (2003) Membranous cells in nasal-associated lymphoid tissue: a portal of entry
for the respiratory mucosal pathogen group A streptococcus. J Immunol 171:2532-2537.
Partanen, T; Kauppinen, T; Hernberg, S; et al. (1990) Formaldehyde exposure and respiratory cancer among
woodworkers: an update. Scand J Work Environ Health 16:394-400.
Passegue, E; Jamieson, CH; Allies, LE; et al. (2003) Normal and leukemic hematopoiesis: are leukemias a stem cell
disorder or a reacquisition of stem cell characteristics? Proc Natl Acad Sci U.S.A. 100(Suppl 1): 11842-11849.
Pauluhn, J. (1996) Predictive testing for respiratory sensitization Toxicol Lett 86:177-185.
Pauluhn, J. (2003) Issues of dosimetry in inhalation toxicity. Toxicol Lett 140-141:229-238.
Pazdrak, K; Gorski, P; Krakowiak, A; et al. (1993) Changes in nasal lavage fluid due to formaldehyde inhalation.
Int Arch Occup Environ Health 64:515-519.
Peddada, SD; Dinse, GE; Kissling, GE. (2007) Incorporating historical control data when comparing tumor
incidence rates. J Am Stat Assoc 102:1212-1220.
Pellegrino, R; Viegi, G; Brusasco, V; et al. (2005) Interpretative strategies for lung function tests. Eur Respir J
26:948-968.
Pinkerton, LE; Hein, MJ; Stayner, LT. (2004) Mortality among a cohort of garment workers exposed to
formaldehyde: an update. Occup Environ Med 61:193-200.
Pitten, FA; Kramer, A; Herrmann, K; et al. (2000) Formaldehyde neurotoxicity in animal experiments. Pathol Res
Pract 196:193-198.
Plesner, BH; and Hansen, K. (1983) Formaldehyde and hexamethylenetetramine in Styles' cell transformation assay.
Carcinogenesis 4:457-459.
Plopper, CG; Weir, AJ; Nishio, SJ; et al. (1994) Elevated susceptibility to 4-ipomeanol cytotoxicity in immature
Clara cells of neonatal rabbits. J Pharmacol Exp Ther 269:867-880.
Poortmans, JR; Kumps, A; Duez, P; et al. (2005) Effect of oral creatine supplementation on urinary methylamine,
formaldehyde, and formate. Med Sci Sports Exerc 37:1717-1720.
This document is a draft for review purposes only and does not constitute Agency policy.
R-36 DRAFT—DO NOT CITE OR QUOTE
-------
Portier, CJ; Hedges, JC; Hoel, DG. (1986) Age-specific models of mortality and tumor onset for historical control
animals in the National Toxicology Program's carcinogenicity experiments. Cancer Res 46:4372-4378.
Pottern, LM; Heineman, EF; Olsen, JH; et al. (1992) Multiple myeloma among Danish women: employment history
and workplace exposures. Cancer Causes Control 3:427-432.
Pourmahabadian, M; Azam, K; Ghasemkhani, M. (2006) Pulmonary function study between formaldehyde exposed
and non-exposed staffs at some of the Tehran Educational Hospitals. J Med Sci 6/4:621-625.
Priha, E; Liesivuori, J; Santa, H; et al. (1996) Reactions of hydrated formaldehyde in nasal mucus. Chemosphere
32:1077-1082.
Proctor, BL; Gaulden, ME; Dowd, MA. (1986) Reactivity and fate of benzene and formaldehyde in culture medium
with and without fetal calf serum; relevance to in vitro mutagenicity testing. Mutat Res 160:259-266.
Pross, HF; Day, JH; Clark, RH; et al. (1987) Immunologic studies of subjects with asthma exposed to formaldehyde
and urea-formaldehyde foam insulation (UFFI) off products. J Allergy Clin Immunol 79:797-810.
Pushkina, N; Gofmekler, V; Klevtsova, G. (1968) Changes in ascorbic-acid and nucleic-acid content during benzol
and formaldehyde action rat [translation]. Bull Exp Biol Med 66:51-53.
Pyatt, D; Natelson, E; Golden, R. (2008) Is inhalation exposure to formaldehyde a biologically plausible cause of
lymphohematopoietic malignancies? Regul Toxicol Pharmacol 51:119-133.
Que, LG; Liu, L; Yan, Y; et al. (2005) Protection from experimental asthma by an endogenous bronchodilator.
Science 308:1618-1621.
Quievryn, G; Zhitkovich, A. (2000) Loss of DNA-protein crosslinks from formaldehyde-exposed cells occurs
through spontaneous hydrolysis and an active repair process linked to proteosome function. Carcinogenesis
21:1573-1580.
Ragan, DL; and Boreiko, CJ. (1981) Initiation of C3H/10T1/2 cell transformation by formaldehyde. Cancer Lett
13:325-331.
Rao, GN; Piegorsch, WW; Haseman, JK. (1987) Influence of body weight on the incidence of spontaneous
neoplasms in rats and mice of long-term studies. Am J Clin Nutr. 45:252-260.
Rapoport, I A. (1946) Carbonyl compounds and the chemical mechanisms of mutation. Compt Rend Acad Sci
(U.S.S.R.) 54:65-67.
Recio, L; Sisk, S; Pluta, L, et al. (1992) p53 mutations in formaldehyde-induced nasal squamous cell carcinomas in
rats. Cancer Res 52:6113-6116.
Report of the Federal Panel on Formaldehyde. (1982) Environ Health Perspect 43:139-168.
Reuzel, PGJ; Wilmer, JWGM; Woutersen, RA; et al. (1990) Interactive effects of ozone and formaldehyde on the
nasal respiratory lining epithelium in rats. J Toxicol Environ Health 29:279-292.
Riedel, F; Hasenauer, E; Earth, PJ; et al. (1996) Formaldehyde exposure enhances inhalative allergic sensitization in
the guinea pig. Allergy 51:94-99.
Ritchie, IM; Lehnen, RG. (1987) Formaldehyde-related health complaints of residents living in mobile and
conventional homes. Am J Public Health 77:323-328.
This document is a draft for review purposes only and does not constitute Agency policy.
R-37 DRAFT—DO NOT CITE OR QUOTE
-------
Robins, J; Ambrun, M; Hute, C; et al. (1988) Estimating the effect of formaldehyde exposure on long cancer and
non-malignant respiratory disease (NMRD) mortality using a new method to control for the health worker survivor
effect. In: Hogstedt, C; Reuterwall, C; eds. Progress in occupational epidemiology. Amsterdam: Elsevier Science;
pp. 75-78.
Roemer, E; Anton, HJ; Kindt, R. (1993) Cell proliferation in the respiratory tract of the rat after acute inhalation of
formaldehyde or acrolein. J Appl Toxicol 13:103-107.
Rogers, J; Mole, M; Chernoff, N; et al. (2002) Developmental toxicity of inhaled methanol in the CD-I mouse, with
application of quantitative dose-response modeling for estimation of benchmark dosages. U.S. Environmental
Protection Agency, Washington, D.C., EPA/600/J-92/350 (NTIS PB93106979).
Rogers, JM; Brannen, KC; Barbee, BD; et al. (2004) Methanol exposure during gastrulation causes
holoprosencephaly, facial dysgenesis, and cervical vertebral malformations in C57BL/6J mice Birth Defects B Dev
Reprod Toxicol 71:80-88.
Ross, WE; Shipley, N. (1980) Relationship between DNA damage and survival in formaldehyde-treated mouse cells.
Mutat Res 79:277-283.
Ross, WE; McMillan, DR; and Ross, CF. (1981) Comparison of DNA damage by methylmelamines and
formaldehyde. J Natl Cancer Inst 67:217-221.
Rotstein, J; Sarma, DSR; Farber, E. (1986) Sequential alterations in growth control and cell dynamics of rat
hepatocytes in early precancerous steps in hepatocarcinogenesis. Cancer Res 46:2377-2385.
Roush, GC; Walrath, J; Stayner, LT; et al. (1987) Nasopharyngeal cancer, sinonasal cancer, and occupations related
to formaldehyde: a case-control study. J Natl Cancer Inst 79:1221-1224.
Rumchev, KB; Spickett, JT; Bulsara, MK; et al. (2002) Domestic exposure to formaldehyde significantly increases
the risk of asthma in young children. Eur Respir J 20:403-408.
Rumchev, K; Spickett, J; Bulsara, M; et al. (2004) Association of domestic exposure to volatile organic compounds
with asthma in young children. Thorax 59: 746-751.
Rusch, GM; Bolte, HF; Rinehart, WE. (1983a) A 26-week inhalation toxicity study with formaldehyde in the
monkey, rat, and hamster. In: Gibson, JE; ed. Formaldehyde toxicity. Washington, DC: Hemisphere Publishing
Corporation; pp. 98-110.
Rusch, GM; Clary, JJ; Rinehart, WE; et al. (1983b) A 26-week inhalation toxicity study with formaldehyde in the
monkey, rat and hamster. Toxicol Appl Pharmacol 68:329-343.
Ryden, E; Ekstrom, C; Hellmer, L; et al. (2000) Comparison of the sensitivities of Salmonella typhimurium strains
TA102andTA2638Ato 16 mutagens. Mutagenesis 15:495-502.
Sadakane, K; Takano, H; Ichinose, T; et al. (2002) Formaldehyde enhances mite allergen-induced eosinophilic
inflammation in the murine airway. J Environ Pathol Toxicol Oncol 21:267-276.
Saillenfait, AM; Bonnet, P; de Ceaurriz, J. (1989) The effects of maternally inhaled formaldehyde on embryonal and
foetal development in rats. Food Chem Toxicol 27:545-548.
Saito, Y; Nishio, K; Yoshida, Y, et al. (2005) Cytotoxic effect of formaldehyde with free radicals via increment of
cellular reactive oxygen species. Toxicology 210:235-245.
Saladino, AJ; Willey, JC; Lechner, JF, et al. (1985) Effects of formaldehyde, acetaldehyde, benzoyl peroxide, and
hydrogen peroxide on cultured normal human bronchial epithelial cells. Cancer Res 45:2522-2526.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 8 DRAFT—DO NOT CITE OR QUOTE
-------
Salkie, ML. (1991) The prevalence of atopy and hypersensitivity to formaldehyde in pathologists. ArchPathol Lab
Medll5(6):614-616
Sallmen, M; Lindbohm, ML; Kyyronen, P; et al. (1995) Reduced fertility among women exposed to organic
solvents. Am J Ind Med 27(5):699-713.
Salthammer T, S Mentese and R Marutzky (2010) Formaldehyde in the Indoor Environment. Chem. Rev., DOI:
10.1021/cr800399g
Sandikci, M; Eren, U; Kum, S. (2007) Effects of formaldehyde and xylene on CD4- and CDS-positive T cells in
bronchus-associated lymphoid tissue in rats. Toxicol Ind Health 23:471-477.
Sanghani, PC; Stone, CL; Ray, BD; et al. (2000) Kinetic mechanism of human glutathione-dependent formaldehyde
dehydrogenase. Biochemistry 39:10720-10729.
Santiago, LY; Harm, MC; Ben-Jebria, A; et al. (2001) Ozone absorption in the human nose during unidirectional
airflow. JApplPhysiol 91:725-732.
Sari-Minodier, I; Orsiere, T; Auquier, P; et al. (2001) Use of the micronucleus assay in the mutagenic risk
assessment: A case of 10 workers occupationally exposed to formaldehyde. Arch Mai Prof 62:75-82.
Sarsilmaz, M; Ozen, O; Akpolat, N; et al. (1999) The histopathologic effects of inhaled formaldehyde on Ley dig
cells of the rats in subacute period [translation]. Firat Univ Med Sci J 13:37-40.
Sarsilmaz, M; Kaplan, S; Songur, A; et al. (2007) Effects of postnatal formaldehyde exposure on pyramidal cell
number, volume of cell layer in hippocampus and hemisphere in therat: A stereological study. Brain Res 1145:157-
167.
Sarrif, AM; Krahn, DF; Donovan, SM, et al. (1997) Evaluation of hexamethylphosphorarnide for gene mutations in
Salmonella typhimurium using plate incorporation, preincubation, and suspension assays. Mutat Res 380:167-177.
Sauder, LR; Chatham, MD; Green, DJ; et al. (1986) Acute pulmonary response to formaldehyde exposure in healthy
nonsmokers. JOccup Med 28:420^24.
Sauder, LR; Green, DJ; Chatham, MD; et al. (1987) Acute pulmonary response of asthmatics to 3.0 ppm
formaldehyde. Toxicol Ind Health 3:569-578.
Schachter, EN; Witek, TJ, Jr; Tosun, T; et al. (1986) A study of respiratory effects from exposure to 2 ppm
formaldehyde in healthy subjects. Arch Environ Health 41:229-23 9.
Schachter, EN; Witek, TJ, Jr; Brody, DJ; et al. (1987) A study of respiratory effects from exposure to 2.0 ppm
formaldehyde in occupationally exposed workers. Environ Res 44:188-205.
Schauenstein, E; Esterbauer, H; Zollner, H. (1977) Aldehydes in biological systems. London: Pion Limited.
Schlosser, PM. (1999) Relative roles of convection and chemical reaction for the disposition of formaldehyde and
ozone in nasal mucus. Inhal Toxicol 11:967-980.
Schlosser, PM; Lilly, PD; Conolly, RB; et al. (2003) Benchmark dose risk assessment for formaldehyde using
airflow modeling and a single-compartment, DNA-protein cross-link dosimetry model to estimate human equivalent
doses. Risk Anal 23:473-487.
Schmid, E; Goggelmann, W; Bauchinger, M. (1986) Formaldehyde-induced cytotoxic, genotoxic and mutagenic
response in human lymphocytes and Salmonella typhimurium. Mutagenesis 1:427-431.
Schmid, O; Speit, G. (2007) Genotoxic effects induced by formaldehyde in human blood and implications for the
interpretation of biomonitoring studies. Mutagenesis 22:69-74.
This document is a draft for review purposes only and does not constitute Agency policy.
R-39 DRAFT—DO NOT CITE OR QUOTE
-------
Schoenberg, JB; Beck, GJ; Bouhuys, A. (1978) Growth and decay of pulmonary function in healthy blacks and
whites. RespirPhysiol 33:367-393.
Schreiber, H; Bibbo, M; Wied, GL; et al. (1979) Bronchial metaplasia as a benign or premalignant lesion. I.
Cytologic and ultrastructural discrimination between acute carcinogen effects and toxin-induced changes. Acta
Cytol 23:496-503.
Schuck, EA; Stephens, ER; Middleton, JT. (1966) Eye irritation response at low concentrations of irritants. Arch
Environ Health 13(5):570-575.
Schulte-Hermann, R; Bursch, W; Grasl-Kraupp, B; et al. (1997) Concepts of cell death and application to
carcinogenesis. Toxicol Pathol 25:89-93.
Schulte-Hermann, R; Bursch, W; Marian, B; et al. (1999) Active cell death (apoptosis) and cellular proliferation as
indicators of exposure to carcinogens. IARC Sci.Publ 273-285.
Seidenberg, J; Anderson, DG; Becker, R. (1986) Validation of an in vivo developmental toxicity screen in the
mouse. Teratog Carcinog Mutagen 6:361-374.
Seidenberg, J; Becker, R. (1987) A summary of the results of 55 chemicals screened for developmental toxicity in
mice. Teratog Carcinog Mutagen 7:17-28.
Seitz, T; Baron, S. (1990) Health hazard evaluation report: Rockcastle Manufacturing, Mount Vernon, Kentucky.
National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services, Cincinnati,
OH; HETA 87-349-2022.
Selevan, SG; Kimmel, CA; Mendola, P. (2000) Identifying critical windows of exposure for children's health.
Environ Health Perspect 108 Suppl 3:451-455.
Sellakumar, AR; Snyder, CA; Solomon, JJ; et al. (1985) Carcinogenicity of formaldehyde and hydrogen chloride in
rats. Toxicol Appl Pharmacol 81:401-406.
Senichenkova, IN. (1991) The embryotoxic effect of industrial environmental pollutants formaldehyde and gasoline
[translation]. Gig Sank 9:3 5-3 8.
Senichenkova, IN; Chebotar, NA. (1996) Effects of gasoline and formaldehyde on prenatal development of rats with
induced iron microelementosis [translation]. Ontogenez 27:108-113.
Shaham, J; Bomstein, Y; Meltzer, A; et al. (1996) DNA-protein crosslinks, a biomarker of exposure to
formaldehyde—in vitro and in vivo studies. Carcinogenesis 17:121-125.
Shaham, J; Bomstein, Y; Melzer, A; et al. (1997) DNA-protein crosslinks and sister chromatid exchanges as
biomarkers of exposure to formaldehyde. Int J Occup Environ Health 3:95-104.
Shaham, J; Gurvich, R; Kaufman, Z. (2002) Sister chromatid exchange in pathology staff occupationally exposed to
formaldehyde. MutatRes514:115-123.
Shaham, J; Bomstein, Y; Gurvich, R; et al. (2003) DNA-protein crosslinks and p53 protein expression in relation to
occupational exposure to formaldehyde. Occup Environ Med 60:403-409.
Shangina, O; Brennan, P; Szeszenia-Dabrowska, N; et al. (2006) Occupational exposure and laryngeal and
hypopharyngeal cancer risk in central and eastern Europe. Am J Epidemiol 164(4): 367-375.
Sharpe, JA; Hostovsky, M; Bilbao, JM; et al. (1982) Methanol optic neuropathy. A histopathological study.
Neurology 32:1093-1100.
This document is a draft for review purposes only and does not constitute Agency policy.
R-40 DRAFT—DO NOT CITE OR QUOTE
-------
Shelby, M; Portier, C; Goldman, L; et al. (2004) NTP-CERHR expert panel report on the reproductive and
developmental toxicity of methanol. Reprod Toxicol 18:303-390.
Sheppard, D; Eschenbacher, W; Epstein, J. (1984) Lack of bronchomotor response to up to 3 ppm formaldehyde in
subjects with asthma. Environ Res 35:133-139.
Sheveleva (1971) In Toxicology of New Industrial Chemicals. 12:78-86.
Shibamoto, T. (2006) Analytical methods for trace levels of reactive carbonyl compounds formed in lipid
peroxidation systems. J Pharm Biomed Anal 41:12-25.
Shin, HS; Ahn, HS; Lee, BH. (2007) Determination of thiazolidine-4 carboxylates in urine by chloroformate
derivatization and gas chromatography-electron impact mass spectrometry. J Mass Spectrom 42:1225-1232.
Shipkovitz, HD. (1966) Formaldehyde vapor in the permanent-press fabrics industry. Occupational Health Program,
Public Health Service, U.S. Department of Health, Education, and Welfare, Cincinnati, OH; Report-TR-52.
Singh, GD; McNamara, JA Jr; Lozanoff, S. (1998) Morphometry of the midfacial complex in subjects with class III
malocclusions: Procrustes, Euclidean, and cephalometric analyses. Clin Anat 11:162-170.
Skrzydlewska, E. (1996) Decreased antioxidant status and increased lipid peroxidation in rats after methanol
intoxication. Rocz Akad Med Bialymst 41:397-404.
Slater, TF. (1984) Free-radical mechanisms in tissue injury. Biochem J 222:1-15.
Slikker, WJ; Andersen, M; Bogdanffy, M; et al. (2004) Dose-dependent transitions in mechanisms of toxicity: Case
studies. Toxicol Appl Pharmacol 201:226-294.
Smedje, G; Norback, D. (2001) Incidence of asthma diagnosis and self-reported allergy in relation to the school
environment—a four-year follow-up study in schoolchildren. Int J Tuberc Lung Dis 5:1059-1066.
Smith, AE; Coit, DW. (1995) Penalty Functions. In Environmental Computation 2. Eds: Back, T; Fogel, DB;
Michalewicz., Z. 41-48. Institute of Physics Publishing and Oxford University Press, New York, NY.
Smith, D; Spanel, P. (2005) Selected ion flow tube mass spectrometry (SIFT-MS) for on-line trace gas analysis.
Mass Spectrom Rev 24:661-700.
Smith, D; Spanel, P. (2007) The challenge of breath analysis for clinical diagnosis and therapeutic monitoring.
Analyst 132:390-396.
Snedecor, GW; Cochran, WG. (1980) Statistical Methods, Seventh Edition. Iowa State University Press, Ames, IA.
Snyder, RD; van Houten, B. (1986) Genotoxicity of formaldehyde and an evaluation of its effects on the DNA repair
process in human diploid fibroblasts. Mutat Res 165:21-30.
Songur, A; Kus, I; Sahin, S; Sogut, S; Ozen, OA; Yaman, M; Sarsilmaz, M. (2005) The changes of zinc, copper, and
iron levels in lung tissue after formaldehyde inhalation during the early postnatal period of rats. Eur J Gen Med
2(2):62-68.
Songur, A; Sarsilmaz, M; Ozen, OA. (2008) The effects of inhaled formaldehyde on oxidant and antioxidant systems
of rat cerebellum during the postnatal development process. Toxicol Mechan Methods 18:569-574.
Sorg, BA; Hochstatter, T. (1999) Behavioral sensitization after repeated formaldehyde exposure in rats. Toxicol Ind
Health 15:346-355.
This document is a draft for review purposes only and does not constitute Agency policy.
R-41 DRAFT—DO NOT CITE OR QUOTE
-------
Sorg, BA; Willis, JR; See, RE; et al. (1996) Repeated low-level formaldehyde exposure produces long-term cross-
sensitization to cocaine: relevance to an animal model for multiple chemical sensitivity. Soc Neurosci Abstr
22:2064
Sorg, BA; Willis, JR; See, RE; et al. (1998) Repeated low-level formaldehyde exposure produces cross-sensitization
to cocaine: Possible relevance to chemical sensitivity in humans. Neuropsychopharm 18:385-394.
Sorg, BA; Bailie, TM; Tschirgi, ML; et al. (2001) Exposure to repeated low-level formaldehyde in rats increases
basal corticosterone levels and enhances the corticosterone response to subsequent formaldehyde. Brain Res
898:314-320.
Sorg, BA; Swindell, S; Tschirgi, ML. (2004) Repeated low level formaldehyde exposure produces enhanced fear
conditioning to odor in male, but not female, rats. Brain Res 1008:11-19.
Spanel, P and Smith, D. (2008) Quantification of trace levels of the potential cancer biomarkets formaldehyde,
acetaldehyde andpropanol in breath by SIFT-MS. J. Breath Res. 2: 1-10.
Speit, G; Merk, O. (2002) Evaluation of mutagenic effects of formaldehyde in vitro: detection of crosslinks and
mutations in mouse lymphoma cells. Mutagenesis 17:183-187.
Speit, G; Schmid, O. (2006) Local genotoxic effects of formaldehyde in humans measured by the micronucleus test
with exfoliated epithelial cells. Mutat Res 613:1-9.
Speit, G; Schutz, P; Merk, O. (2000) Induction and repair of formaldehyde-induced DNA-protein crosslinks in
repair-deficient human cell lines. Mutagenesis 15:85-90.
Speit, G; Schmid, O; Frohler-Keller, M; et al. (2007a) Assessment of local genotoxic effects of formaldehyde in
humans measured by the micronucleus test with exfoliated buccal mucosa cells. Mutat Res 627:129-135.
Speit, G; Schutz, P; Hogel, J, et al. (2007b) Characterization of the genotoxic potential of formaldehyde in V79 cells.
Mutagenesis 22:387-394.
Speit, G; Schmid, O; Neuss, S, et al. (2008) Genotoxic effects of formaldehyde in the human lung cell line A549 and
in primary human nasal epithelial cells. Environ Mol Mutagen 49:300-307.
Speit, G; Zeller, J; Schmid, O, et al. (2009) Inhalation of formaldehyde does not induce systemic genotoxic effects in
rats. Mutat Res. (In Press)
Squire, RA; Cameron, LL. (1984) An analysis of potential carcinogenic risk from formaldehyde. Regul Toxicol
Pharmacol 4:107-129.
Staab, CA; Hellgren, M; Hoog, J-O. (2008a) Dual functions of alcohol dehydrogenase 3: Implications with focus on
formaldehyde dehydrogenase and S-nitrosoglutathione reductase activities. Cell Mol Life Sci 65:3950-3960.
Staab, CA; Alander, J; Brandt, M; et al. (2008b) Reduction of S-nitrosoglutathione by alcohol dehydrogenase 3 is
facilated by substrate alcohols via direct cofactor recycling and leads to GSH-controlled formation of glutathione
transferase inhibitors. Biochem J 41:493-504.
Staab, CA; Hellgren, M; Grafstrom, RC; et al. (2009) Medium-chain fatty acids and glutathione derivatives as
inhibitors of S-nitroglutathione reduction mediated by alcohol dehydrogenase 3. Chem Biol Interact 180:113-118.
Staffclbach T. Ncftcl A, Stauffcr B, cl al. 1991. A record of the from in
ice cores. Nature 349:603-605.
Stayner, LT; Elliot, L; Blade, L; et al. (1988) A retrospective cohort mortality study of workers exposed to
formaldehyde in the garment industry. AmJInMed 13:667-681.
This document is a draft for review purposes only and does not constitute Agency policy.
R-42 DRAFT—DO NOT CITE OR QUOTE
-------
Stedman's Medical Dictionary. (2005) 28th edition. Lippincott Williams, and Wilkins, New York, NY.
Steinhagen, WH; Barrow, CS. (1984) Sensory irritation structure-activity study of inhaled aldehydes inB6C3Fl and
Swiss-Webster mice. Toxicol Appl Pharmacol 72:495-503.
Stellman, SD; Demers, PA; Colin, D; et al. (1998) Cancer mortality and wood dust exposure among participants in
the American cancer Society cancer prevention study-II (CPS-II). Am J Ind Med 34:229-237.
Sterling, TD; Weinkam, JJ. (1988) Reanalysis of lung cancer mortality in a National Cancer Institute study on
mortality among industrial workers exposed to formaldehyde. J Occup Med 30:895-901.
Sterling, TD; Weinkam, JJ. (1989a) Reanalysis of lung cancer mortality in a National Cancer Institute study on
"Mortality among industrial workers exposed to formaldehyde." Exp Pathol 37:128-132.
Sterling, TD; Weinkam, JJ. (1989b) Reanalysis of lung cancer mortality in a National Cancer Institute study of
"Mortality among industrial workers exposed to formaldehyde:" additional discussion. J Occup Med 31:881-884.
Sterling, TD; Weinkam, JJ. (1994) Mortality from respiratory cancers (including lung cancer) among workers
employed in formaldehyde industries. Am J Ind Med 25:593-602.
Stewart, PA; Blair, A; Cubit, DA; et al. (1986) Estimating historical exposures to formaldehyde in a retrospective
mortality study. Appl Ind Hyg 1:34-41.
Stewart, PA; Cubit, D; Blair, A. (1987) Formaldehyde levels in seven industries. Appl Ind Hyg 2:231-236.
Stewart, PA; Herrick, RF; Feigley, CE; et al. (1992) Study design for assessing exposures of embalmers for a case-
control study. Part I. Monitoring results. Appl Occup Environ Hyg 7:532-540.
Stone, DM; Grillo, M; Margolis, FL; et al. (1991) Differential effect of functional olfactory bulb deafferentation on
tyrosine hydroxylase and glutamic acid decarboxylase messenger RNA levels in rodent juxtaglomerular neurons. J
Comp Neurol 311:223-233.
Stroup, NE; Blair, A; Erikson, GE. (1986) Brain cancer and other causes of death in anatomists. J Natl Cancer Inst
77:1217-1224.
Subramaniam, RP; Richardson, RB; Morgan, KT; et al. (1998) Computational fluid dynamics simulations of
inspiratory airflow in the human nose and nasopharynx. Inhal Toxicol 10:92-120.
Subramaniam, RP; Crump, KS; Van Landingham, C; et al. (2007) Uncertainties in the CUT model for formaldehyde-
induced carcinogenicity in the rat: A limited sensitivity analysis-I. Risk Anal 27:1237-1254.
Subramaniam, RP, Chen, C; Crump, KS; et al. (2008) Uncertainties in biologically-based modeling of
formaldehyde-induced respiratory cancer risk: Identification of key issues. Risk Anal 28:907-923.
Sul, D; Kim, H; Oh, E; et al. (2007) Gene expression profiling in lung tissues from rats exposed to formaldehyde.
Arch Toxicol 81:589-597.
Suruda, A; Schulte, P; Boeniger, M; et al. (1993) Cytogenetic effects of formaldehyde exposure in students of
mortuary science. Cancer Epidemiol Biomarkers Prev 2:453^160.
Suskov, II; and Sazonova, LA. (1982) Cytogenetic effects of epoxy, phenolformaldehyde and polyvinylchloride
resins in man. Mutat Res 104:137-140.
Sweeney, MH; Beaumont, JJ; Waxweiler, RJ; et al. (1986) An investigation of mortality from cancer and other
causes of death among workers employed at an East Texas chemical plant. Arch Environ Health 41:23-28.
This document is a draft for review purposes only and does not constitute Agency policy.
R-43 DRAFT—DO NOT CITE OR QUOTE
-------
Swenberg, JA; Kerns, WD; Mitchell, RI; et al. (1980) Induction of squamous cell carcinomas of the rat nasal cavity
by inhalation exposure to formaldehyde vapor. Cancer Res 40:3398-3402.
Swenberg, JA; Gross, EA, Martin, J; et al. (1983) Mechanisms of formaldehyde toxicity. In: Formaldehyde toxicity,
Ed: Gibson, JE. Hemisphere Publishing Corp. Washington, D.C.
Swenberg, JA; Gross, EA; Randall, HW. (1986) Localization and quantitation of cell proliferation following
exposure to nasal irritants. In Toxicology of Nasal Passages, ed: Barrow, CS. Hemisphere Publishing Corp.
Washington, D.C.
Swiecichowski, AL; Long, KJ; Miller, ML; et al. (1993) Formaldehyde-induced airway hyperreactivity in vivo and
ex vivo in guinea pigs. Environ Res 61:185-199.
Szarvas, T; Szatloczky, E; Volford, J; et al. (1986) Determination of endogenous formaldehyde level in human blood
and urine by dimedone-14C radiometric method. J Radioanaly Nucl Chem Letters 106:357-367.
Takahashi, K; Morita, T; Kawazoe, Y. (1985) Mutagenic characteristics of formaldehyde on bacterial systems.
Mutat Res 156:153-161.
Takahashi, M; Hasegawa, R; Furukawa, F; et al. (1986) Effects of ethanol, potassium metabisulfite, formaldehyde
and hydrogen peroxide on gastric carcinogenesis in rats after initiation with N-methyl-N'-nitro-N-nitrosoguanidine.
JpnJ Cancer Res 77:118-124.
Takahashi, S; Tsuji, K; Fujii, K; et al. (2007) Prospective study of clinical symptoms and skin test reactions in
medical students exposed to formaldehyde gas. J Dermatol 34(5):283-289.
Takigawa, T; Usami, M; Yamasaki, Y; et al. (2005) Reduction of indoor formaldehyde concentrations and subjective
symptoms in a gross anatomy laboratory. Bull Environ Contain Toxicol 74:1027-1033.
Tang, X; Bai, Y; Duong, A, et al. (2009) Formaldehyde in China: Production, consumption, exposure levels, and
health effects. Envir. International. 35:1210-1224.
Tarkowski, M; Gorski, P. (1995) Increased IgE antiovalbumin level in mice exposed to formaldehyde. Int Arch
Allergy Immunol 106:422^24.
Tarone, RE. (1975) Tests for trend in life table analysis. Biometrika 62:679-682. (as cited in Battelle Columbus
Laboratories [1981])
Tarone, R; McLaughlin, JK. (2005) Re: "mortality from solid cancers among workers in formaldehyde industries."
AmJEpidemiol 161:1089-1091.
Taskinen, H; Kyyronen, P; Hemminki, K. (1994) Laboratory work and pregnancy outcome. J Occup Med
36:311-319.
Taskinen, HK; Kyyronen, P; Sallmen, M; et al. (1999) Reduced fertility among female wood workers exposed to
formaldehyde. Am JInd Med 36:206-212.
Tavernier, G; Fletcher, G; Gee, I; et al. (2006) IPEADAM study: Indoor endotoxin exposure, family status, and
some housing characteristics in English children. J Allergy Clin Immunol 117:656-662.
Temcharoen, P; Thilly, WG. (1983) Toxic and mutagenic effects of formaldehyde in Salmonella typhimurium.
Mutat Res 119:89-93.
Teng, S; Beard, K; Pourahmad, J; et al. (2001) The formaldehyde metabolic detoxification enzyme systems and
molecular cytotoxic mechanism in isolated rat hepatocytes. Chem Biol Interact 130-132:285-296.
This document is a draft for review purposes only and does not constitute Agency policy.
R-44 DRAFT—DO NOT CITE OR QUOTE
-------
Thomas, RS; Allen, BC; Nong, A; et al. (2007) A method to integrate benchmark dose estimates with genomic data
to assess the functional effects of chemical exposure. Toxicol Sci 98(l):240-248.
Thompson, CM; Grafstrom, RC. (2008) Mechanistic considerations for formaldehyde-induced bronchoconstriction
involving S-nitrosoglutathione reductase. J Toxicol Environ Health A 71:244-248.
Thompson, CM; Subramaniam, RP; Grafstrom, RC. (2008) Mechanistic and dose considerations for supporting
adverse pulmonary physiology in response to formaldehyde. Toxicol Appl Pharmacol 233:355-359.
Thompson, CM; Sonawane, B; Grafstrom, RC. (2009) The ontogeny, distribution and regulation of alcohol
dehydrogenase 3: implications for pulmonary physiology. DrugMetab Dispos. [Epub ahead of print]
Thomson, EJ; Shackleton, S; Harrington, JM. (1984) Chromosome aberrations and sister-chromatid exchange
frequencies in pathology staff occupationally exposed to formaldehyde. Mutat Res 141:89-93.
Thorndike, J; Beck, WS. (1977) Production of formaldehyde from N5-methyltetrahydrofolate by normal and
leukemic leukocytes. Cancer Res 37:1125-1132.
Thrasher, JD; Wojdani, A; Cheung, G; et al. (1987) Evidence for formaldehyde antibodies and altered cellular
immunity in subjects exposed to formaldehyde in mobile homes. Arch Environ Health 42:347-350.
Thrasher, JD; Broughton, A; Card, Z. (1988) Indoor formaldehyde and the elderly. Clin Gerontol 7(3-4):63-66.
Thrasher, JD; Broughton, A; Madison, R. (1990) Immune activation and autoantibodies in humans with long-term
inhalation exposure to formaldehyde. Arch Environ Health 45:217-223.
Thun, MJ; Lakat, M; Airman, R. (1982) Symptom survey of residents of homes insulated with urea formaldehyde
foam. Environ Res 29(2):320-334.
Til, HP; Woutersen, RA; Feron, VJ; et al. (1988) Evaluation of the oral toxicity of acetaldehyde and formaldehyde in
a 4-week drinking-water study in rats. Food Chem Toxicol 26:447-452.
Til, HP; Woutersen, RA; Feron, VJ; et al. (1989) Two-year drinking-water study of formaldehyde in rats. Food
Chem Toxicol 27:77-87.
Titenko-Holland, N; Levine, AJ; Smith, MT; et al. (1996) Quantification of epithelial cell micronuclei by
fluorescence in situ hybridization (FISH) in mortuary science students exposed to formaldehyde. Mutat Res
371:237-248.
Tobe, M; Kaneko, Y; Uchida, Y; et al. (1985) Studies of the inhalation toxicity of formaldehyde. National Sanitary
and Medical Laboratory Service, Toxicity, Tokyo, Japan; T-85-0236. (cited in IPCS, 2002)
Tobe, M; Naito, K; Kurokawa, Y. (1989) Chronic toxicity study on formaldehyde administered orally to rats.
Toxicology 56:79-86.
Togias, A. (1999) Mechanisms of nose-lung interaction. Allergy 54:94-105.
Trezl, L; Hullan, L; Jaszay, ZM; et al. (2003) Antagonistic reactions of arginine and lysine against formaldehyde and
their relation to cell proliferation, apoptosis, folate cycle and photosynthesis. Mol Cell Biochem 244:167-176.
Tsubone, H; Kawata, M. (1991) Stimulation to the trigeminal afferent nerve of the nose by formaldehyde, acrolein,
and acetaldehyde gases. Inhal Toxicol 3:211-222.
Tsukahara, S; Yamamoto, S; Tin-Tin-Tin-Shwe; et al. (2006) Inhalation of low-level formaldehyde increases the
Bcl-2/Bax expression ratio in the hippocampus of immunologically sensitized mice. Neuroimmunomodulation
13:63-68.
This document is a draft for review purposes only and does not constitute Agency policy.
R-45 DRAFT—DO NOT CITE OR QUOTE
-------
Turner, C; Parekh, B; Walton, C; et al. (2008) An exploratory comparative study of volatile compounds in exhaled
breath and emitted by skin using selected ion flow tube mass spectrometry. Rapid Commun Mass Spectrom.
22(4):526-532.
Tuthill, RW. (1984) Woodstoves, formaldehyde, and respiratory disease. Am J Epidemiol 120:952-955.
Tyihak, E; Albert, L; Nemeth, ZI; et al. (1998) Formaldehyde cycle and the natural formaldehyde generators and
capturers. Acta Biol Hung 49:225-238.
Tyihak, E; Bocsi, J; Timar, F; et al. (2001) Formaldehyde promotes and inhibits the proliferation of cultured tumour
and endothelial cells. Cell Prolif 34:135-141.
U.S. EPA (U.S. Environmental Protection Agency). (1985) Health and environmental effects profile for
formaldehyde. Prepared by the U.S. Environmental Protection Agency, Office of Research and Development,
Cincinnati, OH for the Office of Solid Waste and Emergency Response, Washington, D.C.; EPA 600/X-85/362.
U.S. EPA (Environmental Protection Agency). (1986a) Guidelines for the health risk assessment of chemical
mixtures. Federal Register 51(185):34014-34025. Available from: .
U.S. EPA. (1986b) Guidelines for mutagenicity risk assessment. Federal Register 51(185):34006-34012. Available
from: .
U.S. EPA. (1987) Assessment of health risks to garment workers and certain home residents from exposure to
formaldehyde. Office of Pesticides and Toxic Substances, Washington, DC.
U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk assessment.
Prepared by the Environmental Criteria and Assessment Office, Office of Health and Environmental Assessment,
Cincinnati, OH for the Office of Solid Waste and Emergency Response, Washington, DC; EPA 600/6-87/008.
Available from: .
U.S. EPA. (1989) Review of the national ambient air quality standard for ozone, assessment for scientific and
technical information. U.S. Environmental Protection Agency, RTF, NC, EPA-450/2-92.
U.S. EPA. (1991a) Guidelines for developmental toxicity risk assessment. Federal Register 56(234):63798-63826.
Available from: .
U.S. EPA. (1991b) Formaldehyde Risk Assessment Update. Prepared by the Office of Toxic Substances,
Washington, DC.
U.S. EPA. (1991c) Formaldehyde. Integrated Risk Information System (IRIS). National Center for Environmental
Assessment, Washington, DC. Available online at http://www.epa.gov/iris.
U.S. EPA. (1994a) Interim policy for particle size and limit concentration issues in inhalation toxicity studies.
Federal Register 59(206):53799. Available from: .
U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application of inhalation
dosimetry. Office of Research and Development, Washington, DC; EPA/600/8-90/066F. Available from:
.
U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/R-94/007. Available from:
.
U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Federal Register 61(212):56274-56322.
Available from: .
This document is a draft for review purposes only and does not constitute Agency policy.
R-46 DRAFT—DO NOT CITE OR QUOTE
-------
U.S. EPA. (1998) Guidelines for neurotoxicity risk assessment. Federal Register 63(93):26926-26954. Available
from: .
U.S. EPA. (2000a) Science policy council handbook: risk characterization. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA 100-B-00-002. Available from:
.
U.S. EPA. (2000b) Benchmark dose technical guidance document [external review draft]. Risk Assessment Forum,
Washington, DC; EPA/630/R-00/001. Available from: .
U.S. EPA. (2000c) Supplementary guidance for conducting for health risk assessment of chemical mixtures. Risk
Assessment Forum, Washington, DC; EPA/630/R-00/002. Available from:
.
U.S. EPA. (2002a) A review of the reference dose and reference concentration processes. Risk Assessment Forum,
Washington, DC; EPA/630/P-02/0002F. Available from: .
U.S. EPA. (2002b) Toxicological Review and IRIS Summary of 1,3-Butadiene. Integrated Risk Information
System, National Center for Environmental Assessment, Washington, DC. Accessed at
http://www.epa.gov/EPA/iris/subst/0139.htm.
U.S. EPA. (2002c) Memorandum: Interim genomics policy. Office of the Administrator Science Policy Council.
U.S. EPA. (2004) Potential implications of genomics for regulatory and risk assessment applications at EPA.
Prepared by the Genomics Task Force Workgroup. EPA 100/B-04/002.
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum, Washington, DC;
EPA/630/P-03/001B. Available from: .
U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure to carcinogens. Risk
Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available from: .
U.S. EPA. (2006a) Science policy council handbook: peer review. Third edition. Office of Science Policy, Office
of Research and Development, Washington, DC; EPA/100/B-06/002. Available from:
.
U.S. EPA. (2006b) A framework for assessing health risk of environmental exposures to children. National Center
for Environmental Assessment - Washington, DC, EPA/600/R-05/093F. Available from:
.
U.S. EPA. (2006c) Technology transfer network 1999 national-scale air toxics assessment. Available from
http://www.gov/ttn/atw/nata/.
U.S. EPA. (2006d) Air quality criteria for ozone and related photochemical oxidants (Final). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-05/004aF-cF.
U.S. EPA (2008) Reregistration eligibility decision for formaldehyde and paraformaldehyde. Office of Pesticide
Programs. Available from: http://www.epa.gov/pesticides/reregistration/formaldehyde/.
U.S. EPA (2009a) Formaldehyde: Chemical Report. TRI Explorer. Available from
http://www.epa.gov/triexplorer/chemical.htm.
U.S. EPA (2009b) AirData. Available from http://www.epa.gov/air/data/index.html.
Uba, G; Pachorek, D; Bernstein, J; et al. (1989) Prospective study of respiratory effects of formaldehyde among
healthy and asthmatic medical students. Am J Ind Med 15(1):91-101.
This document is a draft for review purposes only and does not constitute Agency policy.
R-47 DRAFT—DO NOT CITE OR QUOTE
-------
Uotila, L; Koivusalo, M. (1974a) Formaldehyde dehydrogenase from human liver. Purification, properties, and
evidence for the formation of glutathione thiol esters by the enzyme. J Biol Chem 249:7653-7663.
Uotila, L, Koivusalo, M. (1974b) Purification and properties of S-formylglutathione hydrolase from human liver. J
Biol Chem 249:7664-7672.
Ura, H; Nowak, P; Litwin, S; et al. (1989) Effects of formaldehyde on normal xenotransplanted human
tracheobronchial epithelium. AmJPathol 134:99-106.
Usanmaz, SE; Akarsu, ES; Vural, N. (2002) Neurotoxic effects of acute and subacute formaldehyde exposures in
mice. Environ Toxicol Pharmacol 11:93-100.
USDA (U.S. Department of Agriculture). (1979) Food intakes: Individuals in 48 States, Year 1977-78, Nationwide
Food Consumption Survey 1977-78, Rep No 1-1. Technical report no. PB91-103523INZ.
van Birgelen, APJM; Chou, BJ; Renne, RA; et al. (2000) Effects of glutaraldehyde in a 2-year inhalation study in
rats and mice. Toxicol Sci 55:195-205.
Van Schoor, J; Joos, GF; Pauwels, RA. (2000) Indirect bronchial hyperresponsiveness in asthma: mechanisms,
pharmacology and implications for clinical research. Eur Respir J 16:514-533.
Vandenplas, O; Fievez, P; Delwiche, JP; et al. (2004) Persistent asthma following accidental exposure to
formaldehyde. Allergy 59:115-116.
Vargova, M; Janota, S; Karelova, J; et al. (1992) Analysis of the health risk of occupational exposure to
formaldehyde using biological markers. Analysis 20:451-454.
Vasudeva, N; Anand, C. (1996) Cytogenetic evaluation of medical students exposed to formaldehyde vapor in the
gross anatomy dissection laboratory. J Am Coll Health 44:177-179.
Vaughan, TL; Strader, C; Davis, S; et al. (1986a) Formaldehyde and cancers of the pharynx, sinus and nasal cavity:
I. Occupational exposures. IntJ Cancer 38:677-683.
Vaughan, TL; Strader, C; Davis, S; et al. (1986b) Formaldehyde and cancers of the pharynx, sinus and nasal cavity:
II. Residential exposures. IntJ Cancer 38:685-688.
Vaughan, TL; Stewart, PA; Teschke, K; et al. (2000) Occupational exposure to formaldehyde and wood dust and
nasopharyngeal carcinoma. Occup Environ Med 57:376-384.
Venkatakrishna-Bhatt, H; Panchal, GM. (1992) Behavioural change in rats due to chronic oral and systemic
formaldehyde. Ind J Physiol Pharmacol 36:270-272.
Venkatakrishna-Bhatt, H and Panchal, GM (1997). Chronic exposure to formaldehyde in drinking water and
behavioral changes in rats. Ind. J. Physioil. & Allied Sci. 51(4):181-184.
Vogelstein, B; Fearon, ER; Hamilton, SR; et al. (1988) Genetic alterations during colorectal-tumor development. N
EnglJ Med 319:525-532.
Von Hippel, PH; and Wong, KY. (1971) Dynamic aspects of native DNA structure: kinetics of the formaldehyde
reaction with calf thymus DNA. J Mol Biol 61:587-613.
Wahlberg, JE. (1993) Measurement of skin-fold thickness in the guinea pig. Assessment of edema-inducing
capacity of cutting fluids, acids, alkalis, formalin and dimethyl sulfoxide. Contact Dermatitis 28:141-145.
Walker, JF. (1975) Formaldehyde. 3rd edition. Huntington, NY: R.E. Krieger Publishing Company.
Walrath, J; Fraumeni, JF, Jr. (1983) Mortality patterns among embalmers. Int J Cancer 31:407^11.
This document is a draft for review purposes only and does not constitute Agency policy.
R-48 DRAFT—DO NOT CITE OR QUOTE
-------
Walrath, J; Fraumeni, JF, Jr. (1984) Cancer and other causes of death among embalmers. Cancer Res 44:463 8-4641.
Wang, B; and Liu, DD. (2006) [Detection of formaldehyde induced developmental toxicity assessed with single cell
gel electrophoresis]. Fen Zi Xi Bao Sheng Wu Xue Bao 39:462-466.
Wang, M; Cheng, G; Villalta, PW; et al. (2007) Development of liquid chromatography electrospray ionization
tandem mass spectrometry methods for analysis of DNA adducts of formaldehyde and their application to rats
treated with N-nitrosodimethylamine or 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone. Chem Res Toxicol
20:1141-1148.
Wang, R; Zhang, Y; Lan, Q; et al. (2009) Occupational exposure to solvents and risk of non-Hodgkin lymphoma in
Connecticut women. Amer. J. Epidem. 169(2): 176-185.
Wang TS, Pysanenko A, Dryahina K, Spanel P, Smith D. (2008) Analysis of breath, exhaled via the mouth and nose,
and the air in the oral cavity. J. Breath Res. 2:1-13.
Wangenheim, J; and Bolcsfoldi, G. (1988) Mouse lymphoma L5178Y thymidine kinase locus assay of 50
compounds. Mutagenesis 3:193 -205.
Wantke, F; Demmer, CM; Tappler, P; et al. (1996a) Exposure to gaseous formaldehyde induces IgE-mediated
sensitization to formaldehyde in school-children. Clin Exp Allergy 26:276-280.
Wantke, F; Focke, M; Hemmer, W; et al. (1996b) Formaldehyde and phenol exposure during an anatomy dissection
course: a possible source of IgE-mediated sensitization. Allergy 51:837-841.
Ward Jr, BJ; Legator, MS; Pereira, MA; et al. (1983) Evaluation in man and animals of tests for the detection of
population exposures to genotoxic chemicals. In: Waters, MD; Sandhu, SS; Lewtas, J; et al.; eds. Short-Term
Bioassays in the Analysis of Complex Environmental Mixtures, New York: Plenum Press, 461-484.
Ward, Jr., JB; Hokanson, JA; Smith, ER; et al. (1984) Sperm count, morphology and fluorescent body frequency in
autopsy service workers exposed to formaldehyde. Mutat Res 30:417-424.
Watanabe, K; Sakamoto, K; Sasaki, T. (1996) Comparisons on chemically-induced mutagenicity among four
bacterial strains, Salmonella typhimurium TA102 and TA2638, and Escherichia coli WP2/pKM101 and WP2
uvrA/pKMlOl: collaborative study I. Mutat Res 361:143-155.
Waydhas, C; Weigl, K; Sies, H. (1978) The disposition of formaldehyde and formate arising from drug N-
demethylations dependent on cytochrome P-450 in hepatocytes and in perfused rat liver. Eur J Biochem 89:143-150.
Weber-Tschopp, A; Fischer, T; Grandjean, E. (1977) Irritating effects of formaldehyde on man. Int Arch Occup
Environ Health 39(4):207-218.
Wehinger, A; Schmid, A; Mechtcheriakov, S; et al. (2007) Lung cancer detection by proton transfer reaction mass-
spectrometric analysis of human breath gas. Int J Mass Spectrometry 265:49-59.
Weibel, ER. (1963) Morphometry of the human lung. Springer-Verlag, Berlin.
Weiler, E; Apfelbach, R. (1992) Age-related differences in olfactory sensitivity in Wistar rats after low-level
formaldehyde gas exposure. Chem Senses 17:717-718.
Weiss, B; Stern, S; Soderholm, SC; et al. (1996) Developmental neurotoxicity of methanol exposure by inhalation
in rats. Research Report (Health Effects Institute) 73:1-64.
Weisskopf, M; Morozova, N; O'Reilly, EJ; et al. (2009) Prospective study of chemical exposures and amyotrophic
lateral sclerosis mortality. J Neurol Neurosurg Psychiatry 80:558-561.
This document is a draft for review purposes only and does not constitute Agency policy.
R-49 DRAFT—DO NOT CITE OR QUOTE
-------
West, GB, and James HB. (2005) The origin of allometric scaling laws in biology from genomes to ecosystems:
towards a quantitative unifying theory of biological structure and organization. The Journal of Experimental
Biology 208(9): 1575-1592. Doi:10.1242/jeb.01589.
West, S; Hildesheim, A; Dosemeci, M. (1993) Non-viral risk factors for nasopharyngeal carcinoma in the
Philippines: results from a case-control study. Int J Cancer 55:722-727.
Westerlund, M; Gaiter, D; Carmine, A; et al. (2005) Tissue- and species-specific expression patterns of class I, III,
and IV ADH and ALDH1 mRNAs in rodent embryos. Cell Tissue Res 322:227-236.
WHO (World Health Organization). (1987) Laboratory manual for the examination of human semen and sperm-
cervical mucus interaction. Cambridge, UK: Cambridge University Press.
Wilcox, AJ; Horney, LF. (1984) Accuracy of spontaneous abortion recall. Am J Epidemiol 120:727-733.
Wilcox, P; Naidoo, A; Wedd, DJ; et al. (1990) Comparison of Salmonella typhimurium TA102 with Escherichia coli
WP2 tester strains. Mutagenesis 5:285-291.
Wilhelmsson, B; Holmstrom, M. (1992) Possible mechanisms of formaldehyde-induced discomfort in the upper
airways. Scand J Work Environ Health 18:403-407.
Wilkins, PJ; and Macleod, HD. (1976) Formaldehyde induced DNA-protein crosslinks in Escherichia Coli. Mutat
Res 36:11-6.
Williams, GM; Mori, H; and McQueen, CA. (1989) Structure-activity relationships in the rat hepatocyte DNA-
repair test for 300 chemicals. Mutat Res 221:263-286.
Wilmer, JW; Woutersen, RA; Appelman, LM; et al. (1987) Subacute (4-week) inhalation toxicity study of
formaldehyde in male rats: 8-hour intermittent versus 8-hour continuous exposures. J Appl Toxicol 7:15-16.
Wilmer, JW; Woutersen, RA; Appelman, LM; et al. (1989) Subchronic (13-week) inhalation toxicity study of
formaldehyde in male rats: 8-hour intermittent versus 8-hour continuous exposures. Toxicol Lett 47:287-293.
Witek, TJ, Jr; Schachter, EN; Tosun, T. (1986) Controlled human studies on the pulmonary effects of indoor air
pollution: experiences with sulfur dioxide and formaldehyde. Environ Int 12:129-135.
Witek, TJ, Jr; Schachter, EN; Tosun, T; et al. (1987) An evaluation of respiratory effects following exposure to 2.0
ppm formaldehyde in asthmatics: lung function, symptoms, and airway reactivity. Arch Environ Health 42:230-237.
Wolf, DC; Gross, EA; Lyght, O, et al. (1995) Immunohistochemical localization of p53, PCNA, and TCP-alpha
proteins in formaldehyde-induced rat nasal squamous cell carcinomas. Toxicol Appl Pharmacol 132:27-35.
Wolfe, RK. (1986) Effects of airborne pollutants on mucociliary clearance. Environ Health Perspect 66: 223-237.
Wong, O. (1983) An epidemiologic mortality study of a cohort of chemical workers potentially exposed to
formaldehyde, with a discussion on SMR and PMR. In: Gibson, J; ed. Formaldehyde toxicity. Washington, DC:
Hemisphere Publishing Corporation; pp. 256-272.
Wong, EY; Ray, R; Gao, DL; et al. (2006) Reproductive history, occupational exposures, and thyroid cancer risk
among women textile workers in Shanghai, China. Int Arch Occup Environ Health 79:251-158.
Wood, RW; Coleman, JB. (1995) Behavioral evaluation of the irritant properties of formaldehyde. Toxicol Appl
Pharmacol 130:67-72.
Wortley, P; Vaughan, TL; Davis, S; et al. (1992) A case-control study of occupational risk factors for laryngeal
cancer. BrJIndMed 49:837-844.
This document is a draft for review purposes only and does not constitute Agency policy.
R-50 DRAFT—DO NOT CITE OR QUOTE
-------
Woutersen, RA; Appelman, LM; Wilmer, JW; et al. (1987) Subchronic (13-week) inhalation toxicity study of
formaldehyde in rats. J Appl Toxicol 7:43-49.
Woutersen, RA; Garderen-Hoetmer, A; Bruijntjes, JP; et al. (1989) Nasal tumours in rats after severe injury to the
nasal mucosa and prolonged exposure to 10 ppm formaldehyde. J Appl Toxicol 9:39-46.
Wu, H-Y; Nguyen, HH; Russell, MW. (1997) Nasal lymphoid tissue (NALT) as a mucosal immune inductive site.
Scand J Immunol 46:506-513.
Wu, H; Romieu, I; Sienra-Monge, J-J; et al. (2007) Genetic variation in S-nitrosoglutathione reductase (GSNOR)
and childhood asthma. J Allergy Clin Immunol 120:322-328.
Xing S-Y; Ye. L; Wang, N-N. (2007) Toxic effect of formaldehyde on reproduction and heredity in male mice. J
JilinUniv 33:716-718.
Xu, B; Aoyama, K; Takeuchi, M; et al. (2002) Expression of cytokine mRNAs in mice cutaneously exposed to
formaldehyde. Immunol Lett 84:49-55.
Yager, JW; Cohn, KL; Spear, RC; et al. (1986) Sister-chromatid exchanges in lymphocytes of anatomy students
exposed to formaldehyde-embalming solution. MutatRes 174:135-139.
Yang, X; Zhang, YP; Chen, D; et al. (2001) Eye irritation caused by formaldehyde as an indoor air pollution—a
controlled human exposure experiment. Biomed Environ Sci 14:229-236.
Yang, YH; Xi, ZG; Chao, FH; et al. (2005) Effects of formaldehyde inhalation on lung of rats. Biomed Environ Sci
18:164-168.
Ye, X; Yan, W; Xie, H; et al. (2005) Cytogenetic analysis of nasal mucosa cells and lymphocytes from high-level
long-term formaldehyde exposed workers and low-level short-term exposed waiters. Mutat Res 588:22-27.
Yi C, Ke K, Xiaohua L and Xu Y (2007) Up-regulation of GSNO reductase in mice lungs by formaldehyde
inhalation. Bioinform Biomed Engineering 6-8: 294-297.
Ying, CJ; Yan, WS; Zhao, MY; et al. (1997) Micronuclei in nasal mucosa, oral mucosa and lymphocytes in students
exposed to formaldehyde vapor in anatomy class. Biomed Environ Sci 10:451-455.
Ying, CJ; Ye, XL; Xie, H; et al. (1999) Lymphocyte subsets and sister-chromatid exchanges in the students exposed
to formaldehyde vapor. Biomed Environ Sci 12:88-94.
Yokley, KA; Tran, H; Schlosser, PM. (2008) Sensory irritation response in rats: modeling, analysis and validation.
Bull Math Biol 70:555-588.
Young, JT. (1981) Histopathologic examination of the rat nasal cavity. Fundam Appl Toxicol 1:309-312.
Yu, L-Q; Jiang, S-F; Leng, S-G; et al. (2005) Early genetic effects on workers occupationally exposed to
formaldehyde. Chin J Prevent Med 39:392-395.
Yu, PH; Zuo, DM. (1996) Formaldehyde produced endogenously via deamination of methylamine. A potential risk
factor for initiation of endothelial injury. Atherosclerosis 120:189-197.
Yu, PH; Lai, CT; Zuo, DM. (1997) Formation of formaldehyde from adrenaline in vivo; a potential risk factor for
stress-related angiopathy. NeurochemRes 22:615-620.
Yu, PH. (1998) Increase of formation of methylamine and formaldehyde in vivo after administration of nicotine and
the potential cytotoxicity. Neurochem Res 23:1205-1210.
This document is a draft for review purposes only and does not constitute Agency policy.
R-51 DRAFT—DO NOT CITE OR QUOTE
-------
Zhang, L; Steinmaus, C; Eastmond, DA; et al. (2009) Formaldehyde exposure and leukemia: A new meta-analysis
and potential mechanisms. Mutat Res 681:150-168.
Zhang, L; Tang, X; Rothman, N; et al. (2010a) Occupational exposure to formaldehyde, hematotoxicity, and
leukemia-specific chromosome changes in cultured myeloid progenitor cells. Cancer Epidemiol Biomarkers Prev
19:80-88.
Zhang, L; Freeman, LE; Nakamura, J; et al. (2010b) Formaldehyde and leukemia: epidemiology, potential
mechanisms, and implications for risk assessment. Environ Mol Mutagen 51:181-191.
Zhao, Z; Zhang, Z; Wang, Z; et al. (2008) Asthmatic symptoms among pupils in relation to winter indoor and
outdoor air pollution in schools in Taiyaun, China. Environ Health Perspect 116:90-97.
Zhitkovich, A; and Costa, M. (1992) A simple, sensitive assay to detect DNA-protein crosslinks in intact cells and in
vivo. Carcinogenesis 13:1485-1489.
Zhong, W; Que Hee, S. (2004a) Quantitation of normal and formaldehyde-modified deoxynucleosides by high-
performance liquid chromatography/UV detection. Biomed Chromatogr 18:462-469.
Zhong, W; Que Hee, SS. (2004b) Formaldehyde-induced DNA adducts as biomarkers of in vitro human nasal
epithelial cell exposure to formaldehyde. Mutat Res 563:13-24.
Zhou, DX; Qiu, SD; Zhang, J; et al. (2006) Reproductive toxicity of formaldehyde to adult male rats and the
functional mechanism concerned. Sichuan Da Xue Xue Bao Yi Xue Ban 37:566-569.
Zielenska, M; Guttenplan, JB. (1988) Mutagenic activity and specificity of N-nitrosomethylaniline and N-
nitrosodiphenylamine in Salmonella. Mutat Res 202:269-76.
Zuercher, AW; Cebra, JJ. (2002) Structural and functional differences between putative mucosal inductive sites of
the rat. Eur J Immunol 32:3191-3196.
Zuercher, AW; Coffin, SE; Thurnheer, MC; et al. (2002) Nasal-associated lymphoid tissue is a mucosal inductive
site for virus-specific humoral and cellular immune responses. J Immunol 168:1796-1803.
Zwart, A; Woutersen, RA; Wilmer, JW; et al. (1988) Cytotoxic and adaptive effects in rat nasal epithelium after 3-
day and 13-week exposure to low concentrations of formaldehyde vapour. Toxicology 51:87-99.
-End of Volume III-
This document is a draft for review purposes only and does not constitute Agency policy.
R-52 DRAFT—DO NOT CITE OR QUOTE
-------
DRAFT - DO NOT CITE OR QUOTE EPA/635/R-10/002A
www.epa.gov/iris
f/EPA
TOXICOLOGICAL REVIEW OF
FORMALDEHYDE -
INHALATION ASSESSMENT
(CAS No. 50-00-0)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
VOLUME IV of IV
Appendices
June 2, 2010
NOTICE
This document is an External 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.
IV-ii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS IN BRIEF
VOLUME I: INTRODUCTION, BACKGROUND, AND TOXICOKINETICS
Chapter 1: Introduction
Chapter 2: B ackground
Chapter 3: Toxicokinetics
VOLUME II: HAZARD CHARACTERIZATION
Chapter 4: Hazard Characterization
VOLUME III: QUANTITATIVE ASSESSMENT, MAJOR CONCLUSIONS IN THE
CHARACTERIZATION OF HAZARD AND DOSE RESPONSE, AND
REFERENCES
Chapter 5: Quantitative Assessment: Inhalation Exposure
Chapter 6: Major Conclusions in the Characterization of Hazard and Dose-Response
References
VOLUME IV: APPENDICES
Appendix A: Summary of External Peer Review and Public Comments and Dispositions
Appendix B: Simulations of Interindividual and Adult-to-Child Variability in Reactive
Gas Uptake in a Small Sample of People (Garcia et al., 2009)
Appendix C: Lifetable Analysis
Appendix D: Model Structure & Calibration in Conolly et al. (2003, 2004)
Appendix E: Evaluation of BBDR Modeling of Nasal Cancer in the F344 Rat: Conolly
et al. (2003) and Alternative Implementations
Appendix F: Sensitivity Analysis of BBDR Model for Formaldehyde Induced
Respiratory Cancer in Humans
Appendix G: Evaluation of the Cancer Dose-Response Modeling of Genomic Data for
Formaldehyde Risk Assessment
Appendix H: Expert Panel Consultation on Quantitative Evaluation of Animal
Toxicology Data for Analyzing Cancer Risk Due to Inhaled Formaldehyde
This document is a draft for review purposes only and does not constitute Agency policy.
IV-iii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS—TOXICOLOGICAL REVIEW OF FORMALDEHYDE
(CAS No. 50-00-0)
VOLUME IV
LIST OF TABLES IV-vi
LIST OF FIGURES IV-vii
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITIONS A-2
APPENDIX B: SIMULATIONS OF INTERINDIVIDUAL AND ADULT-TO-CHILD
VARIABILITY IN REACTIVE GAS UPTAKE IN A SMALL
SAMPLE OF PEOPLE (Garcia et al., 2009) B-2
APPENDIX C: LIFETABLE ANALYSIS C-2
APPENDIX D: MODEL STRUCTURE & CALIBRATION IN CONOLLY ET AL.
(2003,2004) D-2
D.I. DPX AND MUTATIONAL ACTION D-4
D.2. CALIBRATION OF MODEL D-4
D.3. FLUX BINS D-5
D.4. USE OF LABELING DATA D-5
D.5. UPWARD EXTRAPOLATION OF NORMAL CELL DIVISION RATE D-5
D.6. INITIATED CELL DIVISION AND DEATH RATES D-7
D.7. STRUCTURE OF THE CUT HUMAN MODEL D-7
APPENDIX E: EVALUATION OF BBDR MODELING OF NASAL CANCER IN
THE F344 RAT: CONOLLY ET AL. (2003) AND ALTERNATIVE
IMPLEMENTATIONS E-2
E. 1. TABULATION OF ALL IS SUES EVALUATED IN THE RAT MODELS E-2
E.2. STATISTICAL METHODS USED IN EVALUATION E-7
E.3. PRIMARY UNCERTAINTIES IN BBDR MODELING OF THE F344 RAT
DATA E-8
E.3.1. Sensitivity to Use of Historical Controls E-9
E.3.1.1. Use of Historical Controls E-9
E.3.1.2. Influence of Historical Controls on Model Calibration and on
Human Model E-10
E.3.1.3. Influence of Historical Controls on Dose-Response Curve E-13
E.3.1.4. Problem Including 1976 Study for Inhalation Historical Control E-13
E.3.1.5. Effect of Control Data on MOA Inferences E-14
E.3.2. Characterization of Uncertainty-Variability in Cell Replication Rates E-14
E.3.2.1. Dose-Response for O,N as Used in the CUT Clonal Growth
Modeling E-14
This document is a draft for review purposes only and does not constitute Agency policy.
IV-iv DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
E.3.2.2. Time Variability in Labeling Data E-17
E.3.2.3. Site and Time Variability in Derived Cell Replication Rate E-18
E.3.2.4. Alternate Dose-Response Curves for Cell Replication E-22
E.3.3. Uncertainty in Model Specification of Initiated Cell Replication and Death ...E-30
E.3.3.1. Biological Implications of Assumptions in Conolly et al. (2003) E-30
E.3.3.2. Plausible Alternative Assumptions for 0,1 and Pi E-32
E.3.4. Results of Sensitivity Analyses on O,N, oti, and PI E-33
E.3.4.1. Further Constraints E-33
E.3.4.2. Sensitivity of Risk Estimates for the F344 Rat E-34
E.3.4.3. MO A Inferences Revisited E-41
E.3.4.4. Confidence Bounds: Model Uncertainty Versus Statistical
Uncertainty E-42
APPENDIX F: SENSITIVITY ANALYSIS OF BBDR MODEL FOR
FORMALDEHYDE INDUCED RESPIRATORY CANCER IN
HUMANS F-2
F.I. MAJOR UNCERTAINTIES IN THE FORMALDEHYDE HUMAN BBDR
MODEL F-2
F.2. SENSITIVITY ANALYSIS OF HUMAN BBDR MODELING F-4
F.2.1. Effect of Background Rates of Nasal Tumors in Rats on Human Risk
Estimates F-4
F.2.2. Alternative Assumptions Regarding the Rate of Replication of Initiated
Cells F-6
F.2.3. Biological Plausibility of Alternate Assumptions F-10
F.2.4. Effect of Alternate Assumptions for Initiated Cell Kinetics on Human
Risk Estimates F-13
APPENDIX G: EVALUATION OF THE CANCER DOSE-RESPONSE MODELING G-2
G.I. MAJOR CONCLUSIONS IN ANDERSEN ET AL. (2008) G-2
G.2. USE OF MULTIPLE FILTERS ON THE DATA G-3
G.3. DATA FOR LOW-DOSE CANCER RESPONSE G-4
G.4. DIFFICULTIES IN INTERPRETING THE BENCHMARK MODELING G-5
G.5. STATISTICAL SENSITIVITY OF THE DATA FOR DOSE-RESPONSE G-6
G.6. LENGTH OF THE STUDY AND STOCHASTIC EVENTS G-6
G.7. OVERALL CONCLUSION G-7
APPENDIX H: EXPERT PANEL CONSULTATION ON QUANTITATIVE
EVALUATION OF ANIMAL TOXICOLOGY DATA FOR
ANALYZING CANCER RISK DUE TO INHALED FORMALDEHYDE. H-2
This document is a draft for review purposes only and does not constitute Agency policy.
IV-v DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table B-1. Variations in overall nasal uptake, whole nose flux, and key parameters B-4
Table C-l. Extra risk calculation for environmental exposure to 0.0461 ppm
formaldehyde (the LEC0005 for NPC incidence) using a log-linear
exposure-response model based on the cumulative exposure trend
results of Hauptmann et al. (2004), as described in Section 5.2.2 C-3
Table E-l. Evaluation of assumptions and uncertainties in the CUT model for nasal
tumors in the F344 rat E-3
Table E-2. Influence of control data in modeling formaldehyde-induced cancer in
theF344rat E-ll
Table E-3. Variation in number of cells across nasal sites in theF344 rat E-20
Table E-4. Parameter specifications and estimates for clonal growth models of
nasal SCC in the F344 rat using alternative characterization of cell
replication and death rates E-39
Table E-5. Parameter specifications and estimates for clonal growth models of nasal
SCC in the F344 rat using cell replication and death rates as characterized
in Conolly et al. (2003) E-40
Table E-6. Comparison of statistical confidence bounds on added risk for two models E-42
Table F-1. Summary of evaluation of maj or assumptions and results in CUT human
BBDR model F-3
This document is a draft for review purposes only and does not constitute Agency policy.
IV-vi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
Figure B-l. Gas flux across the nasal lining for the case of a "maximum uptake" gas
in Garcia et al. (2009) as a function of axial distance from the nostril B-5
Figure D-l. Dose response of normal (O,N) and initiated (0,1) cell division rate in
Conolly et al. (2003) D-6
Figure D-2. Flux dependence of ratio of initiated and normal cell replication rates
(
-------
Figure F-2. Conolly et al. (2003) hockey-stick model for division rates of initiated
cells in rats and two modified models F-7
Figure F-3. Conolly et al. (2003) J-shape model for division rates of initiated cells in
rats and two modified models F-8
Figure F-4. Very similar model estimates of probability of fatal tumor in rats for
three models in Figure F-2 F-9
Figure F-5. Cell proliferation data from Meng et al. (2010) F-12
Figure F-6. Graphs of the additional human risks estimated by applying these
modified models for ai, using all NTP controls, compared to those
obtained using the original Conolly et al. (2004) model F-13
Figure G-l. Graphs of epithelial hyperplasia versus formaldehyde concentration
with 95% confidence intervals G-7
This document is a draft for review purposes only and does not constitute Agency policy.
IV-viii DRAFT—DO NOT CITE OR QUOTE
-------
Appendix A
-------
1 APPENDIX A
2
3 SUMMARY OF EXTERNAL PEER REVIEW
4 AND PUBLIC COMMENTS AND DISPOSITIONS
5
6
7
8
9
10
11
12
13
14
15 [NOTE: This is a placeholder for Appendix A which will be drafted
16 following External Peer review and receipt of public comments.]
17
This document is a draft for review purposes only and does not constitute Agency policy.
A-2 DRAFT—DO NOT CITE OR QUOTE
-------
Appendix B
-------
1 APPENDIX B
2
3 SIMULATIONS OF INTERINDIVIDUAL AND ADULT-TO-CHILD VARIABILITY IN
4 REACTIVE GAS UPTAKE IN A SMALL SAMPLE OF PEOPLE
5 (GARCIA ET AL., 2009)
6
7
8 Garcia et al. (2009) used computational fluid dynamics to study human variability in the
9 nasal dosimetry of model reactive, water-soluble gases in 5 adults and 2 children, aged 7 and 8
10 years old. They considered two model categories of gases, corresponding to maximal and
11 moderate absorption at the nasal lining. This Appendix was developed in response to EPA
12 reviewers' suggestions that results from the Garcia et al. (2009) work should be used to inform
13 the uncertainty factor considered for interhuman variability in this document. Furthermore the
14 tumor incidence in F344 rats have been used to extrapolate the risk of cancer in the human
15 respiratory tract. This extrapolation was based on internal dose metrics derived using a CFD
16 model constructed from the nasal passages of a single individual (Subramaniam et al. 1998). The
17 adults considered in the Garcia et al. study included that individual.
18 Garcia et al. (2009) mapped out the nasal airway (including the nasopharynx) geometries
19 of these individuals using magnetic resonance imaging or computed tomography scans. The
20 scans chosen for the analysis were from individuals who had normal nasal anatomies with no
21 pathology (as per a review carried out by a ear-nose-throat surgeon). The minute volumes of
22 these individuals were ranged from 6.8 to 9.0 L/min (adults) and 5.5 to 5.8 L/min (children).
23 The sample size in this study is too small to consider the results representative of the population
24 as a whole (as also recognized by the authors). Nonetheless, various comparisons with the
25 characteristics of other study populations add to the strength of this study; we therefore
26 evaluated this study further in this document partly with the goal of impacting on research
27 directions and future interpretations for specific gases. The range of adult minute volumes in
28 this study is reported by the authors to be in good agreement with that obtained in many other
29 studies in the literature. Minute volumes for the children in the study were found to be similar to
30 the average minute volume of 6.1 ± 1.7 L/min obtained by Bennett and Zeman (2004) in a study
31 of 36 children aged 6 to 13 years; the range of nasal surface area values for the adults agreed
32 well with that obtained by Guilmette et al. (1997) for 45 adults; and the range of values for the
33 surface area to volume ratio is in good agreement with that obtained for 40 adult Caucasians
34 studied by Yokley (2006). The surface area to volume ratio is useful for comparing the rate of
35 diffusional transport of a gas out of different cavities; however in the case of the highly
36 nonhomogeneously shaped nasal lumen, this should only be considered a gross indicator.
This document is a draft for review purposes only and does not constitute Agency policy.
B-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 We focus here only on the "maximum uptake" simulations in Garcia et al. (2009). In this
2 case, the model gas was considered so highly reactive and soluble that it was reasonable to
3 assume an infinitely fast reaction of the absorbed gas with compounds in the airway lining.
4 Although such a gas could be reasonably considered a proxy for formaldehyde, these results
5 cannot be utilized to inform quantitative estimates of formaldehyde dosimetry (and that does not
6 appear to have been the intent of the authors either). This is because the same boundary
7 condition corresponding to maximal uptake was applied on the vestibular section as well as on
8 the respiratory and transitional epithelial lining of the nasal cavity. This is not appropriate for
9 formaldehyde as the lining on the nasal vestibule is made of keratinized epithelium which is
10 considerably less absorbing than the transitional or respiratory epithelium (Kimbell et al.,
11 200 la).
12 Table B-l provides results obtained by Garcia et al. (2009) for gas nasal uptake in five
13 adults and two children for the maximal uptake scenario. Although the nasal cavities of the
14 children were smaller in surface area, volume and length, the surface-area-to-volume ratios were
15 similar in the two age groups. Overall uptake efficiency, average flux (rate of gas absorbed per
16 unit surface area of the nasal lining) and maximum flux levels over the entire nasal lining did not
17 vary substantially between adults (1.6-fold difference in average flux and much less in maximum
18 flux), and the mean values of these quantities were comparable between adults and children. The
19 comparisons between adults and children are in agreement with conclusions reached by Ginsberg
20 et al. (2005) that overall extrathoracic absorption of highly and moderately reactive and soluble
21 gases (corresponding to category 1 and 2 reactive gases as per the scheme in EPA [1994]) is
22 similar in adults and children. However, the interindividual variations in each of these three
23 quantities alone are limited in their ability to characterize variability in the interaction of the gas
24 with the nasal lining. For a very reactive and soluble gas, regional absorption of the gas is highly
25 nonhomogeneously distributed over the nasal lining; interindividual variations due to differing
26 spatial patterns of this distribution between individuals could potentially be diluted when flux is
27 averaged over the whole nose. Estimates of maximum gas flux, on the other hand, correspond to
28 extremely small localized regions of hot spots (see Chapter 3), and thus interindividual
29 differences in this quantity provide limited perspective on interindividual variability in flux
30 distribution patterns over the whole nose. Furthermore, numerical error in the calculation (such
31 as mass balance and irregularly shaped elements of the finite-element mesh) is likely to be most
32 pronounced when estimates are considered over extremely small regions. We do not know to
33 what extent these errors impact upon the accuracy in calculations of maximum flux.
This document is a draft for review purposes only and does not constitute Agency policy.
B-3 DRAFT—DO NOT CITE OR QUOTE
-------
Table B-l. Variations in overall nasal uptake, whole nose flux, and key
parameters
1
2
3
4
5
6
7
10
11
12
13
14
15
16
17
18
19
20
21
adultl
adultl
adultl
adultl
adultl
childl
child2
% nasal
93.
92.
93.
89.
91.
uptake
5
4
1
2
5
92
88.
2
MV
(L)
9
6.8
9
7.1
6.9
5.5
5.8
SA/V
(I/mm)
1.12
1.09
0.88
0.87
0.95
1.13
0.95
Avg flux
10"8kg/(s.m2)
left cavity right
1.8
1.5
1.6
1.2
1.4
1.9
1.6
1
1
cavity
.5
.5
1.3
1
1
1
1
.2
.5
.5
.5
Maximum
10'8kg/(s.
left cavity rig
10.8
10.8
11
10.6
10.8
11.8
12.3
flux
m2)
;ht cavity
10.0
10.4
10.6
10.2
10.0
11.0
11.6
MV = minute volume, SA = nasal surface area, V = nasal volume.
Source: Garcia et al. (2009).
On the other hand, Figure 6A of Garcia et al. (2009), reproduced here as Figure B-l,
provides a different perspective on interhuman variability in flux values at specific points on the
nasal walls. In this figure gas flux across the nasal lining is plotted as function of distance from
the nostril along the septal axis of the nose, normalized by the total nasal length along the septal
axis of each subject. The local flux of formaldehyde varies among individuals by a factor of 3 to
5 at various normalized distances along the septal axis of the nose. However, interpretation of
the values in this plot is problematic for reasons explained in their paper:l
The greater variability among individuals seen for wall fluxes at specific sites of
the nasal passages (Figure 6) in comparison to the minimal variability in total
uptake (Table 2) and whole-nose dose (Tables 3 and Tables 4) indicates that
fluxes of equal magnitude do not exactly overlay the same anatomical regions of
the nasal cavity in each individual. This implies that specific anatomical regions
subtended by maximum flux could be offset from one individual to another.
Notwithstanding this difficulty in interpretation, we believe the extents of vertical bars on each
point plotted in Figure B-l provide a better perspective of the interindividual (adult) variability
in local flux than the variation in whole nose average or in maximum flux presented in Table
B-l.
1 The figures and tables in the cited text refer those in Garcia et al. (2009).
This document is a draft for review purposes only and does not constitute Agency policy.
B-4 DRAFT—DO NOT CITE OR QUOTE
-------
A 1.0E-07
-m- Adults - maximum uptake
-*- Children-maximum uptake
1.0E-08
1.0E-09
-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
(Distance from nostrils)/(Length of septum)
Figure B-l. Gas flux across the nasal lining for the case of a "maximum
uptake" gas in Garcia et al. (2009) as a function of axial distance from the
nostril. The vertical bars show range of variation. See the paper for further
details. Figure is reproduced from Garcia et al. (2009).
1 Clearly, multiple measures of variability in dose can be developed depending on the
2 adverse response. The advantage of models such as that developed by Garcia et al. is that they
3 make it possible to explicitly carry out these calculations. For example, if deficit in pulmonary
4 function is the adverse response, and the mechanism of action was a function of total dose to the
5 lung, then interindividual variation in mean whole nose flux or overall nasal uptake efficiency
6 would be most useful. It is possible to conceive of allergic or irritation responses being triggered
7 by some threshold value of local flux. In such a case it may be preferable to calculate the
8 variability associated with the net surface area receiving flux values greater than that threshold.
9 On the other hand, the probability of developing a tumor at a nasal site may be nonlinearly
10 related to the flux at that site and linearly related to the number of cells at that site. In this case,
11 the appropriate metric may be the nasal surface area associated with some intermediate levels of
12 local flux (see appendix in Subramaniam et al., 2008).
13 Various caveats presented by the authors as limitations of their study should be noted:
14 Possible nonuniform distribution of epithelial types, enzymes, glands and other cellular
15 metabolic or clearance machinery were not considered in the model; only effects pertaining to
16 resting breathing were considered; the study sample size was small; children younger than
17 7 years old were not studied; and, the model assumed a rigid nasal geometry.
This document is a draft for review purposes only and does not constitute Agency policy.
B-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 Garcia et al. (2009) conclude their paper as follows:
2
3 "..., our simulations predicted no differences in the nasal dosimetry of reactive,
4 water-soluble gases between children and adults, suggesting that the risk factor of
5 10 typically used to accommodate interhuman variability is adequate."
6
7 In addition to the caveats already recognized by the authors, the above conclusion needs further
8 qualification:
9
10 1. While the uncertainty factor of 10 that is typically applied for interhuman variability is
11 generally considered to be protective of children, it is not based on variations between
12 children and adults. (If there is reasonable evidence that children are more sensitive than
13 adults, the 10-fold factor may be considered inadequate.)
14 2. Assuming that the adverse response under consideration is one for which the localized
15 nature of reactive gas flux across the nasal lining is important, the calculations such as
16 those shown in Figure B-l for the model gas are very relevant to the discussion of
17 interindividual variability. The 3 to 5-fold variation in the local gas flux between adults
18 (and also between the children) in the small sample size in this simulation may be
19 compared with the value of 3.3 used for the pharmacokinetic component of the
20 uncertainty factor for interhuman variability in susceptibility. (EPA practice is often to
21 split this 10-fold uncertainty factor into pharmacokinetic and pharmacodynamic
22 components of 3.3 each.)
This document is a draft for review purposes only and does not constitute Agency policy.
B-6 DRAFT—DO NOT CITE OR QUOTE
-------
Appendix C
-------
1 APPENDIX C
2
3 LIFETABLE ANALYSIS
4
5
6 A spreadsheet illustrating the extra risk calculation for the derivation of the lower 95%
7 bound on the effective concentration associated with a 0.05% extra risk (LECooos) for
8 nasopharyngeal carcinoma (NPC) incidence is presented in Table C-l.
9
This document is a draft for review purposes only and does not constitute Agency policy.
C-2 DRAFT—DO NOT CITE OR QUOTE
-------
This document is a draft for C_3 DIL
review purposes only and does
not constitute Agency policy.
Table C-l. Extra risk calculation3 for environmental exposure to 0.0461 ppm formaldehyde (the LEC0005 for
NPC incidence)11 using a log-linear exposure-response model based on the cumulative exposure trend results of
Hauptmann et al. (2004), as described in Section 5.2.2
A
Interval
number
(i)
1
2
3
4
5
6
7
8
9
10
11
lIII
13
14
I5
16
17
18
B
Age
interval
<1
1-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-59
80-84
C
All cause
mortality
(xl05/yr)
728.7
32.9
16.4
20.9
68.2
96
99
116.3
162.2
237.3
356
518.6
801.8
1257.9
1928.2
2968.1
4556.6
7399.6
D
NPC
incidence
(xl05/yr)
0
0.05
0.03
0.09
0.12
0.16
0.23
0.48
0.55
1.14
1.3
1.72
1.69
1.9 [
2.87
2.1
2.19
1.98
E
All
cause
hazard
rate (h*)
F
Prob of
surviving
interval
(q)
0.0073 | 0.9927
0.0013 I 0.9987
0.0008
0.0010
0.9992
0.9990
0.0034 ' 0.9966
0.0048
0.0050
0.9952
0.9951
0.0058 | 0.9942
0.0081
0.0119
0.0178
0.0259
0.0401
0.0629
0.0964
0.9919
0.9882
0.9824
0.9744
0.9607
0.9390
0.9081
0.1484 1 0.8621
0.2278 I 0.7963
0.3700 I 0.6907
G
Prob of
surviving
up to
interval
(S)
1.0000
0.9927
0.9914
0.9906
0.9896
0.9862
0.9815
0.9766
0.9710
0.9631
0.9518
0.9350
0.9111
0.8753 [
0.8219
0.7464
0.6434
0.5123
H
NPC
cancer
hazard
rate (h)
I
Cond
prob of
NPC
incidence
in
interval
(Ro)
o.ooooo IO.OOOOOG
0.00000
0.00000
0.00000
0.000002
0.000001
0.000004
0.00001 '0.000006
0.00001 0.000008
0.00001 0.000011
J
Exp
duration
mid
interval
(xtime)
0
0
0
0
2.5
7.5
12.5
0.00002 1 0.000023| 17.5
0.00003
0.00006
0.00007
0.00009
0.00008
0.00010
0.00014
0.000027
0.060055"
0.006061"
0.000079
0.000075
o.oooosf
0.066112
0.00011 10.000073
0.00011 10.000063
0.00010 |0.000042
22.5
27.5
32.5
37.5
42.5
47.5
52.5
57.5
62.5
67.5
K
Cum
exp mid
interval
(xdose)
0.0000
0.0000
0.0000
0.0000
0.3506
1.0517
1.7528
2.4539
3.1550
3.8561
4.5572
5.2583
5.9594
6.6605
7.3616
8.0627
8.7638
9.4649
L
Exposed
NPC
hazard
rate (hx)
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0001
0.0001
0.0001
0.0001
0.0002
0.0003
0.0002
0.0002
0.0002
M
Exposed
all cause
hazard
rate
(h*x)
0.0073
0.0013
0.0008
0.0010
0.0034
0.0048 ,
0.0050
0.0058
0.0081
0.0119
0.0178
0.0260
0.0401
0.0630 [
0.0965 "
0.1485
0.2279
0.3701
N
Exposed
prob of
surviving
interval
(qx)
0.9927
0.9987
0.9992
0.9990
0.9966
0.9952
0.9951
0.9942
0.9919
0.9882
0.9823
0.9744
0.9607
0.9390
0.9080
0.8620
0.7962
0.6907
O
Exposed
prob of
surviving
up to
interval
(Sx)
1.0000
0.9927
0.9914
0.9906
0.9896
0.9862
0.9815
0.9766
0.9710
0.9631
0.9517
0.9349
0.9110
0.8751
0.8217
0.7461
0.6431
0.5120
P
Exposed
cond
prob of
NPC in
interval
(Rx)
.....0.00000
0.00000
0.00000
.....0,00000
0.00001
0.00001
0.00001
0.00003
0.00003
0.00008
0.00009
0.00012
0.00012
0.00014
0,00021
0.00014
....0,00013
0.00009
| Ro= 1 0.000725| | Rx = 1 0.001225
Extra Risk = (Rx-Ro)/(l-Ro) = 0.0005
O
O
-------
5 KO
S. S3
Table C-l. Extra risk calculation21 for environmental exposure to 0.0461 ppm formaldehyde (the LEC0005 for
NPC incidence)13 using a log-linear exposure-response model based on the cumulative exposure trend results of
Hauptmann et al. (2004), as described in Section 5.2.2 (continued)
§
O
fe
H
Column B:
Column C:
Column D:
Column E:
Column F:
Column G:
Column H:
Column I:
Column J:
Column K:
ColumnL:
Column M:
Column N:
Column O:
Column P:
5-year age interval (except <1 and 1-4) up to age 85.
all-cause mortality rate for interval i (x 105/year) (2000 data from NCHS).
NPC incidence rate for interval i (x 105/year) (1996-2000 SEER data).
all-cause hazard rate for interval i (h*0 (= all-cause mortality rate x number of years in age interval).0
probability of surviving interval i without being diagnosed with NPC (qO (= exp(-h*0)-
probability of surviving up to interval i without having been diagnosed with NPC (SO (Si = 1; S; = S;-i x qM, for /' > 1).
NPC incidence hazard rate for interval i (hO (= NPC incidence rate x number of years in interval).
conditional probability of being diagnosed with NPC in interval i (= (Vh^) x s; x (1-qO), i.e., conditional upon surviving up to interval i without
having been diagnosed with NPC [Ro, the background lifetime probability of being diagnosed with NPC = the sum of the conditional probabilities
across the intervals].
exposure duration (in years) at mid-interval (xtime).
cumulative exposure mid-interval (xdose) (= exposure level (i.e., 0.0461 ppm) x 365/240 x 20/10 x xtime) [365/240 x 20/10 converts continuous
environmental exposures to corresponding occupational exposures].
NPC incidence hazard rate in exposed people for interval i (hx;) (= h; x (1 + p x xdose), where p = 0.05183 + (1.645 x 0.01915) = 0.08333) [0.05183
per ppm x year is the regression coefficient obtained, along with its SE of 0.01915, from Dr. Hauptmann (see Section 5.2.2. 1). To estimate the
LECooos, i.e., the 95% lower bound on the continuous exposure giving an extra risk of 0.05%, the 95% upper bound on the regression coefficient is
used, i.e., MLE + 1.645 x SE].
all-cause hazard rate in exposed people for interval i (h*xO (= h*; + (hxj - h^).
probability of surviving interval i without being diagnosed with NPC for exposed people (qxO (= exp(-h*xO).
probability of surviving up to interval i without having been diagnosed with NPC for exposed people (SxO (Sxi = 1; Sx; = Sx;-i
, for /' > 1).
conditional probability of being diagnosed with NPC in interval i for exposed people (= (hxi/h*Xi) x Sx; x (1-qx^) [Rx, the lifetime probability of
being diagnosed with NPC for exposed people = the sum of the conditional probabilities across the intervals] .
a Using the methodology of BEIRIV (1988).
b The estimated 95% lower bound on the continuous exposure level of TCE that gives a 0.05% extra lifetime risk of NPC.
0 For the cancer incidence calculation, the all-cause hazard rate for interval i should technically be the rate of either dying of any cause or being diagnosed with
the specific cancer during the interval, i.e., (the all-cause mortality rate for the interval + the cancer-specific incidence rate for the interval—the cancer-specific
mortality rate for the interval [so that a cancer case isn't counted twice, i.e., upon diagnosis and upon death]) x number of years in interval. This adjustment
was ignored here because the NPC incidence rates are small compared with the all-cause mortality rates.
MLE = maximum likelihood estimate, SE = standard error
O
O
-------
Appendix D
-------
1 APPENDIX D
2
3 MODEL STRUCTURE & CALIBRATION IN CONOLLY ET AL. (2003, 2004)
4
5
6 The various studies indicated in Section 5.4.1 were followed by the development of a
7 biologically motivated dose-response model for formaldehyde-induced cancer in the respiratory
8 tract. These efforts are represented in a series of papers and in a health assessment report (CUT
9 model) (Conolly et al., 2004, 2003, 2000; Conolly, 2002; Kimbell et al., 2001a, b; Overton et al.,
10 2001; CUT, 1999). The CUT modeling and available data, and alternatives based on their
11 original model were evaluated extensively for the purpose of this assessment and utilized in
12 calculating the cancer potency. EPA's cancer guidelines (U.S. EPA, 2005a) suggest using a
13 BBDR model for extrapolation when data permits since it facilitates the incorporation of MO A
14 in risk assessment
15 In Conolly et al. (2003), tumor incidence data in the above long-term bioassays were
16 modeled by using an approximation of the two-stage clonal growth model (Moolgavkar et al.,
17 1988) and allowing formaldehyde to have directly mutagenic action. Conolly et al. (2003)
18 combined these data with historical control data on 7,684 animals obtained from National
19 Toxicology Program (NTP) bioassays. These models are based on the Moolgavkar, Venzon, and
20 Knudson (MVK) stochastic two-stage model of cancer (Moolgavkar et al., 1988; Moolgavkar
21 and Knudson, 1981; Moolgavkar and Venzon, 1979), which accounts for growth of a pool of
22 normal cells, mutation of normal cells to initiated cells, clonal expansion and death of initiated
23 cells, and mutation of initiated cells to fully malignant cells.
24 The MVK model for formaldehyde accounted for two MO As that may be relevant to
25 formaldehyde carcinogenicity:
26
27 • An indirect MOA in which the regenerative cell proliferation in response to
28 formaldehyde cytotoxicity increased the probability of errors in DNA replication. This
29 MOA was modeled by using labeling data on normal cells in nasal mucosa of rats
30 exposed to formaldehyde.
31 • A possible direct mutagenic MOA, based on information indicating that formaldehyde is
32 mutagenic (Speit and Merk, 2002; Heck et al., 1990; Grafstrom et al., 1985), was
33 modeled by using rat and rhesus monkey data on formaldehyde production of DPXs.
34
35 The human model for formaldehyde carcinogenicity (Conolly et al., 2004) is
36 conceptually very similar to the rat model. The model uses, as input, results from a dosimetry
37 model for an anatomically realistic representation of the human upper airways and an idealized
This document is a draft for review purposes only and does not constitute Agency policy.
D-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 representation of the lower airways. However, the model does not incorporate any data on
2 human responses to formaldehyde exposure. The rat and human formaldehyde models are
3 detailed further below.
4 The following notations are used in the rest of this appendix:
5
6 N cell, normal cell
7 I cell, initiated cell
8 LI, labeling index (number of labeled cells/(number labeled + unlabeled cells)
9 ULLI, unit length labeling index (number labeled cells/length of basement membrane)
10 N, number of normal cells that are eligible for progression to malignancy
11 (XN, division rate of normal cells (hours"1)
12 HN, rate at which an initiated cell is formed by mutation of a normal cell (per cell division
13 of normal cells)
14 oti, division rate of an initiated cell (hours"1)
15 PI; death rate of an initiated cell (hours"1)
16 jii; rate at which a malignant cell is formed by mutation of an initiated cell (per cell
17 division of initiated cells)
18
19 A novel contribution of the CUT model is that cell replication rates and DPX
20 concentrations are driven by local dose, which is formaldehyde flux to each region of nasal
21 tissue expressed as pmol/mm2-hour. This dosimetry is predicted by computational fluid
22 dynamics (CFD) modeling using anatomically accurate representations of the nasal passages (see
23 Chapter 3). Such a feature is important to incorporating site-specific toxicity in the case of a
24 highly reactive gas like formaldehyde, for which uptake patterns are spatially localized and
25 significantly different across species (see Chapter 3). In the CUT model, each of these
26 parameters is characterized by local flux. The inputs to the two-stage cancer modeling consisted
27 of results from other model predictions as well as empirical data as follows:
28
29 • Regional uptake of formaldehyde in the respiratory tract predicted by using CFD
30 modeling in the F344 rat and human (Kimbell et al., 2001a, b; Overton et al., 2001;
31 Subramaniam et al., 1998)
32 • Concentrations of DPXs predicted by a PBPK model (Conolly et al., 2000) calibrated to
33 fit the DPX data in F344 rat and rhesus monkey (Casanova et al., 1994, 1991) and
34 subsequently scaled up to humans
This document is a draft for review purposes only and does not constitute Agency policy.
D-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 • ON inferred from LI data on rats exposed to formaldehyde (Monticello et al., 1996, 1991,
2 1990)
3 D.I. DPX AND MUTATIONAL ACTION
4 Formaldehyde interacts with DNA to form DPXs. These cross-links are considered to
5 induce mutagenic as well as clastogenic effects. Casanova et al. (1994, 1989) carried out two
6 studies of DPX measurements in F344 rats. In the first study, rats were exposed to
7 concentrations of 0.3, 0.7, 2, 6, and 10 ppm for 6 hours and DPX measurements were made over
8 the whole respiratory mucosa of the rat, while, in the second study, the exposure was to 0.7, 2, 6,
9 or 15 ppm formaldehyde for 3 hours and measurements were made at "high" and "low" tumor
10 sites. Overall, these studies showed statistically significantly elevated levels of DPXs at
11 concentrations >2 ppm, with the trend also indicating elevated DPXs at 0.7 ppm. In Conolly et
12 al. (2003), DPX formation is considered proportional to the intracellular dose that induces
13 mutations. Conolly et al. (2000) used data from the second study to develop a PBPK model that
14 predicted the time course of DPX concentrations as a function of regional formaldehyde flux
15 (estimated in the CFD modeling and expressed as pmol/mm2-hour). In Conolly et al. (2003), this
16 PBPK model was then used to predict regional DPX concentrations (that is, as a function of
17 regional formaldehyde flux) (see Figure 5-11, Chapter 5). These data were incorporated into the
18 two-stage clonal expansion model by defining the mutation rate of normal and initiated cells as
19 the same linear function of DPX concentration as follows:
20
21 UN = Mi = UNbasai + KMU x DPX (D-l)
22
23 The unknown constants UNbasai and KMU were estimated by fitting model predictions to the
24 tumor bioassay data.
25
26 D.2. CALIBRATION OF MODEL
27 The rat model in Conolly et al. (2003) involved six unknown statistical parameters that
28 were estimated by fitting the model to the rat formaldehyde bioassay data shown in Table 5-24 in
29 Chapter 5 (Monticello et al., 1996; Kerns et al., 1983) plus historical data from several thousand
30 control animals from all the rat bioassays conducted by the NTP. These NTP bioassays were
31 conducted from 1976 through 1999 and included 7,684 animals with an incidence of 13 SCCs
32 (i.e., 0.17% incidence). The resulting model predicts the probability of a nasal SCC in the F344
33 rat as a function of age and exposure to formaldehyde. The fit of the Conolly et al. (2003) model
34 to the tumor incidence data is shown in Figure 5-12, Chapter 5.
35
This document is a draft for review purposes only and does not constitute Agency policy.
D-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 D.3. FLUX BINS
2 The spatial distribution of formaldehyde over the nasal lining was characterized by
3 partitioning the nasal surface by formaldehyde flux to the tissue (rate of gas absorbed per unit
4 surface area of the nasal lining), resulting in 20 "flux bins" (see Figure 5-13, Chapter 5). Each
5 bin is comprised of elements (not necessarily contiguous) of the nasal surface that receive a
6 particular interval of formaldehyde flux per ppm of exposure concentration (Kimbell et al.,
7 200la). The spatial coordinates of elements comprising a particular flux bin are fixed for all
8 exposure concentrations, with formaldehyde flux in a bin scaling linearly with exposure
9 concentration (ppm). The number of cells at risk varies across the bins, as shown in Figure 5-14,
10 Chapters.
11
12 D.4. USE OF LABELING DAT A
13 Cell replication rates in Conolly et al. (2003) were obtained by pooling labeling data
14 from two phases of a labeling study in which male F344 rats were exposed to formaldehyde gas
15 at similar concentrations (0, 0.7, 2.0, 6.0, 10.0, or 15.0 ppm). The first phase employed injection
16 labeling with a 2-hour pulse labeling time, and animals were exposed to formaldehyde for early
17 exposure periods of 1, 4, and 9 days and 6 weeks (Monticello et al., 1991). The second phase
18 used osmotic minipumps for labeling with a 120-hour labeling time to quantify labeling in
19 animals exposed for 13, 26, 52, and 78 weeks (Monticello et al., 1996). The combined pulse and
20 continuous labeling data were expressed as one exposure TWA over all sites for each exposure
21 concentration. (XN was calculated from these labeling data by using an approximation from
22 Moolgavkar and Luebeck (1992). A dose-response curve for normal cell replication rates (i.e.,
23 O,N as a function of formaldehyde flux) was then calculated as shown in Figure D-l. These steps
24 are carefully detailed and evaluated in Subramaniam et al. (2008), and discussion of the data will
25 continue in Appendix E in the section on uncertainties in characterizing cell replication rates.
26
27 D.5. UPWARD EXTRAPOLATION OF NORMAL CELL DIVISION RATE
28 The extensive labeling data collected by Monticello et al. (1996, 1991) present an
29 opportunity to use precursor data in assessing cancer risk. However, these empirical data could
30 be used to determine o,N(flux) only for the lower flux range, 0-9,340 pmol/mm2-hour (see
31 Subramaniam et al. [2008] for the reasons), as shown by the solid line in Figure D-l, whereas the
32 highest computed flux at 15.0 ppm exposure was 39,300 pmol/mm2-hour. Therefore Conolly et
33 al. (2003) introduced an adjustable parameter, amax, that represented the value of afflux) at the
34 maximum flux of 39,300 pmol/mm2-hour. amax was estimated by maximizing the likelihood of
This document is a draft for review purposes only and does not constitute Agency policy.
D-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 the two-stage model fit to the tumor incidence data. For 9,340 < flux <
2 39,300 pmol/mm2-hour, afflux) was determined by linear interpolation from a,N(9,340) to an
3 as shown by the dashed line in Figure D-l.
0.05 -
^ 0.04 -
s :
-2 0.03 -
03
01
•§ 0.02 -
Q
ID 0.01 -
O ;
n nn -
0.001
0.000
0.002
0.001
0.000
c
J-shape y
Hockey
•
^
c . . . .,
Estimated aN
Estimated aj u
X
X
) 2000 4000 6000 '
s
^^
/'
S
# 1
—
10000
20000
30000
40000
Flux (pmole/mrrr/h)
Figure D-l. Dose response of normal ((XN) and initiated (oii) cell division rate
in Conolly et al. (2003).
Note: Empirically derived values of aN (TWA over six sites) from Table 1 in
Conolly et al. (2003) and optimized parameter values from their Table 4 were
used. The main panel is for the J-shape dose response. Insets show J-shape and
hockey-stick shape representations at the low end of the flux range. The long
arrow denotes the upper end of the flux range for which the empirical unit-length
labeling data are available for use in the clonal growth model. amax is the value of
aN at the maximum formaldehyde flux delivered at 15 ppm exposure and
estimated by optimizing against the tumor incidence data, ai < aN for flux greater
than the value indicated by the small vertical arrow. Conolly et al. (2004, 2003)
assumed PI = aN at all flux values.
Source: Subramaniam et al. (2008).
This document is a draft for review purposes only and does not constitute Agency policy.
D-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 D.6. INITIATED CELL DIVISION AND DEATH RATES
2 The pool of cells used for obtaining the LI data in Monticello et al. (1996, 1991) consists
3 of largely normal cells with perhaps increasing numbers of initiated cells at higher exposure
4 concentrations. Since the division rates of initiated cells in the nasal epithelium, either
5 background or formaldehyde exposed, could not be inferred from the available empirical data,
6 Conolly et al. (2003) made what they perceived to be a biologically reasonable assumption for
7 0,1, assuming 0,1 to be linked to aN via a two-parameter function:
8
9 ai = (XN x (multb - multc x maxfoiN - o,N(basai), 0]} (D-2)
10
1 1 where OCN = afflux), a,N(basai) is the estimated average cell division rate in unexposed normal
12 cells, and multb and multc are unknown parameters estimated by likelihood optimization against
13 the tumor data.2 The value of o,N(basai) was equal to 3.39 x 10^ hours"1 as determined by Conolly
14 et al. (2003) from the raw averaged unit length labeling index data. The ratio OCI/OCN is plotted
15 against flux in Figure D-2, and ai(flux) is shown by the dotted line in Figure D-l.
16 Death rates of Initiated cells (Pi) are assumed to equal the division rates of Normal cells
17 (ON) for all formaldehyde flux values, that is
18
19 Pi(flux) = aN(flux) (D-3)
20
21 Conolly et al. (2003) stated that this formulation for oti and PI provided the best fit of the
22 model to the tumor data.
23
24 D.7. STRUCTURE OF THE CUT HUMAN MODEL
25 Subsequent to the BBDR model for modeling rat cancer, Conolly et al. (2004) developed
26 a corresponding model for humans for the purpose of extrapolating the risk estimated by the rat
27 model to humans. Also, rather than considering only nasal tumors (as in the rat model), the
28 human model was used to predict the risk of all human respiratory tumors. The human model for
29 formaldehyde carcinogenicity (Conolly et al., 2004) is conceptually very similar to the rat model
30 and follows the schematic in Figure 5-11, Chapter 5. The following points need to be noted:
2 multb and multc were equal to 1.072 and 2.583, respectively (J-shaped ON),, and 1.070 and 2.515, respectively
(hockey-stick shaped aN)
This document is a draft for review purposes only and does not constitute Agency policy.
D-7 DRAFT—DO NOT CITE OR QUOTE
-------
1.08 r
1.04 -
1.00
0.96 -
10000
20000
30000
4000C
Flux (pmole/mrrr/h)
Figure D-2. Flux dependence of ratio of initiated and normal cell replication
rates (OCI/(XN) in CUT model.
Note: Cell replication rate of initiated cells is less than normal cell replication rate
at flux exceeding the value denoted by the arrow. By assumption, the Y-axis also
represents
Source: Subramaniam et al. (2008).
1 • The model does not incorporate any data on human responses to formaldehyde exposure.
2 • The model is based on an anatomically realistic representation of the human nasal
3 passages in a single individual and an idealized representation of the LRT. Local
4 formaldehyde flux to the tissue is estimated by a CFD model for humans (Subramaniam
5 et al., 1998; Kimbell et al., 2001a; Overton et al., 2001).
6 • Rates of cell division and cell death are, with a minor modification, assumed to be the
7 same in humans as in rats.
8 • The concentration of formaldehyde-induced DPXs in humans is estimated by scaling up
9 from values obtained from experiments in the F344 rat and rhesus monkey. This scaling
10 up was discussed in chapter 3.
11 • The statistical parameters for the human model are either estimated by fitting the model
12 to the human background data, assumed to have the same value as obtained in the rat
13 model, or, in one case, fixed at a value suggested by the epidemiologic literature. The
14 delay, D, is fixed at 3.5 years, based on a fit to the incidence of lung cancer in a cohort of
15 British doctors (Doll and Peto, 1978). The two other parameters in the rat model that
16 affect the background rate of cancer (multb and |ibasai) are estimated by fitting to U.S.
17 cancer incidence or mortality data. These parameters affect the baseline values for the
This document is a draft for review purposes only and does not constitute Agency policy.
D-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 human oti, UN, and jii. Since amax, multfc, and KMU do not affect the background cancer
2 rate, they cannot be estimated from the (baseline) U.S. cancer incidence rates. Therefore,
3 in Conolly et al. (2004, 2003), amax and multfc are assumed to have the same values in
4 humans as in rats, and the human value for KMU is obtained by assuming that the ratio
5 KMU/jibasai is invariant across species. Thus,
KMU(human] = KMU(rat] x -—(D-4)
f^Nbasal(rat)
This document is a draft for review purposes only and does not constitute Agency policy.
D-9 DRAFT—DO NOT CITE OR QUOTE
-------
This page intentionally left blank.
This document is a draft for review purposes only and does not constitute Agency policy.
D-10 DRAFT—DO NOT CITE OR QUOTE
-------
Appendix E
-------
1 APPENDIX E
2
3 EVALUATION OF BBDR MODELING OF NASAL CANCER IN THE F344 RAT:
4 CONOLLY ET AL. (2003) AND ALTERNATIVE IMPLEMENTATIONS
5
6
7 A biologically based dose-response model for formaldehyde-induced cancer was
8 developed in a series of papers and in a health assessment report (CUT model) (Conolly et al.,
9 2004, 2003, 2000; Conolly, 2002; Kimbell et al., 2001a, b; Overton et al., 2001; CUT, 1999).
10 The model structure, notations, and calibration have been described in Appendix D. In
11 Chapter 5, an evaluation of the uncertainties of this model and alternative approaches based on
12 its conceptual framework was presented in a summary form. This Appendix provides the
13 relevant details of that evaluation and presents a range of dose-response curves for tumor risk in
14 the rat. It is divided into the following major sections. First, an overview of all the issues that
15 were evaluated is provided in tabular form. The rest of the Appendix then presents only those
16 issues which have a significant impact on model predictions: the use of history controls, the
17 uncertainty and variability in the dose-response for normal cell-replication rates, and sensitivity
18 of model results to uncertainty in the kinetics of initiated cells. These issues have significant
19 impact on inferences regarding mode of action, and this is discussed in some detail in this
20 Appendix. Assumptions and uncertainties related to the human formaldehyde model are
21 discussed in Appendix F.
22
23 E.I. TABULATION OF ALL IS SUES EVALUATED IN THE RAT MODELS
24 Table E-l summarizes model uncertainties and their impact as evaluated by EPA. The
25 key uncertainties are discussed in considerably more detail in additional sections in this
26 Appendix and in published manuscripts (Klein et al., 2010; Crump et al., 2008; Subramaniam et
27 al., 2008, 2007). The results in Subramaniam et al. (2007) and Crump et al. (2008) have been
28 debated further in the literature (Conolly et al., 2009; Crump et al., 2009). Other alternatives to
29 the CUT biological modeling (but based on that original model) are also further explored and
30 evaluated in this appendix.
31
This document is a draft for review purposes only and does not constitute Agency policy.
E-2 DRAFT—DO NOT CITE OR QUOTE
-------
5 3 a
Table E-l. Evaluation of assumptions and uncertainties in the CUT model for nasal tumors in the F344 rat
TO
a
§
Assumptions, approach, and
characterization of input
data in model
Rationale for assumption/
approach
EPA evaluation
Further
elaboration of
evaluation"
<3
Hoogenveen et al. (1999)
solution method, which is valid
only for time-independent
parameters, is accurate enough.
Errors due to this
assumption thought to be
significant only at high
concentration and not at
human exposures.
EPA implemented a solution method valid for time-dependent
parameters. Results did not differ significantly from those obtained
assuming Hoogenveen et al.(1999) solutions. However, impact was
not evaluated for the case where cell replication rates vary in time.
Crump et al.
(2005);
Subramaniam et
al. (2007)
All observed SCC tumors are
rapidly fatal; none are
incidental rumors.
Death is expected to occur
typically within 1-2 weeks
of observed tumor
(personal communication
with R. Conolly).
1) Overall, assumption does not impact model calibration or
prediction.
2) However, since 57 animals were observed to have tumors at
interim sacrifice times, EPA implementation distinguished
between incidental and fatal tumors. Time lag between observable
tumor and time of death was significant compared to time lag
between first malignant cell and observable tumor.
Subramaniam et
al. (2007)
W
oo
Historical controls from entire
NTP database were lumped
with concurrent controls in
studies.
Large number of control
animals (7,684).
Intercurrent mortality was
not expected to be
substantial.
1) Tumor incidence in "all NTP" 10-fold higher than in "all
inhalation NTP" controls. Including all NTP controls is
considered inappropriate.
2) Low-dose response curve is very sensitive to use of historical
controls.
3) Model inference regarding relevance of formaldehyde's
mutagenic potential to its carcinogenicity varies from
"insignificant" to "highly significant," depending on controls
used. (See Appendix F for impact on human risk.)
Table E-2;
Subramaniam et
al. (2007); Sec
E.3.1
LI was derived from
experimentally measured
ULLI.
Derived from correlating
ULLI to LI measured in
same experiment.
Significant variation in number of cells per unit length of basement
membrane. Spread in ULLI/LI -25%. Impact on risk not evaluated.
Subramaniam et
al. (2008);
fe
H
O
O
-------
5 3 a
Table E-l. Evaluation of assumptions and uncertainties in the CUT model for nasal tumors in the F344 rat (continued)
TO
a
§
<3
Assumptions, approach, and
characterization of input
data in model
Rationale for
assumption/approach
EPA evaluation
Further
elaboration of
evaluation8
Pulse and continuous labeling
data were combined in
deriving aN from LI.
All continuous LI values
were normalized by mean
ratio of pulse to continuous
LI for controls.
Formula used for deriving aN from LI is not applicable for pulse
labeling data. Pulse labeling is measure of number of cells in
S-phase, not of their recruitment rate into S-phase; not enough
information to derive aN from pulse data. Impact on risk predictions
could not be evaluated.
Subramaniam et
al. (2008); Section
E.3.2.2
To construct dose response for
ON, labeling data were
weighted by exposure time (t)
and averaged over all nasal
sites (TWA). At an exposure
concentration, flux was
averaged over all nasal sites.
Site-to-site variation in LI
was large and did not vary
consistently with flux. No
reasonable approach was
available for extrapolating
observed time variation in
labeling in rats to humans.
1) TWA assigns low weight to early time LI values, but aN for early
time (t) is very important to the cancer process. Since pulse ULLI
was used for t < 13 weeks, impact of these ULLIs on risk could
not be evaluated.
2) Time dependence in aN derived from continuous ULLI does not
significantly impact model predictions.
3) Site-to-site variation of aN is at least 10-fold and has major impact
on model calibration. Variation in tumor incidence data across
sites is 10-fold.
4) Large differences in number of cells across nasal sites (see Table
E-3), so averaging over sites is problematic.
5) TWA is also problematic because histologic changes, thickening
of epithelium and metaplasia occur at later times for the higher
dose and would affect replication rate.
Figures E-l, E-2,
E-3;
Subramaniam et
al. (2008); Section
E.3.2.3
W
fe
H
O
O
Steady-state flux estimates are
not affected by airway and
tissue reconfiguration due to
long-term dosing.
Histopathologic changes
not likely to be rate-
limiting factors in
dosimetry.
1) Thickening of epithelium and squamous metaplasia occurring at
later times for the higher dose (Kimbell et al. 1997b) will reduce
tissue flux. Not incorporated in model.
2) These effects will push regions of higher flux to more posterior
regions of respiratory tract. Likely to affect calibration of rat
model. Uncertainty not evaluated quantitatively.
3) Calibration of PBPK model for DPXs was seen to be highly
sensitive to tissue thickness.
Subramaniam et
al. (2008);
Cohen-Hubal et
al. (1997); Klein
etal. (2010).
-------
5 3 a
Table E-l. Evaluation of assumptions and uncertainties in the CUT model for nasal tumors in the F344 rat (continued)
TO
a
I
Assumptions, approach, and
characterization of input
data in model
Rationale for
assumption/approach
EPA evaluation
Further
elaboration of
evaluation"
TWA aN(flux) rises above
baseline levels only at
cytolethal dose. Above such
dose, aN(flux) rises sharply due
to regenerative proliferation.
Variability in aN(flux) is
partly represented by also
considering hockey-stick
(threshold in dose) when
TWA indicates J-shape
(inhibition of cell division)
description of aN(flux).
1) Uncertainty and variability in aN were quantitatively evaluated to
be large. In addition, there are several qualitative uncertainties in
characterization of aN(flux) from LI.
2) Several dose-response shapes, including a monotonic increasing
curve without a threshold, were considered in order to adequately
describe highly dispersed cell replication data. This has
substantial impact on low dose risk.
Figures E-l, E-2,
E-3, E-4, E-5;
Subramaniam et
al. (2008); Section
E.3.2
Dose response for a: was
obtained from aN, assuming
ratio (a: /aN) to be a two-
parameter function of flux (see
Figures 5-7, 5-9). Parameters
were estimated by optimizing
model predictions against
tumor incidence data.
(a: /aN) was > 1.0 in line
with the notion of I cells
possessing a growth
advantage overN cells.
Satisfies Occam's razor
principle (Conolly et al.,
2009).
1) (Xi /ON in CUT modeling is < 1.0 (growth disadvantage) for higher
flux values and is > 1.0 only at lower end of flux range in model
(see Figure 5-9).
2) Since there are no data to inform a:, sensitivity of risk estimates to
various functional forms was evaluated. Risk estimates for the rat
were extremely sensitive to alternate biologically plausible
assumptions for a:(flux) and varied by many orders of magnitude
at <1 ppm, including values lower than baseline risk. All these
models described tumor incidence data and cell replication and
DPX data equally well.
Figures D-2, E-5,
E-6;
Subramaniam et
al. (2008); Crump
et al. (2009,
2008); Section
E.3.3
W
Death rate of I cells is
assumed equal to division rate
of N cells
i.e. Pi(flux) = aN(flux).
Based on homeostasis (aN
= PN) and assumption that
formaldehyde is equally
cytotoxic to N cells and I
cells. Satisfies Occam's
Razor principle (Conolly et
al., 2009).
1) In general, data indicate I cells are more resistant to cytolethality
and that ADH3 clearance capacity is greater in transformed cells.
Therefore, plausibility of model assumption, that p: = aN, is
tenuous.
2) Alternate assumption, p: proportional to a:, was examined. Risk
estimates were extremely sensitive to assumptions on p: (see
Figure 5-12).
Subramaniam et
al. (2008); Crump
et al. (2009,
2008); Section
E.3.3
fe
H
O
O
-------
5 3 a
Table E-l. Evaluation of assumptions and uncertainties in the CUT model for nasal tumors in the F344 rat (continued)
TO
a
§
Assumptions, approach, and
characterization of input
data in model
Rationale for
assumption/approach
EPA evaluation
Further
elaboration of
evaluation"
<3
DPX is dose surrogate for
formaldehyde's mutagenic
potential. DPX clearance is
rapid and complete in 18
hours.
Casanova et al. (1994).
Half-life for DPX clearance in in vitro experiments on transformed
cell lines was 7-times longer than estimated by Conolly et al. (2004,
2003) and perhaps 14-times longer with normal (nontransformed)
human cells. Some DPX accumulation is therefore likely. However,
model calibration and dose response in rat was insensitive to this
uncertainty. See section E.3 for effect on scale-up of model to
humans.
Quievryn and
Zhitkovich,
(2000);
Subramaniam et
al. (2007); Section
3.6.6.3
Formaldehyde's mutagenic
action takes place only while
DPX's are in place.
DNA lesions may remain after DPX repair and incomplete repair of
DPX can lead to mutations (Barker et al. 2005). There is some
potential for formaldehyde-induced mutation after DPX clearance.
Thus, it is possible that formaldehyde mutagenicity may be
underrepresented in model. Could not quantitatively evaluate
uncertainty (no data on clearance of secondary lesions).
Subramaniam et
al. (2008); Section
4.3.3.3
W
fe
H
"References stated here are in addition to Conolly et al. (2004, 2003).
Note: Risk estimates discussed in this table are for the F344 rat.
O
O
-------
1 E.2. STATISTICAL METHODS USED IN EVALUATION
2 Parameters of the alternate models shown here were estimated by maximizing the
3 likelihood function defined by the data (Cox and Hinkley, 1974). Such estimates are referred to
4 as maximum likelihood estimates (MLEs). Statistical confidence bounds were computed by
5 using the profile likelihood method (Crump, 2002; Cox and Oakes, 1984; Cox and Hinkley,
6 1974). In this approach, an asymptotic 100(1 - a)% upper (lower) statistical confidence bound
7 for a parameter, /?, in the animal cancer model is calculated as the largest (smallest) value of ft
8 that satisfies
9
10 2[Lmax-L*(p)] = x1.2a (E-l)
11
12 where L indicates the likelihood of the rat bioassay data, Lmax is its maximum value, L*(ft) is, for
13 a fixed value of ft, the maximum value of the log-likelihood with respect to all of the remaining
14 parameters, and xi-2a is the 100(l-2a) percentage point of the chi-square distribution with one
15 degree of freedom. The required bound for a parameter, ft, was determined via a numerical
16 search for a value of ft that satisfies this equation.
17 The additional risk is defined as the probability of an animal dying from an SCC by the
18 age of 790 days, in the absence of other competing risks of death, while exposed throughout life
19 to a prescribed constant air concentration of formaldehyde, minus the corresponding probability
20 in an animal not exposed to formaldehyde. The MLE of additional risk is the additional risk
21 computed using MLEs of the model parameters.
22 The method described above for computing profile likelihood confidence bounds cannot
23 be used with additional risk because additional risk is not a parameter in the cancer model.
24 Instead, an asymptotic 100(1 - a)% upper (lower) statistical confidence bound for additional risk
25 was computed by finding the parameter values that presented the largest (smallest) value of
26 additional risk, subject to the inequality
27
28 2[Lmax-L]
-------
1
2 and []+ equals the quantity in the brackets whenever it is positive and zero otherwise. The
3 multiplicative weight, W, was selected by trial and error so that the final solution satisfied the
4 following equation sufficiently well.
5
6 2(Lmax-L)=x1-2a (E-4)
7
8 The computer code was written in Microsoft Excel® 2002 SP3 Visual Basic. Either the
9 regular Excel Solver or the Frontline Systems Premium Solver® was used to make the required
10 function optimizations. Computation of confidence bounds was highly computationally
11 intensive, and, consequently, confidence bounds were computed only for selected parameters in
12 selected runs. For select cases, the bootstrap method was also used to calculate confidence
13 bounds in order to confirm their accuracy. Values so calculated were found to be in agreement
14 with those calculated by using the likelihood method.
15
16 E.3. PRIMARY UNCERTAINTIES IN BBDR MODELING OF THE F344 RAT DATA
17 In their evaluation, Subramaniam et al. (2007) first attempted to reproduce the Conolly et
18 al. (2003) results under similar conditions and assumptions as employed in their paper, which
19 included the assumption that tumors were rapidly fatal. Figure 5-12 in Chapter 5 shows the
20 results for this case. The predicted probabilities shown in this figure were obtained by
21 Subramaniam et al. (2007) by using the source code made available by Dr. Conolly. These are
22 compared with the best-fitting model and plotted against the Kaplan-Meier (KM) probabilities.
23 Although the results are largely similar, there are some residual differences, and these are
24 detailed in Subramaniam et al. (2007).
25 Given the scope of issues to examine for the uncertainty analyses, the evaluation
26 proceeded in stages. First, the Hoogenveen et al. (1999) solution was replaced by one that is
27 valid for a model with time varying parameters (Crump et al., 2005; first entry in Table E-l), and
28 tumors found at scheduled sacrifices were assumed to be incidental rather than fatal (second
29 entry in Table E-l). Second, weekly averaged solutions for DPX concentration levels were used
30 instead of hourly varying solutions (predicted by a PBPK model). The log-likelihood values and
31 tumor probabilities remained essentially unchanged. Upon quantitative evaluation, these factors,
32 although important from a methodological point of view, were not found to be major
33 determinants of either calibration or prediction of the model for the F344 rat data (Subramaniam
34 et al., 2007).
35 Following Georgieva et al. (2003), Subramaniam et al. (2007) used the DPX clearance
36 rate constant obtained from in vitro data instead of the assumption in Conolly et al. (2003) that
This document is a draft for review purposes only and does not constitute Agency policy.
E-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 all DPXs cleared within 18 hours (Subramaniam et al., 2007). With this revision, weekly
2 average DPX concentrations were larger than those in Conolly et al. (2003) by essentially a
3 constant ratio equal to 4.21 (range of 4.12-4.36) when averaged over flux bin and exposure
4 concentrations. Accordingly, cancer model fits to the rat tumor incidence data using the two sets
5 of DPX concentrations (everything else remaining the same) provided very similar parameter
6 estimates, except that the parameter KMUrat in eq D-l (and eq D-4) (Appendix D) was 4.23
7 times larger with the Conolly et al. (2003) DPX concentrations. In other words, the product
8 KMU x DPX remained substantially unchanged. However, it is important to note that the
9 different clearance rate does significantly impact the scale-up of the two-stage clonal growth
10 model to the human since the parameter KMUhuman is not estimated separately but related to
11 KMUrat (see eq D-4).
12 After making the above modifications, the impact of the other uncertainties in Table E-l
13 were examined. Of the issues in Table E-l, the following uncertainties had large impacts on the
14 modeling of the F344 rat data, and will be discussed in considerably more detail:
15
16 1. use of historical controls,
17 2. uncertainty and variability in characterizing cell replication rates from the
18 labeling data, and
19 3. uncertainty in model specification of initiated cell kinetics.
20
21 E.3.1. Sensitivity to Use of Historical Controls
22 E.3.1.1. Use of Historical Controls
23 Conolly et al. (2003) combined the historical controls arising from the entire NTP
24 database of bioassays. Tumor and survival rates in control groups from different NTP studies
25 are known to vary due to genetic drift in animals over time and differences in laboratory
26 procedures, such as diet, housing, and pathological procedures (Haseman, 1995; Rao et al.,
27 1987). In order to minimize extra variability when historical control data are used, the current
28 NTP practice is to limit the historical control data, as far as possible, to studies involving the
29 same route of exposure and to use historical control data from the most recent studies (Peddada
30 et al., 2007).
31 Bickis and Krewski (1989) analyzed 49 NTP long-term rodent cancer bioassays and
32 found a large difference in determinations of carcinogenicity, depending on the use of historical
33 controls with concurrent control animals. The historical controls used in the CUT modeling
34 controls came from different rat colonies and from experiments conducted in different
35 laboratories over a wide span of years, so it is clearly problematic to assume that background
This document is a draft for review purposes only and does not constitute Agency policy.
E-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 rates in these historical control animals are the same as those in the concurrent control group.
2 There are considerable differences among the background tumor rates of SCCs in all NTP
3 controls (13/7,684 = 0.0017), NTP inhalation controls (1/4,551 = 0.0002), and concurrent
4 controls (0/341 = 0.0). The rate in all NTP controls is significantly higher than that in NTP
5 inhalation controls (p = 0.01, Fisher's exact test). Given these differences, the inclusion of any
6 type of historical controls is problematic and is thought to have limited value if these factors are
7 not controlled for (Haseman, 1995).
8
9 E.3.1.2. Influence of Historical Controls on Model Calibration and on Human Model
10 To investigate the effect of including historical controls in the CUT model, the analyses
11 in Subramaniam et al. (2007) were conducted by using the following sets of data for controls (the
12 fraction of animals with SCCs is denoted in parentheses):
13
14 a) only concurrent controls (0/341),
15 b) concurrent controls plus all the NTP historical control data used by Conolly et al. (2003)
16 (13/8,031),
17 c) concurrent controls plus data from historical controls obtained from NTP inhalation
18 studies (1/4,949) (NTP, 2005).3
19
20 The results of the evaluation are shown in Table E-2. For these analyses, the same normal
21 cell replication rates and the same relationship (see eq D-2 in Appendix D) between initiated cell
22 and normal cell replication rates as used in Conolly et al. (2003) were used. In all cases, weekly
23 averaged values of DPX concentrations were used. Model fits to the tumor incidence data were
24 similar in all cases to that shown in Figure 5-12 (see Subramaniam et al. [2007] for a more
25 complete discussion). The biggest influence of the control data was seen to be on the estimated
26 basal mutation rate in rats, ^Nbasal(mt)-, which, in turn, influences the estimated mutation effect in
27 humans through eq D-4 (Appendix D). amax was also seen to be a sensitive parameter and is
28 discussed later. See Subramaniam et al. (2007) for other parameters in the calibration.
3 Three animals in the inhalation historical controls were diagnosed with nasal SCC. Of these, two of the tumors
were determined to have originated in tissues other than the nasal cavity upon further review (Dr. Kevin Morgan and
Ms. Betsy Gross Bermudez, personal communication). These two tumors were therefore not included on the advice
of Dr. Morgan. See Subramaniam et al. (2007) for more details.
This document is a draft for review purposes only and does not constitute Agency policy.
E-10 DRAFT—DO NOT CITE OR QUOTE
-------
5 3 a
Table E-2. Influence of control data in modeling formaldehyde-induced cancer in the F344 rat
Case
Control animals (combined
with concurrent controls)
Cell replication dose
response
Log-likelihood
UNbasal
KMU
KMU/|lNbasal
C^max
A
All NTP
historical3
J-shaped
-1692.65
1.87x l(T6
1.12x l(T7
0.06
(0.0, 0.40)
0.045
(0.029, 0.045)
D
All NTP
historical3
Hockey stick
-1693.68
2.12 x 10~6
0.0
0.0
(0.0, 0.25)
0.045
(0.029, 0.045)
B
NTP
inhalation
historical3
J-shaped
-1,493.21
7.32 x 10~7
6.84 x 10~7
0.94
(0.26, 6.20)
0.045
(0.026, 0.045)
E
NTP
inhalation
historical3
Hockey stick
-1,493.35
9.32 x 10~7
6.18 x 10~7
0.66
(0.2, 5.20)
0.045
(0.027, 0.045)
C
Concurrent
only3
J-shaped
-1,474.29
0.0
1.20x 10~6
co
(0.42, oo)
0.045
(0.027, 0.045)
F
Concurrent
only3
Hockey stick
-1,474.29
0.0
1.20 x 1Q-6
GO
(0.41, oo)
0.045
(0.027, 0.045)
?Q >! ^
k ^ *S
TO § ^
a ,s. a
^
aValues in parentheses denote lower and upper 90% confidence bounds.
Source: Adapted from Subramaniam et al. (2007).
O
o
-------
1 The ratio KMU/|iNbasai is of particular interest because extrapolation to human in Conolly
2 et al. (2004) assumed its invariance as given by eq D-4 (Appendix D). Now, u.Nbasai in the human
3 is estimated independently by fitting a scaled-up version of the two-stage model to human
4 baseline rates of tumor incidence. Thus, a decrease in the value of UNbasai estimated in the rat
5 modeling increases the formaldehyde-induced mutational effect in the human.
6 The MLE of KMUrat/|iNbasai(rat) is zero in Conolly et al. (2003). However, in the various
7 cases examined in Subramaniam et al. (2007) it takes a range of values from 0 to 0.9 mm3/pmol
8 and undefined (or infinite, when UNbasai = 0). The 95% upper confidence bound on this ratio
9 ranges from 0.25-6.2 (these values would be four times larger had the Conolly et al. [2003] DPX
10 concentrations been used) to infinite. Thus, the extrapolation to human risk by using the
11 approach in Conolly et al. (2004) becomes particularly problematic when only concurrent
12 controls are used, because then the mutational contribution to formaldehyde-induced risk in
13 humans becomes unbounded. This issue will be discussed again toward the end of the
14 discussion on historical controls.
15 It may be noted, however, that absence of tumors in the limited number of concurrent
16 animals does not imply that the calculation will necessarily predict a zero background
17 probability of tumor (i.e., a parameter estimate of UNbasai = 0). Subramaniam et al. (2007)
18 observed such a counterexample estimate for UNbasai in simulations involving the alternate dose-
19 response curves for O,N and 0,1 that are discussed in Section E.3.4. Nonetheless, when UNbasai = 0,
20 an upper bound for UNbasai using the concurrent controls could be inferred. Accordingly, the 90%
21 statistical lower confidence bound on the ratio KMU/jiNbasai is also reported in Table E-2. Such a
22 value would of course provide a lower bound on risk by using this model and would therefore
23 not be conservative.
24 Conolly et al. (2003) estimated KMU to be zero for both their hockey-stick and J-shape
25 dose response models for cell replication. However, the estimate for the coefficient KMU
26 (obtained using the solution of Crump et al. [2005]) is zero only for the case of the model with
27 the hockey-stick curve for cell replication and with control data as used by Conolly et al. (2003).
28 It is positive in all other cases and statistically significantly so in all cases in which either NTP
29 inhalation control data or concurrent controls were used. With concurrent controls only and the
30 J-shape cell replication model, the MLE estimate for KMU (1.2 x 10~6) is larger than the
31 statistical upper bound obtained by Conolly et al. (2003) (8.2 x 10~7). It should also be kept in
32 mind that the estimate would be about 4.2 times larger still had the Conolly et al. (2003) DPX
33 model been used.
34
This document is a draft for review purposes only and does not constitute Agency policy.
E-12 DRAFT—DO NOT CITE OR QUOTE
-------
1 E.3.1.3. Influence of Historical Controls on Dose-Response Curve
2 Subramaniam et al. (2007) showed that inclusion of historical controls had a strong
3 impact on the tumor probability curve below the range of exposures over which tumors were
4 observed in the formaldehyde bioassays. As shown there, the MLE probabilities for occurrence
5 of a fatal tumor at exposure concentrations below 6 ppm were roughly an order of magnitude
6 higher when all the NTP historical controls were used, compared with MLE probabilities
7 predicted when historical controls were drawn only from inhalation bioassays, and many orders
8 of magnitude higher than MLE probabilities predicted when only concurrent controls were used
9 in the analysis. (Note that this comparison should not be inferred to apply to upper bound risk
10 estimates since there were many fewer concurrent than historical controls, so error bounds could
11 be much larger in the case where concurrent controls were used.)
12 However, as shown by these authors, model fits to the tumor data in the 6-15 ppm
13 exposure concentration range were qualitatively indifferent to which of these control data sets
14 was used. This observation emphasizes the statistical aspect of the CUT modeling—that
15 significant interplay among the various adjustable parameters allows the model to achieve a
16 good fit to the tumor incidence data independent of the control data used. On the other hand, the
17 results in Subramaniam et al. (2007) show that changes in the control data affect parameter
18 KMU, resulting in significantly different tumor predictions at lower exposure concentrations.
19 Therefore, the strong influence of using all the NTP historical controls on the low-dose region of
20 the time-to-tumor curves presented in Subramaniam et al. (2007) suggests that large
21 uncertainties may arise in extrapolating to both human and rat (in the low-dose region) from
22 such considerations alone.
23
24 E.3.1.4. Problem Including 1976 Study for Inhalation Historical Control
25 A crucial point needs to be noted with regard to the use of inhalation NTP historical
26 controls (i.e., cases B and E) in the two-stage clonal growth modeling. The single relevant tumor
27 in the NTP inhalation studies came from the very first NTP inhalation study, dated 1976, and the
28 animals in this study were from Hazelton Laboratories, whereas the concurrent animals were all
29 from Charles River Laboratories. Similar problems arise with inclusion of several other NTP
30 inhalation studies. As mentioned before, genetic and other time-related variation can lead to
31 different tumor and survival rates, and in general it is recommended that use of historical
32 controls be restricted to the same kind of bioassays and to studies within a 5-7 year span of the
33 concurrent animals (Peddada et al., 2007). Thus, it is problematic to assume that the tumor in
34 the 1976 NTP study is representative of the risk of SCCs in the formaldehyde bioassays. Even if
35 it were appropriate to consider the 1976 study, this leads to the unstable situation in which,
This document is a draft for review purposes only and does not constitute Agency policy.
E-13 DRAFT—DO NOT CITE OR QUOTE
-------
1 despite all of the "upstream" mechanistic information used to construct the BBDR model, the
2 only piece of data that might keep the model predictions of human risk bounded is a single tumor
3 found among several thousand rats from NTP bioassays (Crump et al., 2008). In summary,
4 although it can be argued that the rate of SCCs among the controls in the rat bioassay is probably
5 not zero, it is also problematic to assume that this rate can be adequately represented by the
6 background rate in NTP historical controls or even in NTP inhalation historical controls.
7
8 E.3.1.5. Effect of Control Data on MOA Inferences
9 Subramaniam et al. (2007) also examined the contribution of the DPX component (which
10 represents the directly mutagenic potential of formaldehyde in the model) to the calculated tumor
11 probability, choosing for their case study the optimized models that use the NTP inhalation
12 control data. In the range of exposures where tumors were observed (6.0-15.0 ppm), the DPX
13 term was found to be responsible for 58-74% of the added tumor probability. Below 6.0 ppm
14 the estimated DPX contribution was extremely sensitive to whether the hockey-stick shape or
15 J-shape was used to characterize the dose response for cell replication, and varied between
16 2% and 80%.
17 The CUT BBDR cancer modeling has contributed to the weight-of-evidence process in
18 various formaldehyde risk assessment efforts and papers by lending weight to the argument that
19 the direct mutations induced by formaldehyde are relatively irrelevant compared to the
20 importance of cytotoxicity-induced cell proliferation in explaining the observed tumorigenicity
21 in rodent bioassays and in projecting those observations to human exposures (Conolly et al.,
22 2004, 2003; Slikker et al., 2004; Bogdanffy et al., 2001, 1999; Conolly, 1995). The reanalyses in
23 Subramaniam et al. (2007) (in particular, the results in the above paragraph) indicate that, if the
24 CUT mathematical modeling were utilized to inform this debate, it would in fact indicate the
25 contrary—that a large contribution from formaldehyde's mutagenic potential may be needed to
26 explain formaldehyde carcinogenicity. This discussion is resumed in the context of uncertainties
27 in model specification for initiated cells.
28
29 E.3.2. Characterization of Uncertainty-Variability in Cell Replication Rates
30 E.3.2.1. Dose-Response for UN as Used in the CUT Clonal Growth Modeling
31 Monticello et al. (1996, 1991) used unit length labeling index (ULLI) to quantify cell
32 replication within the respiratory epithelium. ULLI is a ratio between a count of labeled cells
33 and the corresponding length (in millimeters) of basal membrane examined, whereas the per-cell
34 labeling index (LI) is the ratio of labeled cells to all epithelial cells, in this case, along some
35 length of basal membrane and its associated layer of epithelial cells. Monticello et al. (1996,
This document is a draft for review purposes only and does not constitute Agency policy.
E-14 DRAFT—DO NOT CITE OR QUOTE
-------
1 1991) published ULLI values averaged over replicate animals for each combination of exposure
2 concentration, exposure time, and nasal site. These values are plotted in Figure E-l.
3 In order to utilize the ULLI data in clonal growth modeling, ULLI needed to be related to
4 LI, and thereby to cell replication rate ((XN) of normal cells. Conolly et al. (2003) adopted the
5 following procedure in using these values (Subramaniam et al., 2008):
6
7 1. The injection labeled ULLI data were first normalized by the ratio of the average
8 minipump ULLI for controls to the average injection labeled ULLI for controls.
9 2. Next, these ULLI average values were weighted by the exposure times in Monticello et
10 al. (1996, 1991) and averaged over the nasal sites. Thus, the data were combined into
11 one TWA for each exposure concentration.
12 3. LI was linearly related to the measured ULLI by using data from a different experiment
13 (Monticello et al., 1990) where both quantities had been measured for two sites in the
14 nose.
15 4. Cell replication rates of normal cells (
-------
Exposure < 12 wk (pulse labeling)
q-
8
q
in
in
ci
ALM
10
15
Exposure > 12 wk (continuous labeling)
q-
8
q
in
in
ci
ALM
10
15
8
O)
O
O
CO
CD
"OS
c
03
CD
cl-
in
ci-
in
ci-
in
AMS
5 10 15
MMT
5 10 15
PLM
q-
in
AMS
10 15
o-
in
MMT
10 15
o-
in
• — ~, ---- , ^i*"
PLM
5 10 15
10 15
o-
in
PMS
q-
ci-
V
"V
5 10 15 0 5 10
Formaldehyde exposure concentration, ppm
PMS
15
1
2
3
4
5
Figure E-l. ULLI data for pulse and continuous labeling studies.
Note: Data are from pulse labeling study, left-hand side, at 1-42 days of exposure
and from the continuous-labeling study, right-hand side, at 13-78 weeks of
exposure for five nasal sites ALM, AMS, MMT, PLM, and posterior mid septum
[PMS]). Within each graph, lines with more breaks correspond to shorter
exposure times. Data source: Monticello et al. (1996, 1991).
TTz/'s document /'s a draft for review purposes only and does not constitute Agency policy.
E-16 DRAFT—DO NOT CITE OR QUOTE
-------
1 E.3.2.2. Time Variability in Labeling Data
2 E.3.2.2.1. Short-time exposure effects on cell replication.
3 Figure E-l shows the site and time variation in the raw unit-length labeling index (ULLI)
4 data for 1 day to 78 weeks of exposure duration. The temporal variation in ULLI is quite
5 different between the "early time" (left panel) and "later time" (right panel) and these early-time
6 effects may be quite important to the cancer modeling. At the earliest times in the left panel, the
7 data show an increased trend in labeling at 2 ppm for the sites anterior lateral meatus (ALM),
8 anterior medial septum (AMS), posterior lateral meatus (PLM), and medial maxilloturbinate
9 (MMT) relative to control. Such an increase is generally indicated for low flux values also for
10 the 13-week exposure time. This can be seen in the dose-response plotted as a function of flux
11 in Figure E-4.
12 The early times would be important if, say, repeated episodic exposures were considered,
13 where adequate time has not elapsed for adaptive effects to take place. Such an exposure
14 scenario may be the norm in the human context. In the CUT cancer modeling, the LI was
15 weighted by exposure time. As a consequence, the contribution of the early-time labeling data is
16 minimized in their modeling.
17
18 E.3.2.2.2. Uncertainty due to combining pulse and continuous labeled data.
19 The formula used for obtaining OCN from LI in Conolly et al. (2003) was due to
20 Moolgavkar and Luebeck (1992) who derived this formula for continuous LI, cautioning that it
21 is not applicable for pulse labeled data. However, Conolly et al. (2003) applied this formula to
22 the injection (pulse) labeled data also. Such an application is problematic because 2-hour pulse
23 labeled data represent the pool of cells in S-phase rather than the rate at which cells are recruited
24 to the pool, and because the baseline values of O,N obtained in this manner from both data sets
25 differ considerably. As such, we are not aware of any reasonable manner to derive cell
26 replication rates from these pulse data without acquisition of data at additional time points.
27 Because of these problems in incorporating the pulse-labeled data, further quantitative analysis
28 of cell replication rates is restricted in this document to the continuous labeled data (Monticello
29 et al., 1996), which do not include measurements made before 13 weeks of exposure. It is
30 unfortunate that the continuous labeled data do not include any early measurements.
31
This document is a draft for review purposes only and does not constitute Agency policy.
E-17 DRAFT—DO NOT CITE OR QUOTE
-------
1 E.3.2.3. Site and Time Variability in Derived Cell Replication Rate
2 In the remainder of this section, the factors that are considered in order to represent the
3 uncertainty and variability in the cell replication data when developing alternate dose-response
4 curves for afflux) will be elaborated.
5 The ULLI data for individual animals were provided by CUT, which were transformed to
6 LI values using the linear relationship from step 3 in Section E.3.2.1. For these replicate data,
7 cell replication rates of normal cells ((XN) were then calculated as (XN = (-0.5/t)log(l - LI) as in
8 Step 4. Figure E-2 (adapted from Subramaniam et al., 2008) shows the variability in aN due to
9 replicated animals, exposure times, and nasal sites in the continuous labeled data obtained by
10 Monticello et al. (1996). In this figure, log aN versus site-specific flux are plotted for six sites
11 and four exposure times for four to six replicate animals in each case. (The mean ULLI over
12 these replicates were shown in Figure E-l for each site and time as a function of exposure
13 concentration.) It needs to be noted that these nasal sites differ considerably in the number of
14 cells estimated at these locations as shown in Table E-3. Each point in Figure E-2 represents
15 data from a single site for a single animal at a given time. For comparison, the afflux) in
16 Conolly et al. (2003) is also plotted in this figure at their averaged flux values (filled circles).
17 For flux >9,340 pmol/mm2-hour, Conolly et al. (2003) extrapolated this empirically derived
18 afflux) by using a scheme discussed in Appendix D (see Section D.5) on the upward
19 extrapolation of cell replication rate. The curves shown connecting the filled circles in the figure
20 represent their linear interpolation (long dashes) between the six points. Their linear
21 extrapolation for flux value >9,340 pmol/mm2-hour is also shown (short dashes). Note that the
22 linear interpolation and extrapolation are shown transformed to a logarithmic scale in this plot.
23 As discussed, the raw labeling data plotted in Figure E-l indicates considerable temporal
24 variability. In Figures E-3, fitted dose-response curves showing logio(«w) versus flux with
25 simultaneous confidence limits separately for each time point for two of the largest sites in
26 Table E-3 (ALM and PLM) are plotted for the continuous labeled data. Note that flux levels are
27 different at each site. Simple polynomial models in flux (as a continuous predictor), with time
28 included as a factor (i.e., a class or indicator variable, T; representing the effect of the ith time)
29 were used as follows:
30
31 log(aw) = a + b x flux + c x flux2 + d x flux3 + T; (E-5)
This document is a draft for review purposes only and does not constitute Agency policy.
E-l8 DRAFT—DO NOT CITE OR QUOTE
-------
A
A
V
o
ADS
ALM
AMS
MMT
PLM
PMS
•f
A
A
*
A
*
I
A
A
A
0
2000
4000
6000
8000
10000 12000
Figure E-2. Logarithm of normal cell replication rate OCN versus
formaldehyde flux (in units of pmol/mm2-hour) for the F344 rat nasal
epithelium.
Note: Values were derived from continuous unit length labeled data obtained by
Monticello et al. (1996) for four to six individual animals at all six nasal sites
(legend, sites as denoted in original paper) and four exposure durations (13, 26,
52, 78 weeks). Each point represents a measurement for one rat, at one nasal site,
and at a given exposure time. Filled red circles: afflux) used in Conolly et al.
(2003) plotted at their averaged flux values (see text for details). Long dashed
lines: their linear interpolation between points. Short dashed line: their linear
extrapolation for flux value >9,340 pmol/mm2-hour (see Figure D-l for full range
of extrapolation). Linear interpolation/extrapolation is shown with Y-axis
transformed to logarithmic scale.
Source: Subramaniam et al. (2008).
This document is a draft for review purposes only and does not constitute Agency policy.
E-19 DRAFT—DO NOT CITE OR QUOTE
-------
Table E-3. Variation in number of cells across nasal sites in the F344 rat
Nasal site
Anterior lateral meatus
Posterior lateral meatus
Anterior mid septum
Posterior mid septum
Anterior dorsal septum
Anterior medial maxilloturbinate
No. of cells
976,000
508,000
184,000
190,000
128,000
104,000
Note: Mean number of cells in each side of the nose of control animals.
Source: Monticello et al. (1996).
CM
CO
8 '
o
TO ^fr
O '
LO
13 weeks
•o
0 4000 8000 12000
Flux
CM
CO
8 '
o
TO -sj-
O '
26 weeks
o
.o
0 4000 8000 12000
Flux
CM
TO -
O
LO
52 weeks
0 4000 8000 12000
Flux
CM
8 °?
o
TO
o ^r
10
78 weeks
0 4000 8000 12000
Flux
Figure E-3A. Logarithm of normal cell replication rate versus formaldehyde
flux with simultaneous confidence limits for the ALM.
Source: Subramaniam et al. (2008).
This document is a draft for review purposes only and does not constitute Agency policy.
E-20 DRAFT—DO NOT CITE OR QUOTE
-------
13 weeks
26 weeks
Csl -I
00 .
LO .
CM .
8 CO .
o '
LO .
0 2000 6000
Flux
52 weeks
O)
o
LO .
i
co .
LO .
I
o
0 2000 6000
Flux
0 2000 6000
Flux
78 weeks
o
0 2000 6000
Flux
Figure E-3B. Logarithm of normal cell replication rate versus formaldehyde
flux with simultaneous confidence limits for the PLM.
Source: Subramaniam et al. (2008).
1 The variability considered is that among animals and any measurement error as well as
2 any other design-related components of error. Simultaneous 95% confidence limits for log(oiN)
3 were produced using Scheffe's method (Snedecor and Cochran, 1980). These 95% confidence
4 limits span a range of 0.96 in loglO(oiN), or nearly a 10-fold range in median ON. There is
5 additional dispersion in these data that does not appear in Figures E-2 and E-3 for (XN, derived
6 using the mean value of ULLI/LI; due to variation in the number of cells per mm basement
7 membrane, the ratio of ULLI/LI had a spread of approximately ±25% (0.45 to 0.71, mean 0.60)
8 among the eight observations considered in Monticello et al. (1990). Thus:
9
This document is a draft for review purposes only and does not constitute Agency policy.
E-21 DRAFT—DO NOT CITE OR QUOTE
-------
1 1. As suggested by Table E-3, and Figures E-2 and E-3, the shape of afflux) in Conolly et
2 al. (2003) is therefore likely to be very sensitive to how (XN is weighted and averaged over
3 site and time.
4 2. Averaging of sites could significantly affect model calibration because of substantial
5 nonlinearity in model dependence on O,N at the 10 and 15 ppm doses associated with high
6 cancer incidence.
7 3. Monticello et al. (1996) found a high correlation between tumor rate and the ULLI
8 weighted by the number of cells at a site. Therefore, considering these factors while
9 regressing O,N against tissue dose would be important in the context of site differences in
10 tumor response.
11 4. A further complexity arises because of histologic changes and thickening that occurs in
12 the nasal epithelium over time in the higher dose groups (Morgan, 1997), factors that are
13 likely to affect estimates of local formaldehyde flux, uptake, and replication rates
14 (Subramaniam et al., 2008).
15
16 It is clear from Figures E-l and E-3 that the time dependence in cell replication is
17 significant. It would also be useful to examine if this time dependence affects the results of the
18 time-to-tumor modeling and if early temporal changes in replication rate are important to
19 consider because of the generally cumulative nature of cancer risk. The time window over
20 which formaldehyde-induced cancer risk is most influenced is not known, but the time weighting
21 used by Conolly et al. (2003) assigns a relatively low weight to labeling observed at early times
22 compared with those observed at later time points. Finally, initiated cells are likely to be
23 replicating at higher rates than normal cells as evidenced in several studies on premalignant
24 lesions (Coste et al., 1996; Dragan et al., 1995; Rotstein et al., 1986). Therefore, LI data as an
25 estimator of normal cell replication rate would be most reliable at early times when the mix of
26 cells sampled include fewer preneoplastic or neoplastic cells.
27 The more relevant question, therefore, is whether the afflux) derived in Conolly et al.
28 (2003) by a TWA over all sites has an effect on low-dose risk estimates. Given the above
29 uncertainties and variability not characterized in CUT (1999) or in Conolly et al. (2003), it is
30 important to examine whether additional dose-response curves that fit the cell replication data
31 reasonably well have an impact on estimated risk. Such sensitivity analyses are carried out in
32 the sections that follow.
33
34 E.3.2.4. Alternate Dose-Response Curves for Cell Replication
35 Clearly, a large number of alternative afflux) can be developed. In conjunction with the
36 other uncertainties, mainly the use of control data and alternative model structures for initiated
37 cell kinetics, the number of plausible clonal growth models to be exercised soon require a
This document is a draft for review purposes only and does not constitute Agency policy.
E-22 DRAFT—DO NOT CITE OR QUOTE
-------
1 prohibitively large investment of time. Therefore, detailed analyses were restricted to a select set
2 of biologically plausible choices of curves for afflux), which would allow the identification of
3 a range of plausible risk estimates (MLEs and statistical bounds). This discussion is further
4 informed by recently published dose response data for cell replication (Meng et al., 2010),
5 detailed in section F.2.3.
6 Six alternative equations for (XN were developed by regression analysis of the Monticello
7 et al. (1996) ULLI data. The replicate data corresponding to the summary data presented in this
8 paper were kindly provided to EPA by CUT for further analyses. In each of these equations, ON
9 is expressed as a function of formaldehyde flux to nasal tissue (pmol/mm2-hour) and, in one
10 equation (see eq E-l 1) that explored time-dependence, the duration of exposure to formaldehyde
11 in weeks. All the graphs use flux/10,000 for the X-axis, and the Y-axis expresses logio OCN.
12 One source of uncertainty in the cell replication dose response in Conolly et al. (2003) is
13 the large value of amax (the cell replication rate corresponding to the upper end of the flux range
14 at 15 ppm exposure) in the upward extrapolation from the empirically-determined afflux) (see
15 Figure D-l and surrounding text in Section D.5). The optimal value of amax was found by
16 Conolly et al. (2003) to be 0.0435 hour"1. As noted by the authors, an argument in support of
17 this value is that it corresponds to the inverse of the fastest cell cycle times found in the
18 literature. Since the model treats the induced replication rates as being time invariant, this means
19 that cells in the high-flux region(s) divide at the highest cell turnover rate ever observed
20 throughout most of an animal's life. This does not seem to be biologically plausible
21 (Subramaniam et al., 2008).
22 Our analysis found that a 20% increase or decrease in the estimated value for amax
23 degraded the fit to the tumor incidence data considerably. Because of the interplay between the
24 parameters estimated by optimization, this sensitivity of the model to amax indicates that it is
25 necessary to examine if other plausible values of amax are also indicated by the data and to what
26 extent low dose estimates of risk are influenced by the uncertainty in its value. The need for
27 such an analysis is also indicated by Figure E-2. The value of amax (logioamax = -1.37) in
28 Conolly et al. (2003) is roughly an order of magnitude greater than the values of afflux) at the
29 highest flux levels in this figure. If the data pooled over all sites and times are to be used for
30 afflux), then, based solely on the trend in afflux) in Figure E-2, it appears unlikely that
31 afflux) could increase up to this value of amax. Visually, these empirically derived data
32 collectively suggest that OCN versus flux could be leveling off rather than increasing 10-fold.
33 Therefore, as an alternative to the approach taken in Conolly et al. (2003) of estimating amax via
34 likelihood optimization against the tumor data, regressions of the empirical cell replication data
This document is a draft for review purposes only and does not constitute Agency policy.
E-23 DRAFT—DO NOT CITE OR QUOTE
-------
1 in Figure E-2 were used to extrapolate afflux) outside the range of observation (recognizing the
2 uncertainty and model dependence that still results from extrapolating well outside the range of
3 observed data).
4 In fitting dose-response curves to the cell replication data, a functional form was used
5 that was flexible to allow a variety of monotonic and nonmonotonic shapes, with a parameter
6 that determined the asymptotic behavior of the dose-response function. This allowed the
7 extrapolation of afflux) to higher flux levels by only relying on the empirical cell replication
8 data. Then, there is no need for an adjustable parameter to be estimated by fitting to the tumor
9 data. However, the plausible asymptotes obtained in this manner spanned a large range. In one
10 case below, the asymptote suggested by the fit to the empirical cell replication data was judged
11 to be abnormally high. In this case, the a^ versus flux curve was followed until the biological
12 maximum of amax (as given in Conolly et al. [2003]) was reached.
13 In three of the six regression models below, the data were restricted to the earliest
14 exposure time (13 weeks) in Monticello et al. (1996) for which the cell proliferation rate (a^)
15 could be calculated. The interest in using only the 13-week exposure time arises from
16 observations (Monticello et al., 1996, 1991) that at later times there were more frequent and
17 severe histologic changes, which may have altered formaldehyde uptake and cell proliferation
18 response. Consequently, given that the data in Monticello et al. (1991) for times earlier than
19 13 weeks could not be utilized as explained in Section E.3.2.3, the 13-week responses might
20 better represent proliferation rates for use in a two-stage model of the cancer process than the
21 rest of the Monticello et al.(1996) data.
22 Second, the LI data showed considerable variation among nasal sites, which may be
23 related to the variation in tumor response among sites. Since the cell replication dose-response
24 curves used in the cancer model represent all of the sites, it was attempted to include this
25 variation by weighting the regression by the relative cell populations at risk at each of the sites.
26 This was carried out for some of the models as stated below.
27 Finally, in one of the regression models, derived from fitting to all of the Monticello et al.
28 (1996) ULLI data, time-dependence of aN was considered by using weeks of exposure as a
29 covariate. In this model, time was a regression (continuous) predictor, not a class variable, and
30 its coefficient represents the change in logic (XN per week of exposure
31 The following regression models for UN versus flux, denoted in the equations below as
32 N1-N6 and shown in Figure E-4, as well as the hockey-stick and J-shaped curves used by
33 Conolly et al. (2003), shown in Figure D-l, Appendix D, were next used as inputs to the clonal
34 growth model for cancer:
This document is a draft for review purposes only and does not constitute Agency policy.
E-24 DRAFT—DO NOT CITE OR QUOTE
-------
Y= a + c * exp(- b1*X - sgn(b2)*[b2*X]A2 -...),
param= -2.01;-6.51; 0; 0.673
0.0
0.2
0.4
0.6
0.8
1.0
1
2
Figure E-4, Nl. Various dose-response modeling of normal cell replication
rate.
Note: See text for definitions of N1-N6. Nl: Quadratic; monotone increasing in
flux, derived from fit to all of the Monticello et al. (1996) ULLI data.
loglO(alpha) = -2.565 -0.987 * exp{+2.188*X -(2.162*X)"2 }
4
5
6
7
0.0
0.2
0.4 0.6
weighted mean flux/10,000
0.8
1.0
Figure E-4, N2. Various dose-response modeling of normal cell replication
rate.
Note: See text for definitions of N1-N6. N2: Linear-quadratic; decreasing in flux
for small values of flux, derived from fit to all of the Monticello et al. (1996)
ULLI data.
This document is a draft for review purposes only and does not constitute Agency policy.
E-25 DRAFT—DO NOT CITE OR QUOTE
-------
Time = 13 weeks
0.0
0.2 0.4 0.6
weighted mean flux/10,000
0.8
1.0
1
2
3
Figure E-4, N3. Various dose-response modeling of normal cell replication
rate.
Note: See text for definitions of N1-N6. N3: Linear-quadratic; decreasing in flux
for small values of flux, derived from fit to the 13-week Monticello et al. (1996)
ULLI data, using average flux over all sites for a given ppm exposure and
weighting regression by estimates of the numbers of cells at each of five sites.
Time = 13 weeks
O o
0.0
0.2
0.4
0.6
flux/10,000
0.8
1.0
1.2
Figure E-4, N4. Various dose-response modeling of normal cell replication rate.
Note: See text for definitions of N1-N6. N4: Quadratic; monotone increasing in
flux, derived from unweighted fit to 13-week Monticello et al. (1996) ULLI data.
This document is a draft for review purposes only and does not constitute Agency policy.
E-26 DRAFT—DO NOT CITE OR QUOTE
-------
0.0
0.2
0.4 0.6
weighted mean flux/10,000
0.8
1.0
1
2
Figure E-4, N5. Various dose-response modeling of normal cell replication
rate.
Note: See text for definitions of N1-N6. N5: Linear-quadratic-cubic; initially
increasing slightly with increasing flux, then decreasing slightly, and finally
increasing, derived from fit to 13-week Monticello et al. (1996) ULLI data, using
average flux over all sites for a given ppm exposure and weighting regression by
estimates of the numbers of cells at each of five sites.
This document is a draft for review purposes only and does not constitute Agency policy.
E-27 DRAFT—DO NOT CITE OR QUOTE
-------
All Sites, ~ Time + 2nd order in Flux
Time = 13
Time = 26
T -
0.0 0.2 0.4 0.6 0.8 1.0 1.2
x
T -
CD .
00 .
o o
0.0 0.2 0.4 0.6 0.8 1.0 1.2
x
Time = 52
Time = 78
T -
CD .
00 -
0.0 0.2 0.4 0.6 0.8 1.0 1.2
x
T -
CD .
00 -
O Or
0.0 0.2 0.4 0.6 0.8 1.0 1.2
x
1
2
Figure E-4, N6. Various dose-response modeling of normal cell replication
rate.
Note: See text for definitions of N1-N6. N6: Linear-quadratic-cubic; initially
increasing slightly with increasing flux, then decreasing slightly, and finally
increasing, derived from fit to all Monticello et al. (1996) ULLI data, using weeks
of exposure as a covariate. In this model, time was a regression (continuous)
predictor, not a class variable, and its coefficient represents the decrease in logic
O,N per week of exposure time.
4
5
6
1
Nl: Quadratic; monotone increasing in flux, derived from fit to all of the Monticello et al. (1996)
ULLI data.
aN = Exp{-2.015- 6.513 x Exp[- (6.735* J(T4 x flux)2]}
(E-6)
This document is a draft for review purposes only and does not constitute Agency policy.
E-28 DRAFT—DO NOT CITE OR QUOTE
-------
1 N2: Linear-quadratic; decreasing in flux for small values of flux, derived from fit to all of the
2 Monticello et al. (1996) ULLI data.
3 aN = Exp{-5.906-2.272 * Exp[2.188*10~4 * flux-(2.162* 10'4 *flux)2]} (E-7)
4
5 N3: Linear-quadratic; decreasing in flux for small values of flux, derived from fit to the 13-week
6 Monticello et al. (1996) ULLI data, using average flux over all sites for a given ppm exposure
7 and weighting regression by estimates of the numbers of cells at each of five sites.
8
9 . aN = Exp{-5.274 - 2.792 * Exp[1.407* J(T4 * flux - (1.986* J(T4 x flux)2]} (E-8)
10
11 N4: Quadratic; monotone increasing in flux, derived from unweighted fit to 13-week Monticello
12 etal. (1996) ULLI data.
13
14 aN = Exp{-3.858 - 4.809 x Exp[- (9.293* 10'5 *flux)2]} (E-9)
15
16 N5: Linear-quadratic-cubic; initially increasing slightly with increasing flux, then decreasing
17 slightly, and finally increasing, derived from fit to 13-week Monticello et al. (1996) ULLI data,
18 using average flux over all sites for a given ppm exposure and weighting regression by estimates
19 of the numbers of cells at each of five sites.
20
21 aN = Exp{-5.488 - 2.755 * Exp[-7.808*10~5 * flux + (2.349*10^ * flux)2 (E-10)
22 - (2J66*J(T4 x flux)3]}
23
24 N6: Linear-quadratic-cubic; initially increasing slightly with increasing flux, then decreasing
25 slightly, and finally increasing, derived from fit to all Monticello et al. (1996) ULLI data, using
26 weeks of exposure as a covariate. In this model, time was a regression (continuous) predictor,
27 not a class variable, and its coefficient represents the decrease in logic aN per week of exposure
28 time.
29
30 aN = Exp{7.785* 10'3 * (weeks) - 5.722 - 2.501 * Exp[1.103*10~4 * flux (E-l 1)
31 - (7.223*10~5 * flux)2- (1.575* J(T4 * flux)3]}
32
This document is a draft for review purposes only and does not constitute Agency policy.
E-29 DRAFT—DO NOT CITE OR QUOTE
-------
1 E.3.3. Uncertainty in Model Specification of Initiated Cell Replication and Death
2 E.3.3.1. Biological Implications of Assumptions in Conolly etal. (2003)
3 The results of a two-stage MVK model are extremely sensitive to the values for initiated
4 cell division (0,1) and death (Pi) rates, particularly in the case of a sharply rising dose-response
5 curve as observed of formaldehyde. The pool of cells used for obtaining the available LI data
6 (Monticello et al., 1996, 1991) consists of largely normal cells with perhaps increasing numbers
7 of initiated cells at higher exposure concentrations. As such there is no way of inferring the
8 division rates of initiated cells in the nasal epithelium, either spontaneous (baseline) or induced
9 by exposure to formaldehyde, from the available empirical data. Conolly et al. (2003)
10 considered ai(flux) as a function of afflux) as given by eq D-2 in Appendix D. As shown in
11 Figure D-l (Appendix D), oti is estimated in Conolly et al. (2003) to be very similar to O,N. That
12 is, with eq D-2 assumed to relate ai(flux) to afflux), a J- or hockey-shaped dose-response curve
13 for afflux) necessarily results in a J or hockey shape for ai(flux).
14 The J shape for the TWA afflux) in Conolly et al. (2003) could plausibly be explained,
15 as suggested by the examples in Conolly and Lutz (2004), by a mathematical superposition of
16 dose-response curves describing the effects of the inhibition of cell replication by the formation
17 of DPXs (Heck and Casanova, 1999) and cytotoxicity-induced regenerative replication (Conolly,
18 2002). However, as explained earlier, there is considerable uncertainty and variability, both
19 qualitative and quantitative, in the interpretation of the LI data and in the derivation of normal
20 cell replication rates from the ULLI data. While the TWA values of ULLI indicate a J-shaped
21 dose response for some sites, as also concluded by Gaylor et al. (2004), this is not consistently
22 the case for all exposure times and sites as discussed earlier. Notwithstanding this uncertainty
23 and variability, and in the absence of data, the following essential questions have a significant
24 impact on risk predictions and need resolution if the model structure in eq D-2 is to be used in a
25 biologically based (or motivated) sense:
26
27 • Should mechanisms that might explain a J-shaped dose response for normal cell
28 replication be expected to prevail also for initiated cells? An identical question can be
29 posed for the hockey-stick-shaped curve which indicates a cytotoxicity-driven threshold
30 in dose response.
31 • Would the formaldehyde flux at which the cell replication dose-response curve rises
32 above its baseline be similar in value for both normal and initiated cells as inferred by the
3 3 CUT model in Figure D-1 ?
34
This document is a draft for review purposes only and does not constitute Agency policy.
E-30 DRAFT—DO NOT CITE OR QUOTE
-------
1 The next critical assumption in Conolly et al. (2003) was that made for PI (the death rate
2 of initiated cells), namely, Pi(flux) = afflux) (see eq D-3). The rationale for this assumption is
3 explained by assuming formaldehyde to be equally cytotoxic to initiated and normal cells since
4 the mechanism is presumed to be via its general chemical reactivity (Subramaniam et al. 2008).
5 In essence, this assumption brings the cytotoxic action of formaldehyde to bear strongly on the
6 parameterization of the CUT model.
7 There are no data to evaluate the strength of these assumptions, so Subramaniam et al.
8 (2008) studied the plausibility of various inferences that arise as a result of these assumptions.
9 These inferences are only briefly listed here (see the paper for further discussion).
10
11 • For flux <27,975 pmol/mm2-hour, ai > (XN (see Figures D-l and D-2 of Appendix D).
12 Qualitatively, this concept of a growth advantage is in line with data on epithelial and
13 other tissue types with or without exposure to specific chemicals.
14 • For higher flux levels, however, the model indicates ai <
-------
1 D-3 significantly impact the intended use of the model, namely extrapolation to low-dose human
2 cancer risk and the calculation of an upper bound on human risk. Any such alternate model
3 structure needs to provide a good fit to the time-to-tumor data.
4
5 E.3.3.2. Plausible Alternative Assumptions for ai and fa
6 Therefore, in the additional sensitivity analysis presented here,
7 a) Initiated cell kinetics are considered to be independent of normal cells,
8 b) Initiated cell replication dose-response cannot take a J shape; this is motivated by
9 the consideration that lower-than-baseline turnover rate represents an increased
10 amount of DNA repair taking place, which may not be consistent with impaired
11 DNA repair in initiated cells.
12 Thus, two alternatives were considered to eq D-2 for afflux):
13
14 II: a/ = 77 x [l + exp(j2 / J3)] / {1 + exp[-(flux - y2) / y3]} (E-12)
15
16 12: af = max[ai(ll), aNBasaJ (E-13)
17
18 Here 77,72, and ys are parameters estimated by fitting the cancer model to the rat bioassay
19 data. In eq E-12, a/ increases monotonically with flux from a background level of 77
20 asymptotically up to a maximum value of 77 x [1 + Exp(y2 / Js)]- The choice of this functional
21 form in eq E-12 and eq E-13 was considered in order to be parsimonious while at the same time
22 allowing for a flexible shape to the dose-response curve. The sigmoidal curve allows for the
23 possibility of a slow rise in the curve at low dose and an asymptote.
24 Equation E-13 is a modification of eq E-12 that restricts the rate of division of initiated
25 cells to be at least as large as the spontaneous division rate of unexposed normal cells. There is
26 evidence to suggest (e.g., in the case of liver foci) that initiated cells have a growth advantage
27 over normal cells, with or without exposure to specific chemicals (Ceder et al., 2007;
28 Grasl-Kraupp et al., 2000; Schulte-Hermann et al., 1999; Coste et al., 1996; Dragan et al., 1995).
29 In addition, in most runs, an upper bound (ahigh) is selected for both aN and a7. This value
30 is assumed to represent the largest biologically plausible rate of cell division. Following Conolly
31 et al. (2003), in most cases a^gh is set equal to 0.045 hours"1. If a value of at or aN computed
32 using one of the above formulas exceeded a^gh, the value ofahigh was used in the computation
33 rather than the value obtained by using the formula.
34 As noted above, Conolly et al. (2003) set the rate of death for intermediate cells, ft/, equal
35 to the division rate of normal cells, ft/ = a^. On the other hand, apoptotic rates and cell
This document is a draft for review purposes only and does not constitute Agency policy.
E-32 DRAFT—DO NOT CITE OR QUOTE
-------
1 proliferation rates are thought to be coupled (Schulte-Hermann, 1999; Moolgavkar, 1994), so
2 that death rates of initiated cells would rise concomitantly with an increase in their division rates
3 (Grasl-Kraupp et al., 2000; Schulte-Hermann et al., 1999). Therefore, as an alternative to the
4 Conolly et al. (2003) formulation, it is assumed that the death rate of intermediate cells is
5 proportional to the division rate of intermediate cells.
6 fr = 1:0*0.! (E-14)
7
8 where the constant of proportionality, Kp, is an additional parameter to be estimated by
9 optimization against the tumor incidence data. Such an assumption has also been made by other
10 authors (Luebeck et al., 2000, 1995; Moolgavkar et al., 1993).
11
12 E.3.4. Results of Sensitivity Analyses on (XN? cti, and PI
13 E.3.4.1. Further Constraints
14 The number of models that might be constructed if all the possibilities listed above for
15 a^ a/, and Pi are to be tried in a systematic manner clearly become exponential and daunting.
16 (Optimally, it would have been desirable to elucidate the role of a specific modification while
17 keeping others unchanged to determine risk.) Therefore, in order to carry out a viable sensitivity
18 analysis while at the same time examining the plausible range of risks resulting from variations
19 in parameters and model structures, various uncertainties were combined in any given
20 simulation. By using the constraints described above (see eqs E-6 through E-13 and associated
21 text) for oti, PI, and OCN, 19 models were obtained that provided similarly good fits to the time-to-
22 tumor data (which in some cases contained only five dose groups).
23 However, for many of these models, the optimal ai(flux) displayed a threshold in flux
24 even when the model utilized for afflux) was a monotonic increasing curve without a threshold
25 (i.e., model N4 for OCN in Figure E-4). Indeed, if a thresholded dose-response curve was
26 plausible for 0,1 based on arguments of cytotoxicity, then a threshold is all the more plausible for
27 OCN, and such models are removed from consideration.
28 Secondly, the basal value of ai was required to be at least as large as the basal value of
29 aN. Another constraint was placed on the baseline initiated cell replication rate. In the absence
30 of formaldehyde exposure, 0,1 was not allowed to be greater than two or four times aN, even if
31 such models described the tumor data, including the control data, very well. There are some data
32 that suggest that baseline initiated cells have a small growth advantage over normal cells, so a
33 huge advantage was thought to be biologically less plausible.
This document is a draft for review purposes only and does not constitute Agency policy.
E-33 DRAFT—DO NOT CITE OR QUOTE
-------
1 Finally, since most of the SCCs in the rat bioassays occurred in rats exposed to the
2 highest formaldehyde concentration (15 ppm), the data from this exposure level have a big
3 impact on the estimated model parameters. In most runs that incorporated the 15 ppm data, the
4 model appeared, based on inspection of the KM plots, to fit the 15 ppm data quite well but to fit
5 the lower exposure data less well. Because of the high level of necrosis occurring at 15 ppm, it
6 is possible that the data at this exposure may not be particularly relevant to modeling the sharp
7 upward rise in the dose response at 6 ppm. Furthermore, the principal interest is in the
8 predictions of the model at lower levels to which human populations may be exposed.
9 Consequently, in order to improve the fit of the model at lower exposures, some of the
10 alternative models were constructed with the 15 ppm data omitted.
11
12 E.3.4.2. Sensitivity of Risk Estimates for the F344 Rat
13 Figure E-5 contains plots of the MLE of additional risk computed for the F344 rat at
14 formaldehyde exposures of 0.001, 0.01, 0.1, and 1 ppm for eight models. Two log-log plots are
15 provided. For those models for which the estimates of additional risk are all positive, the
16 additional risks are plotted (panel A), and, for those for which estimates of additional risk are
17 negative, the negatives of additional risks are plotted (panel B). Only five dose groups were
18 considered (i.e., 15 ppm data omitted) for models 8, 5, 15, and 16. Figure E-6 shows the dose-
19 response curves for O,N and 0,1 for these eight cases (panels A and B corresponding to those in
20 Figure E-5). The specification and estimated values of the parameters for these models are
21 provided in Tables E-4 and E-5. The primary results are as follows:
22
23 1. Among the models considered, negative values for additional risk can arise only in
24 models in which the dose response for normal cells is J shaped. Thus, all of the models
25 with negative dose responses for risk have J-shaped dose responses for normal cells.
26 However, the converse is not necessarily true as may be noted from model 8. This model
27 has both a positive dose response for risk and a J-shaped dose response for normal cells.
28 In this case, the strong positive increase in response of initiated cells at low dose was
29 sufficient to counteract the negative response of normal cells.
30 2. For doses below which no tumors were observed, the risk estimates predicted by the
31 different models span a very large range. This result points to large uncertainties in
32 model specification (how to relate the kinetics of normal and initiated cells) as well as in
33 parameter values. As mentioned above, the analysis does not attempt to separate the
34 influence of the different sources of uncertainty, so this range also incorporates the
35 uncertainty arising from the use of different control data and that due to amax.
This document is a draft for review purposes only and does not constitute Agency policy.
E-34 DRAFT—DO NOT CITE OR QUOTE
-------
o S. 5^ 1.E-04 -i
*""+• f^ <*-.
O &
a ^ o'
Q. § D -I F-Ci^i -
~-. t~^ SS ' • *— *-*^
^ ^3 ^2
Sf- ^ <*
iS.
xT o ;-• 1.E-06 -
f^ S
^ Si ^:
^ 1^ 1.E-07 -
•^ TO ^
^ 1.E-08C
V*1
w <
'C 1.E-09-
•o ^
0
g 1.E-10-
<
1.E-11 -
1.E-12 -
1.E-13 -
\
1.E-14/
^.E-^5 -
n
0
U A
O
D
A
O X
]
A
>
.^-^•"•'T
»*«*»***ov" ^
O model 8
D model 15
A model 16
X model 17
• — ^-**^
As in Conoiiy et ai. (2003), Hockey
stick OCN and a, but using concurrent
X controls
f
\
i
0.001 0.01 0.1 1
^ formaldehyde exposure cone (ppm)
Figure E-5A. BBDR models for the rat — models with positive added risk.
O
> Note: All four models provide "similar" fits to tumor data (see text).
H
O
0
-------
a 5 KO
o S. S3
*•
§
w
0.001
1.E-08
-1.E-07 -
(A
"2-1.E-06
LLI
,-1.E-05 -
o
LLI
-1.E-04 -
-1.E-03 J
Exposure cone (ppm)
0.01
0.1
As \r\ Cor\o\\y ei a\. (2003), J-shape aN
and a, bui using Inhalation NTP controls
added to concurrent
fe
H
O
o
Figure E-5B. BBDR rat models resulting in negative added risk.
Note: All four models provide "similar" fits to tumor data (see text).
-------
ON
"""I
Model 8
Model 15
0 10000 20000 30000 40000
Flux (pmole/mm2/hr)
AH 8 models in A & B provide
similar fits to tumor
10000 20000 30000 40000
Flux (pmole/mm2/hr)
Hockey stick aN and a, with concurrent
controls
T3,
s
Model 16
0 10000 20000 30000 40000
Flux (pmole/mm2/hr)
10000 20000 30000 40000
Flux (pmole/mm2/hr)
Figure E-6A. Models resulting in positive added rat risk: Dose-response for
normal and initiated cell replication.
This document is a draft for review purposes only and does not constitute Agency policy.
E-37 DRAFT—DO NOT CITE OR QUOTE
-------
Model 3
10000 20000 30000 40000
Flux (pmole/mm2/hr)
•ON
---a.
Model 4
10000 20000 30000 40000
Flux (pmole/mm2/hr)
AH 8 models in A & B provide
fits to tumor
Model 5
0 10000 20000 30000 40000
Flux (pmole/mm2/hr)
J-shape aN, a, in Conolly (2003) with
Inhalation NTP + concurrent controls
Model 13
0 10000 20000 30000 40000
Flux (pmole/mm2/hr)
Figure E-6B. Models resulting in negative added rat risk: Dose-response for
normal and initiated cell replication.
This document is a draft for review purposes only and does not constitute Agency policy.
E-3 8 DRAFT—DO NOT CITE OR QUOTE
-------
^
K- 5:
Table E-4. Parameter specifications and estimates for clonal growth models of nasal SCC in the F344 rat using
alternative characterization of cell replication and death rates
Parameters
Historical controls added to
concurrent
Number of dose groups
DPX concentration
aN definition
a/ definition
Q-high
/?/ definition
Log-likelihood
^NBasal
KMU
KMX(KMU/juNBasal)
D/
r> $
iJOF
7i
72
7s
Kp
Model 3
Inhalation NTP
6
Subramaniam et
al. (2007)
N3
12
~
fr = Ke x a/
-1495.34
7.518E-7
3.884E-7
0.5166
214.3
75.26
1.164E-5
1427
11944
0.9893
Model 4
Inhalation NTP
6
Subramaniam et
al. (2007)
N6
12
0.045
fr = Ke x a/
-1495.61
1.664E-6
3.471E-7
0.2086
199.7
79.81
1.006E-5
1591
13017
0.9848
Model 5
Inhalation NTP
5
Subramaniam et
al. (2007)
N3
12
~
fr = Ke x a/
-184.02
8.684E-7
0.0
0.0
261.8
119.7
3.168E-5
1825
14207
0.9804
Model 8
Inhalation NTP
5
Subramaniam et
al. (2007)
N6
11
0.045
fr = Ke x a/
-184.22
9.230E-7
0.0
(0.0, 2.093E-6)
0.0
(0.0, 4.696)
254.2
101.1
2.967E-4
3223
15989
0.9504
Model 15
Inhalation NTP
5
Subramaniam et
al. (2007)
N4
11
0.045
fr = Ke x a/
yi < 4 O-NBasal
-182.75
1.037E-6
4.582E-6
(1.8E-6,1.86E-5)
4.420
(1.53, 17.67)
423.2
100.8
6.888E-4
4652
54334
1.006
Model 16
Inhalation NTP
5
Subramaniam et
al. (2007)
N4
11
0.045
fr = Ke x a/
yi < 2 O-NBasal
-186.37
1.662E-7
0.0
0.0
245.1
98.83
3.441E-4
2818
37896
0.9660
§ 5'
I-S
w
OJ
vo
fe
H
o
o
Subramaniam et al. (2007) for an explanation of the time delay constants D0and DOF.
-------
Table E-5. Parameter specifications and estimates for clonal growth models
of nasal SCC in the F344 rat using cell replication and death rates as
characterized in Conolly et al. (2003)
Parameters
Historical controls added
to concurrent
Number of dose groups
DPX concentration
aN definition
a/ definition
O-high
fii definition
Log-likelihood
fiNBasal
KMU
KMX(KMU/juNBasal)
D/
Do/
multib
multic
(%max
Model 13
All NTP
6
Conolly et al. (2000)
J-shape
(TWA, Conolly et al. 2003)
eq. D-l
—
fii = aN
-1692.68
1.731E-6
0.0
0.0
239.5
66.31
1.047
1.510
5.153E-2
Model 17
NO historical controls
6
Subramaniam et al. (2007)
Hockey
(TWA, Conolly et al., 2003)
eq. D-l
—
fii=aN
-1474.29
0.0
1.203E-6
(1.0E-6,1.427E-6)
Infinite
(0.4097,infmite)
243.13
68.83
1.078E+0
3.347
0.045
1
2
3
4
5
6
7
9 See Subramaniam et al. (2007) for an explanation of the time delay constants D0 and DOF.
6
7
3. At the 10 ppb (0.01 ppm) concentration, MLE risks range from -4.0 xio ° to +1.3
At this dose, models that gave only positive risks resulted in a five orders of magnitude
risk range from 1.2 xlO~12to 1.3 xlO~7, while narrowing to a four orders of magnitude
risk range from 1.2 x 10~10 to 1.3 x 10"6 at the 0.1 ppm level. This narrowing continues as
exposure concentration increases, and the curves coalesce to substantially similar values
at 6 ppm and above (not shown). For all these 8 models, the rat added risk at 6.0 ppm
ranged from 1.8 xlO~2to2.1 xio~2.
4. There does not seem to be any systematic effect on additional risk that depends on
whether the 15 ppm data are included in the analysis.
This document is a draft for review purposes only and does not constitute Agency policy.
E-40 DRAFT—DO NOT CITE OR QUOTE
-------
1 5. For all of the models except models 13 and 17 in Figures E-5 and E-6, the additional risk
2 varies substantially linearly with exposure at low exposures between 0.001 and 1.0 ppm
3 (departing only to a small extent from linearity between 0.1 and 1.0 ppm). Models 13
4 and 17 show a quadratic dependence; these models employ the TWA J-shape and hockey
5 stick dose response curves for O,N used in Conolly et al. (2003) and the same equations
6 used by those authors to relate 0,1 and PI to O,N (see eqs D-2 and D-3, Section D-6).
7 However, the control data in Model 17 was different from those used by Conolly et al.;
8 while all NTP controls were added to the concurrent controls in model 13, only
9 concurrent controls were used in model 17.
10
11 The various model choices presented in Figure E-5 all provided equally good fits to the
12 time-to-tumor data although within the context of a significant qualification. It was not possible
13 to simply use the maximized log-likelihood values as a means of comparing the goodness-of-fit
14 to the tumor incidence data across all these model choices. This is because many of the model
15 choices differed in the number of doses or in the number of control animals that were used, so
16 the fits were compared across such models only visually.
17 Wherever results from the BBDR modeling are discussed, values of added risk, as
18 opposed to extra risk, are reported. This is purely for convenience in interpretation. Because of
19 the low background incidence, these values are only negligibly different from the corresponding
20 extra risk estimate. The final risk (or unit risk) estimates provided in this document are based on
21 extra risk estimates.
22
23 E.3.4.3. MOA Inferences Revisited
24 The ratio KMU/^Nbasai represents the added fractional probability of mutation per cell
25 generation (//N - //NbasaiV/^Nbasai due to unit concentration of DPXs. As discussed in Sections
26 E.3.1.2 and E.3.1.5 (see Appendix E), this parameter has a critical impact on the extrapolation as
27 well as on inferring whether the mutagenic action of formaldehyde is relevant to explaining the
28 observed tumor incidence or its carcinogenicity at lower concentrations. In that prior discussion,
29 this ratio was found to be extremely sensitive to the choice of historical control data. The
30 analysis indicates that, for a given set of control data that is used, uncertainties associated with
31 aN and 0,1 also have a large impact on this ratio.
32 As discussed in E.3.1.2, this ratio was infinite when concurrent controls were used
33 because the MLE value for //Nbasai was found to be zero. The use of these concurrent controls,
34 however, does not necessarily imply that //Nbasai will be determined to be zero. In one of the
35 scenarios examined in the sensitivity analysis, where concurrent controls were used along with
36 the combination of dose-response curves eq D-9 for aN (see Figure E-4) and eq E-13 for ai, the
This document is a draft for review purposes only and does not constitute Agency policy.
E-41 DRAFT—DO NOT CITE OR QUOTE
-------
1 optimal value of the ratio KMU///Nbasai was equal to 0.25. For the models in Figure 5-13 A, this
2 ratio was 0 for all except model 17 for which it was infinite. For the models in Figure 5-13B
3 with negative added risk, the ratio ranged from 0-4.5. For some of those models where
4 KMU///Nbasai was finite, the upper confidence bound on this ratio was found to increase by an
5 order of magnitude from the MLE value.
6 Thus, we conclude that the modeling does not help resolve the debate as to the relevance
7 of formaldehyde's mutagenic potential to its carcinogenicity.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
E.3.4.4. Confidence Bounds: Model Uncertainty Versus Statistical Uncertainty
For models 15 and 17 in Figures E-5A and E-6A, 90% CIs for additional risk were
calculated by using the profile likelihood method. Table E-6 compares the lower and upper
confidence bounds for these models for 0.001 ppm, 0.1 ppm (doses well below the range where
tumors were observed), and 6 ppm (the lowest dose where tumors were observed) with the MLE
risk estimates at these doses. In both cases, these intervals were quite narrow compared with the
differences in risk predicted by different models in Figure E-5. This suggests that model
uncertainty is of more consequence in the formaldehyde animal model than is statistical
uncertainty. We also estimated confidence bounds using the bootstrap method for select models,
and determined that these estimates were in agreement with the bounds calculated using the
profile likelihood method. These results are not presented here. We return to the calculation of
confidence limits when determining points of departure (PODs).
Table E-6. Comparison of statistical confidence bounds on added risk for
two models
Dose (ppm)
0.001
0.1
6
Model
Model 15
Model 17
Model 15
Model 17
Model 15
Model 17
Lower
bound
4.4 x l(T9
1.2 x l(T14
4.5 x l(T7
1.2 x l(T10
1.8 x l(T2
1.3 x l(T2
MLE
1.3 x l(T8
1.2 x 10-
14
1.3 x 1Q-6
1.2 x 10-
10
2.1 x 1(T2
1.8 x l(T2
Upper
bound
1.6 x l(T8
1.3 x l(T14
1.7 x 1Q-6
1.3 x l(T10
2.3 x l(T2
3.0 x l(T2
In conclusion, it is demonstrated that the different formaldehyde clonal growth models
can fit the data about equally well and still produce considerable variation in additional risk and
biological inferences at low exposures. However, even with these large variations, the highest
This document is a draft for review purposes only and does not constitute Agency policy.
E-42 DRAFT—DO NOT CITE OR QUOTE
-------
1 MLE added risk for the F344 rat is only of the order of lO"6 at 0.1 ppm. Thus, with regard to
2 calculating a reasonable upper bound that includes model and statistical uncertainty, the relevant
3 question is whether the range arising out of uncertainties in the rat model amplifies when
4 extrapolated to the human. Thus, in Appendix F, the human model in Conolly et al. (2004) will
5 be examined.
This document is a draft for review purposes only and does not constitute Agency policy.
E-43 DRAFT—DO NOT CITE OR QUOTE
-------
This page intentionally left blank.
This document is a draft for review purposes only and does not constitute Agency policy.
E-44 DRAFT—DO NOT CITE OR QUOTE
-------
Appendix F
-------
1 APPENDIX F
2
3 SENSITIVITY ANALYSIS OF BBDR MODEL FOR FORMALDEHYDE INDUCED
4 RESPIRATORY CANCER IN HUMANS
5
6
7 F.I. MAJOR UNCERTAINTIES IN THE FORM ALDEHYDE HUM AN BBDR MODEL
8 Subsequent to the BBDR model for modeling rat cancer, Conolly et al. (2004) developed
9 a corresponding model for humans for the purpose of extrapolating the risk estimated by the rat
10 model to humans. Also, rather than considering only nasal tumors, it is used to predict the risk
11 of all human respiratory tumors. The human model for formaldehyde carcinogenicity (Conolly
12 et al., 2004) is conceptually very similar to the rat model and follows the schematic in Figure
13 5-11 in Chapter 5. The model structure, notations, and calibration are described in Appendix D.
14 Unlike the sensitivity analysis of the rat modeling where a number of issues were examined, a
15 much more restricted analysis will be presented here for the sake of brevity. A more extensive
16 analysis was carried out initially that carried forward several of the rat models from Appendix E
17 to the human, and the lessons learned from those exercises are in agreement with the more
18 restricted presentation that follows. Table F-l lists the major uncertainties and assumptions in
19 the human extrapolation model in Conolly et al. (2004).
20
This document is a draft for review purposes only and does not constitute Agency policy.
F-2 DRAFT—DO NOT CITE OR QUOTE
-------
5 3 a
Table F-l. Summary of evaluation of major assumptions and results in CUT human BBDR model
Assumptions"
Rationale in Conolly et al. (2003) or
CUT (1999)
EPA evaluation
Further elaboration
TO
a
I
Cell division rates derived from rat
labeling data were assumed
applicable to human (except for
assuming different fraction of cells
with replicative potential).
There are no equivalent LI data for
human or guidance for extrapolating cell
division rate across species.
Enzymatic metabolism plays a role in mitosis.
Therefore, we expect interspecies difference in cell
division rate. Basal cell division rates in humans are
expected to be much more variable than in laboratory
animals.
Subramaniam et al.
(2008)
Parameters for enzymatic
metabolism of formaldehyde in
human PBPK model for DPX
concentrations: Km varies by order
of magnitude between rat and
monkey but is same for monkey and
human. Vmax/Km is similar for rat
and monkey but 6-fold lower for
human.
See text (Section 3.6.6.2)
See text (Section 3.6.6.2)
Section 3.6.6.2;
Conolly et al. (2000);
Subramaniam et al.
(2008); Klein et al.
(2010)
Anatomically realistic representation
of nasal passages.
Reduces uncertainty (over default
calculation carried out by averaging dose
over entire nasal surface).
Computer representation pertains to that of one
individual (Caucasian male adult). There is
considerable interindividual variability in nasal
anatomy. Susceptible individuals are even more
variable.
Kimbelletal. (200la,
b); Subramaniam et
al. (2008, 1998)
asd is species invariant
(used to estimate human).
Human cells are more difficult to
transform than rodent, both
spontaneously and by exposure to
formaldehyde.
is 0 when concurrent controls or inhalation NTP
controls in time frame of concurrent bioassays are
used. Leads to infinitely large KMU for human.
Subramaniam et al.
(2007); Crump et al.
(2009, 2008).
fe
H
Conservative assumptions were
made. Results are conservative in
the face of model uncertainties.
1) Hockey-stick dose-response for aN was
included even though TWA indicated
J-shape.
2) Overall respiratory tract cancer
incidence data for human baseline rates
were used.
3) Risk was evaluated at statistical upper
bound of the proportionality parameter
relating DPXs to the probability of
mutation.
Results in Conolly et al. (2004) are not conservative in
the face of model uncertainties: (a) Human risk
estimates are very sensitive to use of historical controls
in the analysis of the animal bioassay. (b) Human risk
estimates are unboundedly large when concurrent
controls are used in rat model, (c) Minor perturbations
in model assumptions regarding division and death
rates of initiated cells lead to upper bound risks that
were more than 1,000-fold greater than the highest
estimates in Conolly et al. (2004).
Conolly et al. (2004);
Subramaniam et al.
(2007); Crump et al.
(2009, 2008).
""Assumptions in this table are in addition to those listed for the BBDR model for the F344 rat.
o
o
-------
1 F.2. SENSITIVIT Y ANALYSIS OF HUM AN BBDR MODEL ING
2 Crump et al. (2008) carried out a limited sensitivity analysis of the Conolly et al. (2004)
3 human model. This analysis was limited to evaluating the effect on the human model of the
4 following. These evaluations have been the subject of some debate in the literature and at
5 various conferences (Conolly, 2009; Conolly et al., 2009, 2008; Crump et al., 2009).
6
7 1. The use of the alternative sets of control data for the rat bioassay data that were considered in
8 the sensitivity analysis of the rat model in Appendix E.
9 2. Minor perturbations in model assumptions regarding the effect of formaldehyde on the
10 division and death rates of initiated cells (oti, PI).
11 • As mentioned in Section D.7 one (of the two) adjustable parameter in the expression
12 for the human 0,1 in Conolly et al. (2004) was determined from the model fit to the rat
13 tumor incidence data while the second parameter was determined from background
14 rates of cancer incidence in the human. Therefore, variations considered in oti were
15 constrained to only those that (a) did not meaningfully degrade the fit of the model to
16 the rat tumor incidence data and (b) were in concordance with background rates in the
17 human.
18 • Crump et al. (2008) also evaluated these variations with respect to their biological
19 plausibility. The sensitivity analysis on assumed initiated cell kinetics was thought to
20 be particularly important since there were no data to even crudely inform the kinetics
21 of initiated cells for use in the models, even in rats, and the two-stage clonal
22 expansion model is very sensitive to initiated cell kinetics (Gaylor and Zheng, 1996;
23 Crump, 1994a, b).
24
25 Crump et al. (2008) note that, since the purpose of their analysis was to carry out a
26 sensitivity analysis, in order to illustrate certain points, only risks to the general U.S. population
27 from constant lifetime exposure to various levels of formaldehyde under the Conolly et al.
28 (2004) environmental scenario (8 hours/day sleeping, 8 hours/day sitting, and 8 hours/day
29 engaged in light activity) are considered. Fits based on the hockey-stick and J-shape models
30 were identical, and, of the three estimated parameters (jibasal, multb, and D), only the estimate
31 of jibasal differed between the two models.
32
33 F.2.1. Effect of Background Rates of Nasal Tumors in Rats on Human Risk Estimates
34 Crump et al. (2008) quantitatively evaluated the impact of different control groups on
35 estimates of additional human risk as follows:
This document is a draft for review purposes only and does not constitute Agency policy.
F-4 DRAFT—DO NOT CITE OR QUOTE
-------
4
5
6
7
1. Concurrent controls plus all NTP controls:, the same as used by Conolly et al. (2004);
2. Concurrent controls plus controls from NTP inhalation studies;
3. Only concurrent controls;
4. Each set of control data was applied with both the J shape and hockey-stick models in
Conolly et al. (2004) for afflux) and ai(flux) for a total of six analyses;.
5. Uncertainties associated with (XN or ai are not addressed. Parameters amax, multfc, and
KMU were estimated in exactly the same manner as in Conolly et al. (2004).
9 Crump et al. (2008) present the following dose-response predictions of additional risk in
10 humans from constant lifetime exposure to various levels of formaldehyde arising from
11 exercising the above six cases. Their plots are reproduced in Figure F-l, where the
12 corresponding curves based on Conolly et al. (2004) are also shown for comparison.
0.1
jockey, Concurrent Controls, MLE and 95% UB;
l-Shape, Concurrent Controls, 95% UB
1E-6-
ft
ockey, Inh. NTP
Controls. 95% UB
J-Shape, Concurrent
ontrols, MLE
Hockey, All NTP
Controls, 95% UB
Hockey, Inh. NTP
Controls, MLE „ .
Conolly et al. (2064),
\
1 P 7
1E-8n
1E
„ — ^ — "
' - " "
-3
Hockey UB /
J-Shape, All NTP Controls, MLE and 95% UB;
J-Shape, Inh. NTP Controls, MLE and 95% ilB;
Conolly et al. (2004) J-Shace UB '
tsn 05 ' 'ii'"
Hockey,
All NTP
Controls
MLE
1
Formaldehyde Exposure (ppm)
Figure F-l. Effect of choice of NTP bioassays for historical controls on
human risk.
Note: Estimates of additional human risk of respiratory cancer by age 80 from
lifetime exposure to formaldehyde are obtained by using different control groups
of rats.
Source: Crump et al. (2008).
This document is a draft for review purposes only and does not constitute Agency policy.
F-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 The lowest dotted curve in Figure F-l represents the highest estimates of human risk
2 developed by Conolly et al. (2004). This resulted from use of the hockey-stick model for cell
3 division rates in conjunction with the statistical upper bound for the parameter KMU. As
4 indicated by the downward block arrows in the figure, their corresponding estimates based on
5 the J-shape model were all negative for exposures below 1 ppm.
6 Consider next the solid curves in the figure, which show predicted MLE added risks that
7 were positive and less than 0.5. Crump et al. (2008) next examined the added risk obtained
8 when the MLE estimate of (KMUI^basai) in these cases is replaced by the 95% upper bound of
9 this parameter ratio. The upper bound risk estimates in Conolly et al. (2004) were calculated in a
10 similar manner (but using all NTP historical controls). Except for minor differences, risk
11 estimates corresponding to such an upper bound and using all NTP controls were very similar in
12 the two efforts (Crump et al., 2008; Conolly et al., 2004).
13 Figure F-l shows that the choice of controls to include in the rat model can make an
14 enormous difference in estimates of additional human risk. For the J-shaped model for cell
15 replication rate both estimates based on the MLE and those based on the 95% upper bound on
16 KMUI^basai are negative for formaldehyde exposures below 1 ppm. However, when only
17 concurrent controls are used in the model in Crump et al. (2008), the MLE from the J-shape
18 model is positive and is more than three orders of magnitude higher than the highest estimates
19 obtained by Conolly et al. (2004). Using only concurrent controls, estimates based on the 95%
20 upper bound on KMU/jUbasai are unboundedly large (block arrows at the top of the figure). For
21 the hockey-stick shaped model for cell replication rate, when all NTP controls are used, the
22 estimates based on the MLEs are zero for exposures less than about 0.5 ppm. If only inhalation
23 controls are added, the MLEs are about seven times larger than the Conolly et al. (2004) upper
24 bound estimates, and the estimates based on the 95% upper bound on KMUIpbasai are about 50
25 times larger than the Conolly et al. (2004) estimates. If only concurrent controls are used, both
26 the MLE estimates and those based on the 95% upper bound on KMUI^basai are unboundedly
27 large.
28
29 F.2.2. Alternative Assumptions Regarding the Rate of Replication of Initiated Cells
30 For the human model, Conolly et al. (2004) made the same assumptions for relating
31 ai(flux) and Pi(flux) to aN(flux) as in their rat model (Conolly et al., 2003). That is, these
32 quantities were related by using eqs D-2 and D-3 (see Appendix D). As discussed in the context
33 of the rat modeling, by extending the shape of these curves to humans, the authors' model brings
34 the cytotoxic action of formaldehyde to bear strongly on the parameterization of the human
35 model as well.
This document is a draft for review purposes only and does not constitute Agency policy.
F-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 In the sensitivity analyses of the rat modeling in Appendix E, it was concluded that other
2 biologically plausible assumptions for oti and PI resulted in several orders of magnitude
3 variations in the low dose risk relative to those obtained by models based on the assumptions in
4 Conolly et al. (2003) but that the highest risks were nonetheless of the order of 10~6 at the 10 ppb
5 level. This section examines how these uncertainties in the rat model propagate to the human
6 model.
7 Crump et al. (2008) made minor modifications to the assumed division rates of initiated
8 cells in Conolly et al. (2004), while all other aspects of the model and input data were kept
9 unchanged. Two alternatives were considered for each of the J-shape and hockey-stick models.
10 Figure F-2 shows the hockey-stick model for initiated cells in rats. In the first modification to
11 the hockey-stick model (hockey-stick Mod 1), rather than having a threshold at a flux of
12 1,240 pmol/m2-hour, the division rate increases linearly with increasing flux until the graph
13 intersects the original curve at 4,500 pmol/m2-hour, where it then assumes the same value as in
14 the original curve for larger values of flux. The second modification (hockey-stick Mod 2) is
15 similar, except the modified curve intersects the original curve at a flux of 3,000 pmol/m2-hour.
a.
ro
0.05 n
0.04-
0.03-
0.02-
0.01-
Mod1
0.00
0.0010-
10000 20000 30000
Formaldehyde Flux [pmol/(m2-h)]
40000
Figure F-2. Conolly et al. (2003) hockey-stick model for division rates of
initiated cells in rats and two modified models.
Source: Crump et al. (2008).
This document is a draft for review purposes only and does not constitute Agency policy.
F-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 Figure F-3 shows the rat J-shape model for initiated cells. In the first modification to this
2 dose response (J-shape Mod 1), rather than having a J shape, the division rate of initiated cells
3 remains constant at the basal value until the original curve rises above the basal value and has
4 the same value as the original curve for larger values of flux. In the second modification
5 (J-shape Mod 2), the J shape is retained but somewhat mitigated. In this modification, the
6 division rate initially decreases in a linear manner similar to that of the original model but with a
7 less negative slope until it intersects the original curve at a flux of 1,240 |im/m2-hour, where it
8 then follows the original curve for higher values of flux.
0.05n
0.02-
0.01 -
000
10000 20000 30000
Formaldehyde Flux [pmol/(m'-h)]
40000
Figure F-3. Conolly et al. (2003) J-shape model for division rates of initiated
cells in rats and two modified models.
Source: Crump et al. (2008).
9 Since the first constraint on the variation in oti was in concordance with the rat time-to-
10 tumor incidence data, Crump et al. (2008) applied each of the modified models in Figures F-2
11 and F-3 to the version of the formaldehyde models in Subramaniam et al. (2007) that employed
12 all NTP controls and the hockey-stick curve for O,N. These authors restricted their analysis to
13 this case since their stated purpose was only a sensitivity analysis as opposed to developing
14 alternate credible risk estimates. Figure F-4 reproduces (from Crump et al. [2008]) curves of the
15 cumulative probability of a rat dying from a nasal SCC by a given age for bioassay exposure
16 groups of 6, 10, and 15 ppm. For comparison purposes, the corresponding KM (nonparametric)
17 estimates of the probability of death from a nasal tumor are also shown. Three sets of
18 probabilities are graphed: the original unmodified one and the ones obtained by using hockey-
This document is a draft for review purposes only and does not constitute Agency policy.
F-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 stick Mod 1 and Mod 2. Crump et al. (2008) state that the changes in the tumor probability
2 resulting from these modifications are so slight that the three models cannot be readily
3 distinguished in this graph.4 Thus, the modifications considered to the models for the division
4 rates of initiated cells caused an inconsequential change in the fit of the model-predicted tumor
5 incidence to the animal tumor data.
01
0.9n
0.8^
0.6^
0.5-
0.1-
0.0-
15 ppm
0 200 400 600
Time (days)
800
1000
Figure F-4. Very similar model estimates of probability of fatal tumor in
rats for three models in Figure F-2.
Note: The differences are visually indistinguishable. Models were derived from
the implementation of Conolly et al. (2003) with the hockey-stick curves for
ai(flux) and aN(flux) and variants derived from modifications (Mod 1 and Mod 2,
Figure F-2) to ai(flux). Model probabilities are compared to KM estimates. The
three sets of model estimates are so similar that they cannot be distinguished on
this graph.
Source: Crump et al. (2008).
4 The largest change in the tumor probability resulting from this modification for any dose group and any age up
through 900 days was found to be less than 0.002, a change so small that it would be impossible to detect, even in
the largest bioassays ever conducted. The changes in tumor probability resulting from the other modifications
described earlier were found to be even smaller. These comparisons were made in Crump et al. (2008) without re-
optimizing the likelihood. The authors note that re-optimization of the model subsequent to the variations would
have made the fit of modified models even better.
This document is a draft for review purposes only and does not constitute Agency policy.
F-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 The above modifications did not affect the basal rate of cell division in the model and
2 likewise had no effect on the fit to the human background data (Crump et al., 2008).
3 Crump et al. (2008) noted that, although the threshold model for initiated cells in Conolly
4 et al. (2003) was replaced with a model that had a small positive slope at the origin, the resulting
5 curves, hockey-stick Mod 1 and hockey-stick Mod 2, could have been shifted slightly to the right
6 along the flux axis in order to introduce a threshold for a/ without materially affecting the risk
7 estimates resulting from these modified curves. Thus, "the assumption of a linear no-threshold
8 response is not an essential feature of the modifications to the hockey-stick model; clearly
9 threshold models exist that would produce essentially the same effect" (Crump et al. 2008).
10
11 F.2.3. Biological Plausibility of Alternate Assumptions
12 These very small variations made to the 0,1 in Conolly et al. (2003) are seen to be
13
14 • consistent with the tumor-incidence data (see Figure F-4);
15 • small compared with the variability and uncertainty in the cell replication rates
16 characterized from the available empirical data (at the formaldehyde flux where 0,1 was
17 varied);
18 • supported (qualitatively) by limited data, suggesting increased cell proliferation at doses
19 below cytotoxic;
20 • perturbations that one should expect on any dose response derived from laboratory
21 animal data because of human population variability in cell replication;
22 • and biologically plausible because cell cycle control in initiated cells is likely to be
23 disrupted.
24
25 The averaged cell replication rate constants as tabulated in Table 1 of Conolly et al.
26 (2003) and shown by the red curve in Figure E-2 of Appendix E (for various exposure
27 concentrations and corresponding average formaldehyde flux values in the F344 rat nose)
28 demonstrate an increase over baseline values only at exposure concentrations of 6 ppm and
29 higher. Increased cell proliferation at these concentrations of formaldehyde, whether transient or
30 sustained, have been associated in the literature with epithelial response to the cytotoxic
31 properties of formaldehyde (Conolly, 2002; Monticello and Morgan, 1997; Monticello et al.,
32 1996, 1991). The labeling data are considered to show a lack of cytotoxicity and regenerative
33 cell proliferation in the F344 rat at exposures of 2 ppm and below (Conolly, 2002). In the
34 Conolly et al. (2003) modeling, it is further assumed that the formaldehyde flux levels at which
35 cell replication exceeds baseline rates remain essentially unchanged when extrapolated to the
36 human and for initiated cells for the rat as well as the human. These assumptions need to be first
This document is a draft for review purposes only and does not constitute Agency policy.
F-10 DRAFT—DO NOT CITE OR QUOTE
-------
1 viewed in the context of the uncertainty and variability in the data on normal cells discussed in
2 Appendix E.
3 Arguments for a hockey-stick or J shape over the background have been made in the
4 literature for sustained and chronic cell replication rates. However, the analyses of the cell
5 replication data show that the data are not consistently (over each site and time) indicative of a
6 hockey-stick or J shape as the best representation of the data (see Appendix E). This uncertainty
7 is particularly prominent when examining the cell replication data at the 13-week exposure time
8 and the pooled data from the PLM nasal site from Monticello et al. (1996) (see Figures E-l
9 [dotted curve], E-3B, and E-4 of Appendix E). The earliest exposure time in this experiment was
10 at 13 weeks, and the 13-week cell replication data appear to be more representative of a
11 monotonic increasing dose response without a threshold; it is possible that early times are of
12 more relevance to the carcinogenesis as well as for considering typical (frequent short duration)
13 human exposures.
14 Recently, Meng et al. (2010) measured cell replication in the anterior lateral meatus of
15 the F344 rat using continuous labeling on rats exposed to all the concentration levels in the
16 Monticello et al. (1996) experiment. Labeling index (i.e., LI, as opposed to ULLI in the
17 Monticello experiment) was measured as the percentage of BrdU-labeled cells among the total
18 number of cells counted at the nasal site. Their data are reproduced below in Figure F-5, where
19 the asterick denotes the observation of a statistically significant difference from the control
20 group (Dunnett's test,/? < 0.01). These data appear to be consistent with a monotonically
21 increasing dose-response shape for cell replication. Linear regression provided good fits to all of
22 the data (R2 = 0.97) as well as to the subset of the data obtained by deleting the higher dose data
23 at 10 ppm and 15 ppm exposures (R2 = 0.84). We cite these data in support of considering the
24 modifications carried out in Figure F-2.
25 For initiated cells, there are no data on which to evaluate the modifications made in
26 Section F.2.2 to these rates. However, some perspective can be gained by comparing them to the
27 variability in the division rates obtained from the data on normal cells used to construct the
28 formaldehyde model. As shown in Figure E-2 and discussed further in Subramaniam et al.
29 (2008), these data show roughly an order of magnitude variation in the cell replication rate at a
30 given flux. As part of a statistical evaluation of these data, a standard deviation of 0.32 was
31 calculated for the log-transforms of individual measurements of division rates of normal cells
32 (Crump et al., 2008). By comparison, the maximum change in the log-transform division rate of
33 initiated cells resulting from hockey-stick Mod 2 was only 0.20, and the average change would
34 be considerably smaller. Thus, although there are no data for initiated cells, it can be said that
This document is a draft for review purposes only and does not constitute Agency policy.
F-11 DRAFT—DO NOT CITE OR QUOTE
-------
1 the modifications introduced in Crump et al. (2008) for initiated cells are extremely small in
2 comparison to the dispersion in the data for normal cells.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
0 5 10 15 10
Formaldehyde (ppm)
Figure F-5. Cell proliferation data from Meng et al. (2010). The Y-axis
shows the percentage of BrdU-labeled cells among the total number of cells
counted in the ALM section of the rat nose.
Reproduced with permission from Meng et al. (2010).
Subramaniam et al. (2008) also point to some additional, albeit limited, data, suggesting
that exposure to formaldehyde could result in increased cell replication at doses far below those
that are considered to be cytotoxic. Tyihak et al. (2001) treated different human cell lines in
culture to various doses (0.1-10 mM) of formaldehyde and found that the mitotic index
increased at the lowest dose of 0.1 mM. These findings considered along with human population
variability and susceptibility (for example, polymorphisms in ADH3 [Hedberg et al., 2001])
indicate that it is necessary to consider the possibility of small increases in the human a/ over
baseline levels at exposures well below those at which cytotoxicity-driven proliferative response
is thought to occur.
Heck and Casanova (1999) have provided arguments to explain that the formation of
DPXs by formaldehyde leads to inhibition of cell replication (i.e., if this effect alone is
considered, normal cell replication rate of the exposed cells would be less than the baseline rate).
However, this hypothesis was posed for normal cells. Subramaniam et al. (2008) argue that if an
initiated cell is created by a specific mutation that impairs cell cycle control, the effect would be
This document is a draft for review purposes only and does not constitute Agency policy.
F-12 DRAFT—DO NOT CITE OR QUOTE
-------
1 to mitigate the DPX-induced inhibition in cell replication, either partially or fully, depending on
2 the extent to which the cell cycle control has been disrupted. In the absence of data on initiated
3 cells, the above argument provided biological motivation to the modification applied to the
4 J-shape model for cell division (Crump et al. 2008).
5 Thus, the previous paragraphs suggest that the changes made in the analysis in Crump et
6 al. (2008) to the assumption by Conolly et al. (2003) regarding the dose response for the division
7 rate of initiated cells are plausible.
8
9 F.2.4. Effect of Alternate Assumptions for Initiated Cell Kinetics on Human Risk
10 Estimates
11 Figure F-6 contains graphs of the additional human risks estimated (in Crump et al.
12 [2008]) by applying these modified models for a/ and using all NTP controls, compared with
13 those obtained by using the original Conolly et al. (2004) model. Each of the four modified
14 models presents a very different picture from that of Conolly et al. (2004). At low exposures,
15 these risks are three to four orders of magnitude larger than the largest estimates obtained by
16 Conolly et al. (2004).
17
J-Shape Mod 2
18
19
20
21
Conolly et al. (2004)
"Risk Conservative"
Hockey Stick Model
Unmodified
Hockey Stick
Model
1E-3
0.01 0.1
Formaldehyde Exposure (ppm)
Figure F-6. Graphs of the additional human risks estimated by applying
these modified models for 04, using all NTP controls, compared to those
obtained using the original Conolly et al. (2004) model.
Source: Crump et al. (2008).
This document is a draft for review purposes only and does not constitute Agency policy.
F-13 DRAFT—DO NOT CITE OR QUOTE
-------
1 These results have been criticized by Conolly et al. (2009) as being unrealistically large
2 and above the realm of any epidemiologic estimate for formaldehyde SCC. Thus, they argue that
3 the parameter adjustments made in Crump et al. (2008) are inappropriate. Crump et al. (2009)
4 rebutted these points by arguing that the purpose of their work was not to provide a more reliable
5 or plausible model but to carry out a sensitivity analysis. They argued that the changes made to
6 the model (in their analyses) were reasonable since they did not violate any biological
7 constraints or the available data. Further, they pointed out that "by appropriately mitigating the
8 small modifications [they] made to the division rates of initiated cells, the model [would]
9 provide any desired risk ranging from that estimated by the original model up to risks 1,000-fold
10 larger than the conservative estimate in Conolly et al. (2004)."
11 Crump et al. (2008) also evaluated the assumption in eq D-3 of the CUT modeling
12 pertaining to initiated cell death rates (/?/) by making small changes to /?/. They report that they
13 obtained similarly large values for estimates of additional human risk at low exposures.
14 Obtaining reliable data on cell death rates in the nasal epithelium appears to be an unusually
15 difficult proposition (Hester et al., 2003; Monticello and Morgan, 1997), and, even if data are
16 obtained, they are likely to be extremely variable.
This document is a draft for review purposes only and does not constitute Agency policy.
F-14 DRAFT—DO NOT CITE OR QUOTE
-------
Appendix G
-------
1 APPENDIX G
2
3 EVALUATION OF THE CANCER DOSE-RESPONSE MODELING
4 OF GENOMIC DATA FOR FORMALDEHYDE RISK ASSESSMENT
5
6
7 G.I. MAJOR CONCLUSIONS IN ANDERSEN ET AL. (2008)
8 In Chapter 4, the gene microarray data from animal studies on formaldehyde (Andersen
9 et al., 2008; Thomas et al., 2007) were described. The analysis of these animal high throughput
10 data and the conclusions reached in these two groundbreaking papers were closely examined for
11 use in this assessment. Studies on high throughput animal data provide a wealth of information
12 that helps further understanding of the relevant mechanisms. However, such studies have
13 generally not made quantitative bottom-line inferences that inform low dose human risk. The
14 above-mentioned studies are a notable exception due to the breadth of their conclusions on low
15 dose MO As, their pioneering application of the benchmark dose (BMD) methodology to
16 genomic data, their use of BMD-response analysis that identified dose estimates at which
17 specific cellular processes were significantly altered, and the fact that they were accompanied by
18 recommendation in the literature urging use of these results in setting exposure standards for
19 formaldehyde (Daston, 2008).
20 We focus here on the conclusions in these papers with regard to modeling the cancer
21 dose-response for formaldehyde. In addition to supporting our disposition of these analyses for
22 this assessment, this write-up serves the purpose of exemplifying critical issues that need to be
23 considered for the future.
24 The overall BMD determined in Andersen et al. (2008) for all genes with significant
25 dose-response averaged 6.4 ppm. These analyses indicated a general progression with the lowest
26 BMD values (i.e., the most sensitive epithelial responses) for extracellular and cell membrane
27 components and higher BMD values for intracellular processes. Overall, these authors
28 concluded that
29
30 • Genomic changes, including those suggestive of mutagenic effects, did not temporally
31 precede or occur at lower doses than phenotypic changes in the tissue
32 • Genomic changes were no more sensitive than tissue responses
33 • Formaldehyde, being an endogenous chemical, is well handled until some threshold is
34 achieved. Above these doses, toxicity rapidly ensues with concomitant genomic and
35 histologic changes.
36 • Linear extrapolations, or extrapolations that specify similar MO As at high and low doses
37 would be inappropriate.
This document is a draft for review purposes only and does not constitute Agency policy.
G-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 These findings were judged to have significant implications on the debated MOA for
2 formaldehyde carcinogenicity, confirming results from earlier bioassays and dose-response
3 modeling that the mutagenicity of formaldehyde was too weak to be of relevance to its
4 carcinogenicity. Daston (2008) judged the method in these efforts to be extremely sensitive and
5 therefore suited to examining whether responses at the molecular level take place at doses below
6 which frank adverse effects occur. Daston (2008) argued that"... if there are pleiotropic effects
7 at lower exposure levels that would elicit a different profile of gene expression, those genes
8 would not go unnoticed" and thus concluded that "the gene expression data confirm that the
9 responses are not linear at low doses."
10 In the analyses that follow, we point to some significant quantitative factors that impact
11 on these conclusions.
12
13 G.2. USE OF MULTIPLE FILTERS ON THE DATA
14 The analyses in these papers involved the following sequence of data filters.
15
16 1. Gene probe sets that differed in expression in response to treatment were identified by
17 one-way analysis of variance. Probability values were adjusted for multiple comparisons
18 by using a false discovery rate of 5%.
19 2. Next, in addition to the above statistical filter, the output was further screened by
20 selecting only those genes that exhibited a change from the control group that was greater
21 than or equal to 1.5-fold (logarithmic).
22 3. The gene probe sets that demonstrated significant dose-response behavior were then
23 matched to their corresponding biological process and molecular function gene ontology
24 (GO) categories (considering only those involving more than three genes) and grouped
25 into process categories such as cell division, DNA repair, cellular proliferation,
26 apoptosis, and related molecular function categories.
27
28 A large number of genes are expressed in these studies; therefore, clearly some
29 appropriate filter needs to be used for meaningful interpretation of the vast database. Tissue
30 pathology served as a phenotypic anchor for the interpretation of microarray results, and the
31 genomic study confirmed (and improved on) the qualitative and quantitative understanding
32 derived from the histopathology and observation of frank effects. It is possible that the
33 combination of filters used by these authors is adequate for an inquiry into some mechanisms
34 associated with the specific phenotypic effects. However, the studies reached bottom-line
35 conclusions with regard to the low-dose MOA and approach to be considered for quantitative
36 extrapolation. These conclusions necessarily involve questions as to whether there were gene
This document is a draft for review purposes only and does not constitute Agency policy.
G-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 expression changes at low dose and at early exposure times that may be relevant to initiating
2 carcinogenesis and finally as to whether there is a threshold in dose associated with
3 formaldehyde carcinogenesis. However, collectively, the three filters employed in these studies
4 likely constitute overly stringent criteria, taking away the resolution needed to observe critical
5 gene changes needed to delineate low dose effects. An indication that this may indeed be the
6 case can be seen by examining the correlations in their findings with the observed trend in the
7 data on DPXs formed by formaldehyde. This is detailed in the following section.
8
9 G.3. DAT A FOR LOW-DOSE CANCER RESPONSE
10 A significant finding in Thomas et al. (2007) is that BMD estimates for the GO
11 categories applicable to cell proliferation and DNA damage were similar to values obtained for
12 cell labeling indices and DPXs in earlier studies and to BMD estimates obtained for the onset of
13 nasal tumors. The mean BMD for the GO category of "positive regulation of cell proliferation"
14 was 5.7 ppm; in comparison, Schlosser et al. (2003) obtained a 10% BMD of 4.9 ppm for the cell
15 labeling index. The GO category associated with "response to DNA damage stimulus," seen as a
16 genomic correlate to a mutagenic effect, had a mean BMD of 6.31 ppm. Thomas et al. (2007)
17 compare this finding with significant increase at 6 ppm of DPXs following a 3-hour exposure in
18 the study by Casanova et al. (1994). The formation and repair of DPXs have been considered to
19 be one of the potential mechanisms associated with the genotoxic action of formaldehyde
20 (Conolly et al., 2003, 2000). Based on earlier work in the same laboratory (Conolly et al., 2004,
21 2003; Conolly, 2002), Slikker et al. (2004) concluded that there is a dose threshold (at about
22 6 ppm) to formaldehyde carcinogen!city and that the putative mutagenic action of formaldehyde
23 is not relevant to its carcinogenicity. Therefore, the finding that a significant genomic response
24 (e.g., induction of DNA repair genes) is not observed at doses lower than those that induce
25 tumors in rodent bioassays is seen by these authors (Andersen et al., 2008; Daston, 2008;
26 Thomas et al., 2007) to further buttress the above conclusions related to the mode of action for
27 formaldehyde-induced respiratory cancer.
28 However, phenotypic anchoring to the DPX data drawn only from Casanova et al. (1994)
29 misses critical low-dose data that informs mode of action. In an earlier study, Casanova et al.
30 (1989) observed statistically significantly elevated (over controls) levels of DPXs at 2 ppm and a
31 trend towards elevated DPXs at 0.7 ppm. In analysis of low-dose data, the trend in the dose-
32 response is critically important because data inherently lack the power to establish statistical
33 significance. Furthermore, the two studies by Casanova and coworkers are different in some
34 respects. The earlier study was a 6-hour exposure, while the later study was a 3-hour study; thus,
35 on this account alone, it appears more relevant to compare with the older study. Exposures in
This document is a draft for review purposes only and does not constitute Agency policy.
G-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 the earlier study were additionally at 0.3 and 10 ppm, thus affording a lower exposure
2 concentration. In the earlier study, tissue from the whole nose was analyzed, whereas in the later
3 study tissue from two specific regions was obtained from the "high" tumor (Level II) and "low"
4 tumor regions. Together, these data suggest that DPXs occur at exposure concentrations
5 considerably lower than those that elicited transcriptional changes. One possible explanation is
6 that the increase in DPXs was not sufficient to induce DNA repair genes. Alternatively, these
7 discrepancies may be due to the stringent filters and the low statistical power of the Andersen et
8 al. (2008) study. These disparities between the gene array study and the DPXs question the
9 ability of the studies in Andersen et al. (2008) and Thomas et al. (2007) to inform the presence or
10 absence of a mutational MOA for formaldehyde, and in essence, to inform the low-dose response
11 curve for formaldehyde-induced cancer.
12 In another instance, Andersen et al. (2008) clearly stated that no genes were significantly
13 altered by exposure to 0.7 ppm, yet they state that there was "a trend toward altered expression at
14 0.7 ppm" in some genes with U and inverted U shape dose-responses (Figures 4 and 5 of their
15 paper). While these changes may not be statistically significant, they could be biologically
16 significant.
17
18 G.4. DIFFICULTIES IN INTERPRETING THE BENCHMARK MODELING
19 The benchmark analyses are summarized in Thomas et al. (2007) as average BMD
20 estimates for genes in a given GO that were statistically significantly dose related. The
21 benchmark modeling was then used by the authors to identify that the dose below individual
22 cellular processes was judged to be "not altered."
23 The BMD definition used by these authors is quite stringent: it defines an effect so that
24 only 0.005 of controls will be considered affected and sets the BMR corresponding to this dose
25 at 0.105. The net effect is that the BMD is the air level, such that the increase in the mean
26 response is 1.349 x standard deviation. This is essentially an arbitrary definition. For
27 comparison, if 0.05 of controls are considered affected and the BMR is set at 0.1 (common
28 values that are applied to whole animal data), the BMD is the air level such that the increase in
29 the mean response is 0.608 x standard deviation. Thus, if this definition had been used (as is
30 traditionally the case), the BMD estimates would all be 2.2 times smaller than those obtained by
31 Schlosser et al. (2003). Furthermore, the analysis assumes equal variance in all dose groups.
32 Thus, further consideration of these issues with regard to interpretation of the BMR obtained
33 from these studies is needed before it can be used in regulatory exposure setting. Secondly,
34 lower confidence limits on the BMDs need to be derived for the data in Andersen et al. (2008).
35
This document is a draft for review purposes only and does not constitute Agency policy.
G-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 G.5. STATISTICAL SENSITIVITY OF THE DATA FOR DOSE-RESPONSE
2 Another cautionary note pertains to the qualification of gene array studies as being
3 extremely sensitive. Such a qualification should actually refer to the fact that only tiny amounts
4 of mRNA are needed, that is, the sensitivity of the assay per se for measuring gene expression.
5 However, this should not be confused with the sensitivity needed to identify the very small dose-
6 related changes at low dose. Andersen et al. (2008) reports on results of studies that involve
7 small numbers of animals in each dose group (five or eight). Despite the limited power in such
8 studies, the paper equates the absence of a statistically significant effect with no effect. This
9 limitation is generally true of studies of the dose responses of changes in gene expression
10 conducted to date; they have generally relied on very few animals (<10 per dose group). Since
11 there will likely always be background amounts of gene expression, quantifying the dose
12 response requires statistically significant changes in gene expression as a function of dose. If the
13 genomic data involve even fewer animals per group than the histopathological data, they have
14 even less power to delineate the dose response; in particular, whether there is a threshold at low
15 exposures. This is illustrated by the example in Figure G-l of the dose responses for epithelial
16 hyperplasia. The data in this figure are from lesion 2 in Andersen et al. (2008); the linear
17 regressions and confidence limits were determined by EPA These appear equally consistent
18 with both a threshold at around 1 ppm and a linear response down to zero.
19
20 G.6. LENGTH OF THE STUDY AND STOCHASTIC EVENTS
21 Another significant consideration with regard to MO A conclusions that are pertinent to
22 the disease process is the length of the study, 15 days. If formaldehyde-induced tumor formation
23 is a stochastic process (e.g., genotoxicity), then exposure of a small number of animals to low
24 concentrations for 15 days may not be long enough to detect changes that might occur under
25 long-term exposure scenarios.
26 Relatedly, it has been suggested that gene (and protein) expression is a stochastic process
27 whereby steady state gene expression obeys Poisson statistics (i.e., distribution of rare events),
28 and that events of interest may occur in a single cell or small number of cells in which larger
29 tissue samples can average out such stochastic events and prevent the detection of nonaverage
30 behavior (Quakenbush, 2007). Given the implied difficulty in such an analysis, duration of
31 exposure may be one of the most tenable ways of addressing whether a chemical increases the
32 probability of an adverse response.
This document is a draft for review purposes only and does not constitute Agency policy.
G-6 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
Dayl
01234567
Day 1 Recovery
Day5
Day 6
Day 6 Recovery
Day 15
Figure G-l. Graphs of epithelial hyperplasia (Lesion 2) versus formaldehyde
concentration (ppm) with 95% confidence intervals (with linear fit by eye).
Source: Fit to data from Andersen et al. (2008).
3 G.7. OVERALL CONCLUSION
4 We believe our analyses of the presentations in Andersen et al. (2008) and Daston (2008)
5 are generally useful with regard to future developments in quantitative analyses of genomic data
6 if they are to be of relevance to risk assessment. For risk assessment, rather than focusing on
7 what responses are statistically significant, an analysis should focus on (1) what range of values
8 of critical parameters (e.g., gene expression) are consistent with the data, and (2) what these
9 values imply for whole animal risk. This is of course, an extremely difficult proposition because
This document is a draft for review purposes only and does not constitute Agency policy.
G-l DRAFT—DO NOT CITE OR QUOTE
-------
1 we do not know nearly enough about how changes in genes quantitatively affect whole animal
2 risk, or even which genes are important.
3
4
This document is a draft for review purposes only and does not constitute Agency policy.
G-8 DRAFT—DO NOT CITE OR QUOTE
-------
Appendix H
-------
1 APPENDIX H
2
3 EXPERT PANEL CONSULTATION ON QUANTITATIVE EVALUATION OF
4 ANIMAL TOXICOLOGY DATA FOR ANALYZING CANCER RISK DUE TO
5 INHALED FORM ALDEHYDE
6
7
8 The National Center for Environmental Assessment convened an expert panel of
9 scientists for advice on evaluating available approaches for incorporating biological information
10 in analyzing animal tumor data for assessing cancer risk due to inhaled formaldehyde. This
11 Appendix pertains to the major deliberations and results of that meeting and is divided into three
12 sections.
13
14 A. Scope and Agenda of Meeting on Quantitative Evaluation of Animal Toxicology Data for
15 Analyzing Cancer Risk due to Inhaled Formaldehyde. October 28 & 29, 2004.
16 B. Summary of Consultative Meeting on CUT Formaldehyde Model. October 28 & 29,
17 2004.
18 C. Meeting Report from Dr. Rory B. Conolly
19
20
This document is a draft for review purposes only and does not constitute Agency policy.
H-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 A. Scope and Agenda of Meeting on Quantitative Evaluation of Animal Toxicology Data
2 for Analyzing Cancer Risk due to Inhaled Formaldehyde
3 October 28 & 29, 2004. Washington, DC.
4
5 This meeting is to assist EPA in evaluating available approaches for incorporating biological
6 information in analyzing animal tumor data for assessing cancer risk due to inhaled
7 formaldehyde. The CUT Centers for Health Research (CUT) has published a novel risk
8 assessment that links site-specific predictions of flux using computational fluid dynamics (CFD)
9 modeling with a two-stage clonal growth model of cancer to analyze nasal tumor incidence in
10 two rodent bioassays. The rodent models are used with corresponding human models for low-
11 dose extrapolation of cancer risk to people.
12
13 Key predictions of the CUT effort are a zero maximum likelihood estimate of the probability of
14 formaldehyde-induced mutation per cell generation in the rat and a de minimus additional
15 lifetime risk in nonsmokers due to continuous environmental exposure below 0.2 ppm. The
16 National Center for Environmental Assessment is carrying out sensitivity analyses and
17 examining variations of the CUT model in order to understand the implications of the model
18 structure and parameters on model predictions. In this meeting, we wish to focus on the
19 strengths and key uncertainties of this model, the extent to which assumptions in the CUT model
20 are supported by biological data, and examine the impact of uncertainty and variability on the
21 overall quantitative risk characterization.
22
23 Broadly, the discussions will focus on the following areas:
24
25 • Impact of uncertainties in dosimetry on human risk estimates
26 • Uncertainties in the use of experimental data on labeling index
27 • The model structure related to initiated cells and DNA protein cross-links
28 • Considerations of time-to-tumor in the clonal growth modeling
29 • Inferences and information on the role of mutation and cytotoxicity in estimating human risk
30 • Relative merits of benchmark dose modeling vs. the 2-stage clonal growth model
31
32 Discussions on Mode of Action are expected to be an integral part of several of the sessions.
33 Therefore a specific time-slot is not set aside for this purpose.
34
35 The meeting will have a panel discussion format. There will be no formal presentations unless
36 necessary to elucidate an issue. Various attachments referred to in the Agenda below, as well as
37 the relevant manuscripts will be sent separately.
38
39 Specifically, we suggest the following issues upon which to focus the discussion in the above
40 areas, and approximate time frames and discussion leads, although discussants should feel free to
41 bring up other critical issues.
42
43
This document is a draft for review purposes only and does not constitute Agency policy.
H-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 I. Introduction and purpose of discussion
2 Peter Preuss.
3 9:OOAM,Oct28
4
5 II. Impact of uncertainties in dosimetry on risk estimates
6 Lead discussant: Linda Hanna
1 9:15-11 AMOct 28
8
9 Boundary conditions
10 The CFD modeling specified a mass transfer coefficient as a boundary condition on the
11 nasal lining, adjusting the value of this coefficient on the "absorbing" portion of the
12 lining so as to match simulated overall uptake in the rat nose to the experimentally
13 determined average overall uptake. This value was then used for the corresponding
14 human nasal lining. Are these boundary conditions appropriate surrogates for the
15 underlying pharmacokinetics, including saturation in metabolism and mucociliary
16 clearance, particularly with reference to humans?
17
18 Turbulence
19 Turbulent flow has been seen to occur in experimental models of the human nose at some
20 of the higher flow rates at which the CFD models were used in CIIT's assessment. It is
21 not likely that the CUT CFD model can reliably identify signatures of transition to
22 turbulent behavior. Turbulent flow can significantly alter regional uptake patterns.
23 Additionally, significant mass balance errors were seen at the higher flow rates in the
24 human flow models. Discuss if these are likely to impact significantly on risk estimates.
25
26 Interindividual variability
27 The CUT assessment has focused on the nasal anatomy of a single individual. Discuss
28 the implications of interindividual variations in nasal anatomy on the population
29 distribution in risk.
30
31 III. Uncertainties in the use of experimental data on labeling index
32 Lead discussant: George Lucier
33 11AM- 11:45 AM, 1:00 - 3:15PMOct 28
34
35 Cell-replication rate and its relationship to flux is a critical determinant of risk. Therefore
36 uncertainties and variability in measurement of the unit length labeling index and its use in the
37 CUT clonal growth modeling need to be characterized.
38
39 1. Discuss the strengths, uncertainties and limitations associated with estimating cell
40 replication rates from the unit length labeling index (ULLI).
41 a. For example, a constant ratio of the measured ULLI to the labeling index (LI) that
42 is used in the model is assumed. Is it valid to assume this ratio to be constant
43 across nasal sites, dose and exposure time.
44 b. How uncertain is this ratio?
45
This document is a draft for review purposes only and does not constitute Agency policy.
H-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 2. Considering the large patterns of variability in the ULLI data, discuss the validity of
2 using ULLI averaged over site and exposure times.
3 a. The averaging loses information on the sequential effect of change with time, and
4 on significant differences among sites.
5 b. How sensitive is the clonal growth modeling result to these variations in the dose-
6 response function for cell replication rates vs. flux to the tissue? A discussion of
7 this question in this session is intended to serve as input to later deliberations on
8 the issue.
9
10 3. Discuss the validity of combining data collected in different experiments using different
11 labeling methods, and the validity of estimating cell replication rates from LI or ULLI
12 measured in a single pulse labeling experiment.
13
14 See attachment C: "ULLI Dose-Response Modeling and Statistical Analysis "for a
15 discussion of these issues, and Moolgavkar andLuebeck (1992).
16
17 IV Model Structure: Birth and death rates for Initiated cells, Role of DPX
18 Lead discussant: Kenny Crump
19 3:30- 6:00 PMOct 28.
20
21 Parameters for initiated cells
22
23 1. The CUT analysis of ULLI data allows for a virtual threshold in dose in the replication
24 rate of normal cells. Discuss the validity of ascribing such a behavior to initiated cells
25 considering the sensitivity of 2-stage model results to the initiated cell replication rates.
26
27 2. Discuss the treatment of death rate for initiated cells in the model (set equal to birth rate
28 of normal cells in Conolly et al., 2003) and implications for confidence in model
29 predictions.
30
31 Also see Attachment A (memo from Rory Conolly) and Attachment D (EPA discussion of
32 CUT clonal growth modeling and some sensitivity analyses. . .)
33
34 Treatment ofDNA protein cross-links (DPX) in clonal expansion model
35
36 3. Formaldehyde-induced mutation is modeled as taking place only while DPX are in place
37 with DPX undergoing rapid repair. Discuss the possibility of persistent genetic damage
38 that extends beyond the DPX half-life and enhances mutation. How might this issue be
39 included in the model structure?
40
41
This document is a draft for review purposes only and does not constitute Agency policy.
H-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 V. Considerations of time-to-tumor in the CUT clonal growth modeling
2 Lead discussant: Christopher Portier
3 8:30- 11:00 AM, Oct 29.
4
5 1. A number of issues affect likelihood values and the model fit to the time-to-tumor data.
6 Discuss assumptions in the treatment of time-to-tumor in the CUT clonal expansion
7 model, and their impact on parameter estimates. For example,
8 a. Results in Conolly et al. (2003, 2004) are derived considering all tumors to be
9 fatal. Note in this context that serially sacrificed animals have been combined
10 with those experiencing mortality—the effect of this is visible as irregularities in
11 the time-to-tumor curve.
12 b. How is the time variability in ULLI likely to impact on the time-to-tumor
13 predictions?
14
15 2. Long delay times are predicted by the model for observation of detectable tumor. Is this
16 compatible with the assumption of rapidly fatal tumors?
17
18 3. Discuss the weight to be given to differences in likelihood when comparing with
19 variations on the Conolly et al. (2003) model structure such as in Attachment A or D.
20
21 VI. Inferences on the role of formaldehyde-induced mutation and cell proliferation
22 Lead discussant: Dale Hattis
23 77:75 - 12:00 PM, 1:00 - 4:00 PM, Oct 29.
24
25 1. The model structure in Conolly et al. (2003) predicts a zero maximum likelihood estimate
26 for the constant of proportionality (KMU) linking DPX to the probability of
27 formaldehyde-induced mutation per cell generation. Examine the strength of this
28 conclusion, and the extent to which an insignificant probability of formaldehyde-induced
29 mutation per cell generation is supported by data.
30
31 2. Discuss the biological relevance and validity of model-estimated parameters, particularly
32 in the context of low-dose predictions.
33 a. Discuss possible avenues to validate CUT cancer model predictions.
34
35 3. Discuss the validity of using cell replication rates determined for the rat to predict human
36 risk in a population.
37
38 4. In the face of uncertainties, are the results in Conolly et al. (2003, 2004) conservative in
39 the sense of overpredicting risk?
40 a. Discuss the extent to which sensitivity analyses have addressed this issue and the
41 extent to which sensitivity analyses can speak to the strength of the model. [See
42 Attachments A: Memo from Conolly, andD: EPA discussion of CUT clonal
43 growth modeling and some sensitivity analyses... .7
44
This document is a draft for review purposes only and does not constitute Agency policy.
H-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 VII. Benchmark Dose Modeling
2 Lead discussant: Kenny Crump
3 4:15-5:30PM,Oct29.
4
5 Discuss the relative merits of using a benchmark dose approach that incorporates
6 biological modeling (such as estimating flux to tissue or DPX levels) as compared with
7 the CUT 2-stage model for cancer. (See attachment E and Schlosser et al, 2003.)
This document is a draft for review purposes only and does not constitute Agency policy.
H-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 B. Summary of Consultative Meeting on CUT Formaldehyde Model
2 October 28 & 29 2004, NCEA, Washington, DC
O
4 Date: November 10, 2004
5 Ravi P. Subramaniam, Ph.D.
6 Quantitative Risk Methods Group
7 National Center for Environmental Assessment, ORD, US EPA
8
9 This is a broad summary of the most important issues at the formaldehyde meeting.
10 It was generally felt by consultants that the broad framework of the approach adopted by
11 CUT, namely the use of a two-stage model for cancer, the linking of localized flux to cell
12 replication rates and DPX concentration, and the expression of formaldehyde-induced mutation
13 as a linear function of DPX, was reasonable.
14 Potential errors in the dosimetry modeling were seen not to have a significant effect on
15 risk estimates. The boundary conditions used were discussed to be a reasonable representation
16 of the pharmacokinetics for both rats and humans. The discussion on the impact of
17 interindividual variability of nasal anatomy was not particularly conclusive. It was determined
18 that there was likely to be much less variability in reactive gas uptake than that seen in
19 parti culates.
20 Crucial errors were however identified on several fronts in the manner in which the
21 clonal growth model had been implemented in the CUT effort. Dr. Portier felt that the
22 calculation of probability was seriously flawed on account of lumping serially-sacrificed animals
23 and animals that died of tumor together, while at the same time assuming rapid fatality of all
24 tumors. This was seen to significantly alter the calculation of tumor probability (the shape of the
25 dose-response curve), and his insight was that a correction was likely to allow for a substantially
26 higher value for the probability of formaldehyde-induced mutation at low-dose. The best
27 estimate for this probability is now zero in the model. Drs. Crump, Portier and Hattis argued that
28 replacing this estimate by an upper confidence bound on KMU (the coefficient determining the
29 role of DPX in the probability of mutation per cell generation), keeping other structural problems
30 in the model unexplored, or other parameters fixed, would not be enough. There was a
31 discussion on the need to provide confidence bounds on risk determined by allowing all the
32 parameters to vary. Drs. Crump and Hattis (and Portier?) felt such an estimate would be very
33 different from that calculated based on individual parameters.
34 Drs. Crump, Hattis and Portier urged us not to be constrained by the optimal likelihood
35 values of a single plausible model, and underscored the need to explore a variety of biologically
36 reasonable model structures as a requisite for utilizing such a model in risk assessment.
This document is a draft for review purposes only and does not constitute Agency policy.
H-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 Likelihood was seen to be an inadequate expression of what is to be considered an optimal
2 model (okay only for comparing models that were nested, etc.). These models should allow the
3 expression of variability and uncertainty in the data, as well as in underlying assumptions in
4 model specification. Dr. Crump (and Hattis also?) felt that alternate model structures, if
5 explored, could potentially lead to risk estimates, for the range below the observed data, that
6 were higher by several thousands.
7 Dr. Crump cautioned that extrapolating to human using the hockey or J-shaped cell
8 replication curve used in the rodent carried with it a large uncertainty that had not been
9 characterized in the Conolly modeling.
10 Dr. Portier expressed concern over the manner in which historical and concurrent
11 controls were lumped together. The thrust of Portier's comments was that such a combination of
12 controls was generally not done. The large number of historical controls was likely to
13 significantly bias the impact of the bioassay data in determining the time-to-tumor fits.
14 There were various discussions about the pros and cons of constructing a joint likelihood
15 of the cell replication data and the tumor data, and the weights to be assigned to the separate
16 likelihoods. This was considered to be problematic by Dr. Portier.
17 Dr. Crump's opinion was that the Conolly model, and those explored by EPA, fit the
18 tumor data poorly, and that an improved description of the tumor data was needed before the
19 model could be used for low-dose and interspecies extrapolation.
20 Drs. Lucier and Hattis placed emphasis on including the early-time cell replication data
21 instead of constructing a time-weighted average. It was felt that the two Monticello experiments
22 could not be combined together as in Conolly et al. Dr. Lucier felt that the early-time data would
23 have a greater impact in the progression of carcinogenesis. In general, the effect of "time" was
24 considered to have significant effects on the time-to-tumor modeling, and they urged us to
25 incorporate time-dependent terms in the modeling. CUT expressed willingness to provide the
26 original cell replication data to us for further analysis. (Further discussion on this matter did not
27 take place in the open forum.)
28 Preliminary indications are, particularly based on Dr. Portier's insight, that the currently-
29 held "de-minimus" picture of low-dose risk, as expressed in Conolly et al. (2004), is not likely to
30 be the case if these various suggestions are incorporated in the modeling.
31
This document is a draft for review purposes only and does not constitute Agency policy.
H-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 C. Meeting Report from Dr. Rory B. Conolly
2
3 Rory B. Conolly, Sc.D., D.A.B.T.
4 106 Michael's Way
5 Chapel Hill, NC 27516
6 Voice: 919.929.2258
7
8 July 24, 2005
9
10 Dr. Bobette Norse
11 ORAU Procurement - MS-04
12 P.O. Box 117
13 Oak Ridge, TN 37831-0117
14 Phone: 865-576-3051
15 Fax: 865-576-9385
16
17 Dear Dr. Nourse,
18
19 The following is my final written report on the formaldehyde review meeting held at the
20 U.S. EPA in Washington, D.C. on 28-29 October, 2004.
21 EPA provided no guiding philosophical statement about the criteria being used to
22 evaluate the CUT assessment. The new Guidelines for Carcinogen Assessment state that the
23 preferred default approach is to use a biologically based model. Since the key components of the
24 CUT assessment have been published in the peer-reviewed literature and have undergone several
25 peer reviews other than the current NCEA effort, one has to wonder just how high the bar is set
26 for acceptance of biologically based assessments. Given the time and resources expended on the
27 CUT assessment and the richness of the supporting data base, I find it difficult to imagine what
28 an acceptable biologically-based assessment might look like if in the end the CUT assessment is
29 deemed not acceptable by NCEA. If this is in fact the outcome it will have major implications
30 for the likelihood that anyone will be willing to commit the significant resources needed to
31 develop of these kinds of risk assessment models.
32 The documents provided in advance of the October 2004 review meeting were
33 collectively a discussion of uncertainty about the CUT work. With respect to the clonal growth
34 model, however, no new risk predictions were provided, so there was no way to judge how the
35 uncertainties that NCEA identified might impact predicted risk. Evaluation of the significance
36 of "uncertainties" when the impact of the uncertainties on the predicted risk is not known is itself
37 an uncertain process.
38 A related concern is that there did not seem to be any consideration of the historical
39 context of the CUT assessment. EPA developed formaldehyde assessments in 1987 and 1991.
40 The 1987 assessment used ppm as the input and the LMS model for the dose-response
41 prediction. The 1991 assessment used DPX as a dosimeter and the LMS model. BMD
42 assessments have since become available from other sources such as Paul Schlosser's work. The
43 risk predictions of the BMD models are similar to the 1991 LMS assessment. Both the DPX-
44 LMS and BMD assessments predicted somewhat less risk than the 1987 assessment, establishing
45 the trend of less risk with increased incorporation of relevant data. I have always argued
This document is a draft for review purposes only and does not constitute Agency policy.
H-10 DRAFT—DO NOT CITE OR QUOTE
-------
1 (probably initially at the 1998 Ottawa review) that the historical context is the appropriate
2 context for evaluating the CUT clonal growth model. For a "level playing field" the
3 uncertainties of the 1987 and 1991 assessments, and of the more recent BMD models, should be
4 analyzed to the same degree as the clonal growth model. Does NCEA think that, because the
5 LMS and BMD approaches used structurally simpler dose-response models and much more
6 limited data inputs, they are less uncertain? The NCEA analysis seemed to be implying that use
7 of more data and of a biologically more realistic model structure actually makes the CUT
8 approach more uncertain than the LMS and BMD approaches. I encourage NCEA to consider
9 how uncertainties that can be evaluated explicitly in the structurally rich CUT model compare to
10 hidden uncertainties in the simpler models, where the hidden uncertainties encompass, for
11 example:
12
13 1. Missing or incomplete descriptions of the regional dosimetry of formaldehyde.
14 2. Lack of simultaneous incorporation of the directly mutagenic and
15 cytolethal/regenerative proliferation modes of action.
16 3. Lack of explicit consideration of the multistage nature of cancer.
17 4. Lack of consideration of the growth kinetics of initiated cell populations
18 5. Lack of evaluation of the measured J-shaped dose response for regenerative cellular
19 proliferation.
20
21 A careful, balanced comparison of the CUT assessment with the previous assessments along
22 these lines would be informative with respect to the suitability of the CUT assessment as the
23 basis for a new IRIS listing for formaldehyde.
24 A further concern involves the peer-review of the CUT formaldehyde assessment held in
25 Ottawa in 1998. This review was sponsored by the U.S. EPA and Health Canada and involved
26 what was arguably a world-class review panel. The CUT assessment was not in its final form at
27 that time, though we did provide a detailed description of the overall approach and the specific
28 methods we were using to generate dose-response predictions. The 1999 CUT document and the
29 subsequent peer-reviewed publications are responsive to the comments and suggestions raised by
30 the reviewers. My concern is that no information was provided on the role that Ottawa review
31 plays in the ongoing review of the CUT formaldehyde assessment by NCEA. Should the
32 October 2004 review be viewed as standing on the shoulders of the 1998 review or as being in
33 parallel to it? It was not at all clear to me that the October 2004 review in any way utilized the
34 judgments of the 1998 review. It seems that the 2004 review was more of a parallel effort and
35 that the 1998 review was ignored and was effectively a waste of time and money. I would like to
36 have some clear understanding of how the 2004 review effort should be viewed relative to that of
37 1998.
38 In closing, let me reiterate that while the detailed examination of the CUT formaldehyde
39 assessment is laudable, this examination should be conduced with an eye to the historical context
40 of formaldehyde risk assessment on the one had and, on the other hand, to a concern for
41 encouraging, and not discouraging, development of biologically based risk assessment models.
42
43 Sincerely yours,
44
45 Rory B. Conolly, Sc.D., D.A.B.T.
This document is a draft for review purposes only and does not constitute Agency policy.
H-11 DRAFT—DO NOT CITE OR QUOTE
-------
-End of Volume IV-
This document is a draft for review purposes only and does not constitute Agency policy.
H-12 DRAFT—DO NOT CITE OR QUOTE
------- |