United States
Environmental Protection
Agency
Hesnarch and Development
Office of Health and
Environmental Assessment
Washington DC 20460
EPA-600 8-84 '004F
September 1 985
Final Report
&EPA
Health Assessment
Document for
Chloroform
-------
-------
-------
PREFACE
The Office of Health and Environmental Assessment has prepared this
health assessment to serve as a "source document" for EPA use. This health
assessment document was developed for use by the Office of Air Quality
Planning and Standards to support decision-making regarding possible
regulation of chloroform as a hazardous air pollutant. However the scope of
this document has since been expanded to address multimedia aspects.
In the development of the assessment document, the scientific literature
has been inventoried, key studies have been evaluated and summary/conclusions
have been prepared in order to quantitatively identify the toxicity of
chloroform and related characteristics. Observed effect levels and other
measures of dose-response relationships are discussed, where appropriate, to
-------
TABLE OF CONTENTS
LIST OF TABLES x
LIST OF FIGURES xiv
AUTHORS, CONTRIBUTORS, AND REVIEWERS xvi
1. SUMMARY AND CONCLUSIONS 1-1
1.1. INTRODUCTION 1-1
1.2. PHYSICAL AND CHEMICAL PROPERTIES, AND ANALYSIS 1-2
1.3. PHARMACOKINETICS 1-2
1.4. HEALTH EFFECTS OVERVIEW 1-5
1.4.1 Toxicity 1-5
1.4.2 Reproductive Effects 1-7
1.4.3 Mutagenicity 1-7
1.4.4 Carcinogenicity 1-9
1.4.5 Quantitative Risk Assessment 1-12
2. INTRODUCTION 2-1
3. BACKGROUND INFORMATION 3-1
3.1. INTRODUCTION 3-1
3.2. PHYSICAL AND CHEMICAL PROPERTIES 3-2
3.3. SAMPLING AND ANALYSIS 3-4
3.3.1. Chloroform in Air 3-4
3.3.2. Chloroform in Water 3-5
3.3.3. Chloroform in Blood 3-6
3.3.4. Chloroform in Urine 3-6
3.3.5. Chloroform in Tissue 3-6
3.4. EMISSIONS FROM PRODUCTION AND USE 3-7
3.4.1. Emissions from Production 3-7
3.4.1.1. Direct Production 3-7
3.4.1.2. Indirect Production 3-12
3.4.2. Emissions from Use 3-20
3.4.2.1. Emissions from Pharmaceutical
Manufacturing 3-21
3.4.2.2. Emissions from Fluorocarbon-22
Production 3-22
3.4.2.3. Emissions from Hypalon®
Manufacture 3-22
-------
TABLE OF CONTENTS (continued)
3.4.2.4. Chloroform Emissions from Grain
Fumigation 3-23
3.4.2.5. Chloroform Losses from Loading and
Transportation 3-24
3.4.2.6. Miscellaneous Use Emissions 3-25
3.4.2.7. Summary of Chloroform Discharges
from Use 3-25
3.4.3. Summary 3-25
3.5. AMBIENT AIR CONCENTRATIONS 3-26
3.6. ATMOSPHERIC REACTIVITY 3-32
3.7. ECOLOGICAL EFFECTS/ENVIRONMENTAL PERSISTENCE 3-33
3.7.1. Ecological Effects 3-33
3.7.1.1. Terrestrial 3-33
3.7.1.2. Aquatic 3-34
3.7.2 Environmental Persistence 3-37
3.8. EXISTING CRITERIA, STANDARDS, AND GUIDELINES 3-40
3.8.1. Air 3-40
3.8.2. Water 3-42
3.8.3. Food 3-43
3.8.4. Drugs and Cosmetics 3-43
3.9. RELATIVE SOURCE CONTRIBUTIONS 3-43
3.10. REFERENCES FOR CHAPTER 3 3-44
4. DISPOSITION AND RELEVANT PHARMACOKINETICS 4-1
4.1. INTRODUCTION 4-1
4.2. ABSORPTION 4-2
4.2.1. Dermal Absorption 4-2
4.2.2. Oral , 4-3
4.2.3. Pulmonary Absorption 4-6
4.3. TISSUE DISTRIBUTION 4-12
4.4. EXCRETION 4-20
4.4.1. Pulmonary Excretion 4-20
4.4.2. Other Routes of Excretion 4-30
4.4.3. Adipose Tissue Storage 4-31
4.5. BIOTRANSFORMATION OF CHLOROFORM 4-32
-------
TABLE OF CONTENTS (continued)
4.5.1. Known Metabolites 4_32
4.5.2. Magnitude of Chloroform Metabolism 4-36
4.5.3. Enzymatic Pathways of Biotransformation 4-39
4.6. COVALENT BINDING TO CELLULAR MACROMOLECULES 4-45
4.6.1. Proteins and Lipids 4-45
4.6.1.1. Genetic Strain Difference 4-51
4.6.1.2. Sex Difference 4-53
4.6.1.3. Inter-species Difference 4-53
4.6.1.4. Age Difference 4-56
4.6.2. Nucleic Acids 4-56
4.6.3. Role of Phosgene 4-58
4.6.4. Role of Glutathione 4-59
4.7. SUMMARY 4-61
4.8. REFERENCES FOR CHAPTER 4 4-65
5. TOXICITY 5-1
5.1. EFFECTS OF ACUTE EXPOSURE TO CHLOROFORM 5-1
5.1.1. Humans 5-1
5.1.1.1. Acute Inhalation Exposure in Humans. . . . 5-1
5.1.1.2. Acute Oral Exposure in Humans 5-5
5.1.1.3. Acute Dermal and Ocular Exposure
in Humans 5-6
5.1.2. Experimental Animals 5-7
5.1.2.1. Acute Inhalation Exposure in Animals . . . 5-7
5.1.2.2. Acute Oral Exposure in Animals 5-8
5.1.2.3. Acute Dermal and Ocular Exposure
in Animals 5-11
5.1.2.4. Intraperitoneal and Subcutaneous
Administration in Animals 5-12
5.2. EFFECTS OF CHRONIC EXPOSURE TO CHLOROFORM 5-13
5.2.1. Humans 5-13
5.2.1.1. Chronic Inhalation Exposure in Humans . .5-13
5.2.1.2. Chronic Oral Exposure in Humans 5-15
5.2.2. Experimental Animals 5-16
5.2.2.1. Chronic Inhalation Exposure in Animals . 5-16
5.2.2.2. Chronic Oral Exposure in Animals 5-17
VI
-------
TABLE OF CONTENTS (continued)
5.3. INVESTIGATION OF TARGET ORGAN TOXICITY IN EXPERIMENTAL
ANIMALS 5-31
5.3.1. Hepatotoxicity 5-31
5.3.2. Nephrotoxicity 5-39
5.4. FACTORS MODIFYING THE TOXICITY OF CHLOROFORM 5-49
5.4.1. Factors that Increase the Toxicity 5-50
5.4.2. Factors that Decrease the Toxicity 5-58
5.5. SUMMARY: CORRELATION OF EXPOSURE AND EFFECT 5-60
5.5.1. Effects of Acute Inhalation Exposure 5-60
5.5.2. Effects of Acute Oral Exposure 5-61
5.5.3. Effects of Dermal Exposure 5-62
5.5.4. Effects of Chronic Inhalation Exposure 5-63
5.5.5. Effects of Chronic Oral Exposure 5-64
5.5.6. Target Organ Toxicity 5-66
5.5.7. Factors that Modify the Toxicity of Chloroform . . .5-76
5.6. REFERENCES FOR CHAPTER 5 5-77
6. TERATOGENICITY AND REPRODUCTIVE EFFECTS 6-1
6.1. SUMMARY 6-13
6.2 REFERENCES FOR CHAPTER 6 6-14
7. MUTAGENICITY 7-1
7.1. INTRODUCTION 7-1
7.2. COVALENT BINDING TO MACROMOLECULES 7-1
7.3. MUTAGENICITY STUDIES IN BACTERIAL TEST SYSTEMS 7-4
7.4. MUTAGENICITY STUDIES IN EUCARYOTIC TEST SYSTEMS 7-10
7.5. OTHER STUDIES INDICATIVE OF DNA DAMAGE 7-16
7.6. CYTOGENETIC STUDIES 7-21
7.7. SUGGESTED ADDITIONAL TESTING 7-25
7.8. SUMMARY AND CONCLUSIONS 7-26
7.9. REFERENCES FOR CHAPTER 7 7-27
8. CARCINOGENICITY 8-1
8.1. ANIMAL STUDIES 8-1
8.1.1 Oral Administration (Gavage): Rat 8-2
8.1.1.1 National Cancer Institute (1976) 8-2
8.1.1.2 Palmer et al. (1979) 8-9
8.1.2 Oral Administration (Gavage): Mouse 8-11
-------
TABLE OF CONTENTS (continued)
8.1.2.1 National Cancer Institute (1976) . . .8-11
8.1.2.2 Roe et al. (1979) 8-14
8.1.2.3 Eschenbrenner and Miller (1945) . . 8-18
8.1.2.4 Rudali (1967) 8-21
8.1.3 Oral Administration (Drinking Water): Rat
and Mouse 8-22
8.1.3.1 Jorgenson et al. (1985) 8-22
8.1.4 Oral Administration (Capsules): Dog 8-26
8.1.4.1 Heywood et al. (1979) 8-26
8.1.5 Intraperitoneal Administration: Mouse 8-29
8.1.5.1 Roe et al. (1969) 8-29
8.1.5.2 Theiss et al. (1977) 8-30
8.1.6 Evaluation of Chloroform Carcinogenicity
by Reuber (1979) 8-31
8.1.7 Oral Administration (Drinking Water): Mouse:
Promotion of Experimental Tumors 8-32
8.1.7.1 Cape! et al. (1979) 8-32
8.2. CELL TRANSFORMATION ASSAY 8-38
8.2.1 Styles (1979) 8-38
8.3. EPIDEMIOLOGIC STUDIES 8-41
8.3.1 Young et al. (1981) 8-42
8.3.2 Hogan et al. (1979) 8-46
8.3.3 Cantor et al. (1978) 8-47
8.3.4 Gottlieb et al. (1981) 8-52
8.3.5 Alavanja et al. (1978) 8-54
8.3.6 Brenniman et al. (1978) 8-56
8.3.7 Struba et al. (1979) 8-58
8.3.8 Discussion 8-60
8.4. RISK ESTIMATES FROM ANIMAL DATA 8-63
8.4.1 Possible Mechanisms Leading to a
Carcinogenic Response for Chloroform 8-64
8.4.2 Selection of Animal Data Sets 8-67
8.4.2.1 NCI 1976 Bioassay (Mice): Liver
Tumors 8-67
8.4.2.2 NCI 1976 Bioassay (Rats): Kidney
Tumors 8-68
VI
-------
TABLE OF CONTENTS (continued)
8.4.2.3 Roe et al. 1979 Bioassay (Mice):
Kidney Tumors 8-69
8.4.2.4 Jorgenson et al. 1985 Bioassay
(Rats): Kidney Tumors 8-69
8.4.3 Interspecies Dose Conversion 8-71
8.4.3.1 General Considerations 8-71
8.4.3.2 Calculation of Human Equivalent Doses . . 8-79
8.4.4 Choice of Risk Model 8-87
8.4.4.1 General Considerations 8-87
8.4.4.2 Mathematical Description of Low-Dose
Extrapolation Model 8-90
8.4.4.3 Adjustment for Less than Lifespan
Duration of Experiment 8-91
8.4.4.4 Additional Low-Dose Extrapolation .... 8-92
8.4.5 Unit Risk Estimates 8-93
8.4.5.1 Definition of Unit Risk 8-93
8.4.5.2 Calculation of the Slope of the
Dose-Risk Relationship for Chloroform . .8-93
8.4.5.3 Risk Associated with 1 jag/m3 of
Chloroform in Air 8-96
8.4.5.4 Risk Associated with 1 ng/L of
Chloroform in Drinking Water 8-96
8.4.5.5 Interpretation of Unit Risk Estimates . . 8-97
8.4.5.6 Reconciliation of Unit Risk Estimates
with Epidemiological Evidence 8-98
8.4.5.7 Discussion 8-98
8.5. RELATIVE CARCINOGENIC POTENCY 8-100
8.5.1 Derivation of Concept 8-100
8.5.2 Potency Index 8-100
8.6. SUMMARY 8-106
8.6.1 Qualitative 8-106
8.6.2 Quantitative 8-110
8.7. CONCLUSIONS 8-112
8.8. REFERENCES FOR CHAPTER 8 8-115
APPENDIX 8A COMPARISON AMONG DIFFERENT EXTRAPOLATION MODELS. . . 8A-1
IX
-------
LIST OF TABLES
Table Pa^e
3-1 Physical Properties of Chloroform 3-3
3-2 Chloroform Producers, Production Sites, and Capacities .... 3-9
3-3 Estimated Chloroform Discharges from Direct Sources 3-13
3-4 Ethylene Dichloride Producers, Production Sites, and
Capacities 3-15
3-5 Chloroform Discharges from Indirect Sources 3-21
3-6 Chlorodifluoromethane Producers and Production Sites 3-23
3-7 Chloroform Discharges from Use 3-26
3-8 Relative Source Contribution for Chloroform 3-27
3-9 Ambient Levels of Chloroform 3-28
3-10 Acute and Chronic Effects of Chloroform on Aquatic
Organisms 3-35
3-11 Values for kQH 3-38
3-12 Summary of EXAMS Models of the Fate of Chloroform 3-41
4-1 Physical Properties of Chloroform and Other
Chloromethanes 4-4
4-2 Partition Coefficients for Human Tissue at 37°C 4-4
4-3 Retention and Excretion of Chloroform by Man During and
After Inhalation Exposure to Anesthetic Concentrations .... 4-8
4-4 Chloroform Content in United Kingdom Foodstuffs and
in Human Autopsy Tissue 4-13
4-5 Concentration of Chloroform in Various Tissues of Two
Dogs After 2.5 Hours Anesthesia 4-16
4-6 Concentrations of Radioactivity (Chloroform Plus
Metabolites) in Various Tissues of the Mouse (NMRI) 4-17
4-7 Tissue Distribution of 14C-Chloroform Radioactivity
in CF/LP Mice After Oral Administration (60 mg/kg) 4-19
4-8 Pulmonary Excretion of 13CHC13 Following Oral Dose 4-25
-------
LIST OF TABLES (continued)
Table Page
4-9 Species Difference in the Metabolism of 14C-Chloroform . . . . 4-28
4-10 Kinetic Parameters for Chloroform After I.V. Administration
to Rats 4-29
4-11 Levels of Chloroform in Breath of Fasted Normal Healthy
Men 4-33
4-12 Covalent Binding of Radioactivity From ^C-Chloroform and
C-Carbon Tetrachloride in Microsomal Incubation Jji Vitro. . 4-49
4-13 Mouse Strain Difference in Covalent Binding of Radioactivity
From 14C-Chloroform 4-52
4-14 in Vivo Covalent Binding of Radioactivity From 14CHC13
in Liver and Kidney of Male and Female Mice (C57BL/6) .... 4-54
4-15 I_n Vitro Covalent Binding of Radioactivity from 14CHClo
to Microsomal Protein from Liver and Kidney of Male and
Female Mice (C57BL/67 4-54
4-16 Covalent Binding of Radioactivity from ^4C-Chloroform and
C-Carbon Tetrachloride in Rat Liver Nuclear and Microsomal
Incubation Iji Vitro 4-58
4-17 Effect of Glutathione, Air, N2 or CO: 02 Atmosphere
on the In Vitro Covalent Binding of CCl^, CHCl^ and CBrCl
to Rat Liver Microsomal Protein 4-60
4-18 Effects of 24-Hour Food Deprivation on Chloroform and
Carbon Tetrachloride In Vitro Microsomal Metabolism,
Protein, and P-450 Liver Contents of Rats 4-62
5-1 Relationship of Chloroform Concentration in Inspired
Air and Blood to Anesthesia 5-2
5-2 Dose-Response Relationships 5-6
5-3 Effects of Inhalation Exposure of Animals to Chloroform,
Five Days/Week for Six Months 5-18
5-4 Effects of Subchronic or Chronic Oral Administration of
Chloroform to Animals 5-20
5-5 Target Organ Toxicity of Chloroform 5-67
6-1 Summary of Results of the Schwetz et al. (1974) Study 6-3
6-2 Summary of Results of the Murray et al. (1979) Study 6-6
xi
-------
LIST OF TABLES (continued)
Table
6-3 Summary of Effects of the Thompson et al. (1974) Study. . . . 6-10
7-1 Genetic Effects of Chloroform on Strain 07 of
S. Cerevisiae 7-12
7-2 Mitotic Index, Anaphase/Metaphase, and Presence of
Complete C-Mitosis in Grasshopper Embryos after Exposure
to CHC13 Vapor 7-25
8-1 Effect of Chloroform on Kidney Epithelial Tumor Incidence
in Osborne-Mendel Rats 8-6
8-2 Effect of Chloroform on Thyroid Tumor Incidence in Female
Osborne-Mendel Rats 8-7
8-3 Toothpaste Formulation for Chloroform Administration 8-9
8-4 Effects of Chloroform on Hepatocellular Carcinoma Incidence
in B6C3F1 Mice 8-13
8-5 Kidney Tumor Incidence in Male ICI Mice Treated with
Chloroform 8-17
8-6 Liver and Kidney Necrosis and Hepatomas in Strain A Mice
Following Repeated Oral Administration of Chloroform
in Olive Oil 8-20
8-7 Relative Tumor Incidence in Male Osborne-Mendel Rats
Treated with Chloroform in Drinking Water 8-24
8-8 Liver Tumor Incidence Rates in Female B6C3F1 Mice Treated
with Chloroform in Drinking Water 8-25
8-9 SGPT Changes in Beagle Dogs Treated with Chloroform 8-28
8-10 Effect of Oral Chloroform Ingestion on the Growth of Ehrlich
Ascite Tumors 8-35
8-11 Effect of Oral Chloroform Ingestion on Metastatic
"Tumor Takes" with B16 Melanoma 8-36
8-12 Effect of Oral Chloroform Ingestion on the Growth and
Spread of the Lewis Lung Tumor 8-37
8-13 Correlation Coefficients Between Residual Mortality
Rates in White Males and THM Levels in Drinking Water
by Region and by Percent of the County Population
Served in the United States 8-50
XI 1
-------
LIST OF TABLES (continued)
Table Page
8-14 Correlation Coefficients Between Bladder Cancer
Mortality Rates by Sex and BTHM Levels in Drinking
Water by Region of the United States 8-50
8-15 Risk of Mortality from Cancer of the Rectum Associated
with Levels of Organics in Drinking Water 8-54
8-16 Cancer Risk Odds Ratios and 95% Confidence Intervals
(Chlorinated Versus Unchlorinated) 8-61
8-17 Incidence of Tumors in Experimental Animals 8-68
8-18 Species Difference in the Metabolism of 14C-chloroform
(Oral Dose of 60 mg/kg) 8-78
8-19 Pulmonary Excretion of Chloroform Following Oral Dose. . . . 8-78
8-20 Continuous Human Equivalent Doses and Incidence of
Hepatocellular Carcinomas in Male and Female B6C3F1 Mice. . . 8-84
8-21 Continuous Human Equivalent Doses and Incidence of
Renal Tubular-Cell Adenocarcinomas in Male
Osborne-Mendel Rats 8-84
8-22 Continuous Human Equivalent Doses and Incidence of
Malignant Kidney Tumors in Male ICI Mice 8-85
8-23 Continuous Human Equivalent Doses and Incidence of
Renal Tubular-Cell Adenocarcinomas in Male
Osborne-Mendel Rats 8-85
8-24 Upper-Bound Estimates of Cancer Risk of 1 mg/kg/day,
Calculated by Different Models on the Basis of Different
Data Sets 8-95
8-25 Relative Carcinogenic Potencies Among 55 Chemicals Evaluated
by the Carcinogen Assessment Group as Suspect Human
Carcinogens . • 8-102
8A-1 Maximum Likelihood Estimate of the Parameters for Each
of the Four Extrapolation Models, Based on Different
Data Sets 8A-2
XI 1 1
-------
LIST OF FIGURES
Page
Rate of Rise of Alveolar (Arterial) Concentration Toward
Inspired Concentration For Five Anesthetic Agents of
Differing Ostwald Solubilities 4-9
4-2 Arteriovenous Blood Concentrations of a Patient During
Anesthesia with Chloroform 4-10
4-3 Exponential Decay of Chloroform, Carbon Tetrachloride,
Perchloroethylene and Trichloroethylene in Exhaled
Breath of 48 Year-old Male Accidentally Exposed to
Vapors of These Solvents 4-21
4-4 Relationship Between Total 8-Hour Pulmonary Excretion of
Chloroform Following 0.5-g Oral Dose in Man and the
Deviation of Body Weight From Ideal 4-26
4-5 Blood and Adipose Tissue Concentrations of Chloroform During
and After Anesthesia in a Dog 4-32
4-6 Metabolic Pathways of Chloroform Biotransformation 4-34
4-7 Metabolic Pathways of Carbon Tetrachloride
Biotransformation 4-41
4-8 Rate of Carbon Monoxide Formation After Addition of Various
Halomethanes to Sodium Dithionite-reduced Liver Microsomal
Preparations From Phenobarbitol-treated Rats 4-46
4-9 Effect of Increasing Dosage of i.p.-Injected ^C-Chloroform
on Extent of Covalent Binding of Radioactivity In Vivo to
Liver and Kidney Proteins of Male Mice 6 Hours
after Administration 4-50
4-10 Comparison of Irreversible Binding of Radioactivity from
14C-CHCl3 to Protein and Lipid of Microsomes from
Normal Rabbit, Rat, Mouse, and Human Liver Incubated
In Vitro at 37°C in 02 4-55
5-1 Probable Pathways of Metabolism of Chloroform in the
Kidney 5-44
8-1 Survival Curves for Fisher 344 Rats in a Carcinogenicity
Bioassay on Chloroform 8-4
8-2 Negative Result in Transformation Assay of Chloroform
which was also Negative in the Ames Assay 8-40
xiv
-------
LIST OF FIGURES (continued)
Figure Page
8-3 Freguency Distribution of CHC13 Levels in 68 U.S.
Drinking Water Supplies 8-49
8-4 Effect of Increasing Dosage of I.P.-injected 14C-chloroform
on Extent of Covalent Binding of Radioactivity i_n vivo
to Liver and Kidney Proteins of Male Mice 6 Hours After
Administration 8-74
8-5 Comparison of Irreversible Binding of Radioactivity
from 14C-CHC13 to Protein and Lipid of Microsomes from
Normal Rabbit, Rat, Mouse, and Human Liver Incubated
i_n vitro at 37°C in 02 8-75
8-6 Allometric Relationship (Y=aWn) Between Species Body
Weight (in order: mouse, rat, squirrel monkey, and man)
and the Amount Metabolized of a Common Oral Dose of
Chloroform as Calculated from the Data of Fry et al.,
(1972) and Brown et al., (1974) 8-80
8-7 The Relationship Between the Equivalent Human Dose and
Bioassay Tumor Incidence 8-86
8-8 Histogram Representing the Frequency Distribution of the
Potency Indices of 55 Suspect Carcinogens Evaluated by
the Carcinogen Assessment Group 8-101
xv
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
The EPA Office of Health and Evironmental Assessment (OHEA) is
responsible for the preparation of this health assessment document. The OHEA
Environmental Criteria and Assessment Office (ECAO, Research Triangle Park,
NC 27711) had overall responsibility for coordiantion and direction of the
document production effort (Si Duk Lee, Ph.D., Project Manager, ECAO,
919_541_4159).
AUTHORS
Larry Anderson, Ph.D.
Carcinogen Assessment Group
Office of Health and Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
David Bayliss, M.S.
Carcinogen Assessment Group
Office of Health and Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Chao W. Chen, Ph.D.
Carcinogen Assessment Group
Office of Health and Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Joan P. Coleman, Ph.D.
Syracuse Research Corporation
Syracuse, NY
I.W.F. Davidson, Ph.D.
Bowman Gray School of Medicine
Winston-Salem, NC
D. Anthony Gray, Ph.D.
Syracuse Research Corporation
Syracuse, NY
Si Duk Lee, Ph.D.
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
U.S. Environmental Protection Agency
Research Triangle Park, NC
Chapter 8
Chapter 8
Chapter 8
Chapter 5
Chapter 4
Chapter 3
Chapter 2
xvi
-------
Sheila Rosenthal, Ph.D. Chapter 7
Reproductive Effects Assessment Group
Office of Health and Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Carol Sakai, Ph.D. Chapter 6
Reproductive Effects Assessment Group
Office of Health and Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Sharon B. Wilbur, M.A. Chapter 5
Syracuse Research Corporation
Syracuse, NY
U.S. Environmental Protection Agency Peer Reviewers
Karen Blanchard
Office of Air Quality Planning and Standards
Research Triangle Park, NC
Lester D. Grant, Ph.D.
Office of Health and Environmental Assessment
Environmental Criteria and Assessment Office
Research Triangle Park, NC
Joseph Padgett
Office of Air Quality Planning and Standards
Research Triangle Park, NC
Jerry F. Stara, D.V.M.
Office of Health and Environmental Assessment
Environmental Criteria and Assessment Office
Cincinnati, OH
Environmental Criteria and Assessment Office Support Staff
F. Vandiver Bradow
Allen Hoyt
xvn
-------
The OHEA Carcinogen Assessment Group (CAG) was responsible for
preparation of the sections on carcinogenicity. Participating members of the
CAG are listed below (principal authors of present carcinogenicity materials
are designated by an asterisk (*).
Participating Members of the Carcinogen Assessment Group
Roy E. Albert, M.D. (Chairman)
Elizabeth L. Anderson, Ph.D.*
Larry D. Anderson, Ph.D.*
Steven Bayard, Ph.D.
David L. Bayliss, M.S.*
Robert P. Bellies, Ph.D.*
Chao W. Chen, Ph.D.*
Margaret M.L. Chu, Ph.D.
James Cogliano, Ph.D.
Bernard H. Haberman, D.V.M., M.S,
Charalingayya B. Hiremath, Ph.D.
Robert E McGaughy, Ph.D.
Dharm W. Singh, D.V.M. Ph.D.
Todd W. Thorslund, Sc.D.
The OHEA Reproductive Effects Assessment Group (CAG) was responsible for
preparation of the sections on mutagenicity, teratogenicity,and reproductive
effects. Participating members of the REAG are listed below (principal
authors of present sections are designated by an asterisk (*).
Participating Members of the Reproductive Effects Assessment Group
Eric D. Clegg, Ph.D.
John R. Fowle, III, Ph.D.
David Jacobsen-Kram, Ph.D.
K.S. Lavappa, Ph.D.
Sheila Rosenthal, Ph.D.*
Carol Sakai, Ph.D.*
Lawrence R. Valcovic, Ph.D.
Vicki Vaughn-Dellarco, Ph.D.
Peter E. Voytek, Ph.D. (Director)
xvi
-------
External Peer Reviewers
Dr. Karim Ahmed
Natural Resources Defense Fund
122 East 42nd Street
New York, NY 10168
Dr. Eula Bingham
Graduate Studies and Research
University of Cincinnati (ML-627)
Cincinnati, OH 45221
Dr. James Buss
Chemical Industry Institute of
Toxicology
Research Triangle Park, NC 27709
Dr. I.W.F. Davidson
Wake Forest University
Bowman Gray School of Medicine
Winston-Salem, NC
Dr. Larry Fishbein
National Center for Toxicological
Research
Jefferson, AR 72079
(501) 542-4390
Dr. Derek Hodgson
Department of Chemistry
University of North Carolina
Chapel Hill, NC 27514
Dr. Marshall Johnson
Thomas Jefferson Medical College
Department of Anatomy
1020 Locust Street
Philadelphia, PA 19107
Dr. Trent Lewis
National Institute of Occupational
Safety and Health
26 Columbia Parkway
Cincinnati, OH 45226
(513) 684-8394
Dr. Richard Reitz
Dow Chemical, USA
Toxicological Research Laboratory
1803 Building
Midland, MI 48640
Dr. Bernard Schwetz
National Institute of
Environmental Health Sciences
Research Triangle Park, NC 27709
Dr. James Selkirk
Oak Ridge National Laboratory
Oak Ridge, TN 37820
(615) 624-0831
Dr. Samuel Shibko
Food and Drug Administration
Division of Toxicology
200 C Street, SW
Washington, DC 20204
Dr. Robert Tardiff
1423 Trapline Court
Vienna, VA 22180
(703) 276-7700
Dr. Norman M. Trieff
University of Texas Medical Branch
Department of Pathology
Galveston, TX 77550
(409) 761-1895
Dr. Benjamin Van Duuren
Institute of Environmental Medicine
New York University Medical Center
New York, NY 10016
(212) 340-5629
Dr. James Withey
Health and Protection Branch
Department of National Health &
Welfare
Tunney's Pasture
Ottawa, Ontario KIA 01Z Canada
Mr. Matthew Van Hook
Consultant
1133 North Harrison Street
Arlington, VA 22205
xix
-------
SCIENCE ADVISORY BOARD
ENVIRONMENTAL HEALTH COMMITTEE
CHLORINATED ORGANICS SUBCOMMITTEE
The content of this health assessment document on chloroform was
independently peer-reviewed in public session by the Chlorinated Organics
Subcommittee of the Environmental Health Committee of the Environmental
Protection Agency's Science Advisory Board.
ACTING CHAIRMAN. ENVIRONMENTAL HEALTH COMMITTEE
Dr. John Doull, Professor of Pharmacology and Toxicology, University of
Kansas Medical Center, Kansas City, Kansas 66207
EXECUTIVE SECRETARY, SCIENCE ADVISORY BOARD
Dr. Daniel Byrd III, Executive Secretary, Science Advisory Board, A-101 F,
U.S. Environmental Protection Agency, Washington, DC 20460
MEMBERS
Dr. Seymour Abrahamson, Professor of Zoology and Genetics, Department of
Zoology, University of Wisconsin, 500 Highland Avenue, Madison,
Wisconsin 53706
Dr. Ahmed E. Ahmed, Associate Professor of Pathology, Pharmacology, and
Toxicology, The University of Texas Medical Branch, Galveston, Texas
77550.
Dr. George T. Bryan, Profesor of Human Oncology, K4/528 C.S.C Clinical
Sciences, University of Wisconsin, 500 Highland Avenue, Madison,
Wisconsin 53792.
Dr. Ronald D. Hood, Professor and Coordinator, Cell and Developmental Biology
Section, Department of Biology, The University of Alabama, and Principal
Associate, R.D. Hood and Associates, Consulting Toxicologists, P.O. Box
1927, University. Alabama 35486.
Dr. K. Roger Hornbrook, Department of Pharmacology, P.O. Box 26901,
University of Oklahoma, Oklahoma City, Oklahoma 73190.
Dr. Thomas Starr, Chemical Industry Institute of Toxicology, P.O. Box 12137,
Research Triangle Park, North Carolina 27709.
xx
-------
1. SUMMARY AND CONCLUSIONS
1.1 INTRODUCTION
Chloroform is a dense, colorless, volatile liquid used primarily in the
production of chlorodifluoromethane (90%) and for export (5%). Non-
consumptive uses (5%) include use as a solvent, as a cleaning agent, and as a
fumigant ingredient. Although chloroform production and capacity have
declined recently, 1981 data place direct United States production of
chloroform at 184 million kg, with indirect production estimated at
13.2 million kg (-193 million kg overall). Also, based on 1981 data, the
amount of chloroform emitted to air is estimated to be 7.2 million kg, with
emissions to water of 2.6 million kg, and emissions to land of 0.6 million
kg. Total United States emissions are estimated (1981) to be 10.4
million kg.
Chloroform is ubiquitous in the environment, having been found in urban
and non-urban locations. There have been reports of a northern hemisphere
background average of 14 ppt (10~^ v/v), with an average in the southern
hemisphere of <3 ppt, and a global average of 8 ppt. However, a more recent
report suggests the ratio of hemispheric concentrations (north v. south) may
be less dramatic, more on the order of 1.6. This same research also suggests
that chloroform in the atmosphere may, on a global basis, be largely natural
in origin (from tropical oceans, in particular), rather than mostly
anthropogenic as previously thought. For the most part, urban ambient air
concentrations remain <1000 ppt, and rural or remote locations can be
<10 ppt. There are some notable exceptions, however, but the reasons for them
1-1
-------
are not readily apparent. The highest values reported were in Rutherford,
New Jersey (31,000 ppt), and Niagara Falls, New York (21,611 ppt).
1.2 PHYSICAL AND CHEMICAL PROPERTIES, AND ANALYSIS
Hydroxyl radical oxidation is the primary atmospheric reaction of
chloroform. Based on the rate constant for reaction with chloroform, a half-
life of 11.5 weeks is expected. The principal products from this reaction
are HC1 and C02. It has been estimated that roughly 1% of the tropospheric
chloroform will diffuse into the stratosphere, based on a lifetime of 0.2 to
d.3-year and a troposphere-to-stratosphere turnover time of 30 years. An
EXAMS model of chloroform in water confirms other data suggesting that the
N
major removal process for chloroform in water is evaporation.-
The -best analyti-eal~method for detection of chloroform appears to be gas
chromatography with electron capture or electrolytic conductivity detection.
This gives a detection limit of <5 ppt.
1.3 PHARMACOKINETICS
The pharmacokinetics and metabolism of chloroform have been studied in
both humans and experimental animals. Chloroform is rapidly and extensively
absorbed through the respiratory and gastrointestinal tracts. Absorption
through the skin would make a significant contribution to body burden only in
instances of contact of the skin with liquid chloroform.
The limited available data suggest that, in a human at rest, at least
2 hours are required to reach an apparent equilibrium of the body with the
inhaled chloroform concentration. The magnitude of chloroform uptake into
the body (dose or body burden) is directly proportional to the concentration
of chloroform in the inspired air, the duration of exposure, and the
respiratory minute volume.
1-2
-------
The absorption of chloroform from the gastrointestinal tract appears to
be virtually complete, judging from recovery of unchanged chloroform and
metabolites in the exhaled air of humans and in the exhaled air, urine,
feces, and carcass of experimental animals. Chloroform given in a corn oil
vehicle to experimental animals is absorbed more slowly than chloroform given
in water. Peak blood levels occurred at =1 hour after oral administration
of chloroform in olive oil to humans or animals.
After inhalation or ingestion, highest concentrations of chloroform are
found in tissues with higher /lipj'd contents. Results from the administration
-of ^C-labeled chloroform to animals indicate that the distribution of
radioactivity (reflecting both chloroform and its metabelites) may be
affected by the route of exposure. Oral administration appeared to result in
the accumulation of a greater proportion of radioactivity in the liver than
did inhalation exposure, but differences in experimental protocols make this
interpretation tentative. Differences in the distribution of chloroform and
its metabolites between male and female animals were found only in mice and
not in rats or squirrel monkeys. The kidneys of male mice accumulated
strikingly more radioactivity than did those of female mice.
Chlorofwm has been detected in fetal liver. Chloroform would be
expected to appear in human milk, because it has been found in cow's milk,
cheese, and butter.
Chloroform is metabolized via microsomal cytochrome P-450 oxidation to
trichloromethanol, which spontaneously dehydrochlorinates to the toxic
reactive intermediate compound, phosgene. The end products of the phosgene
reaction with cellular water are C02 and hydrochloric acid, but significant
amounts of phosgene and other reactive intermediates bind covalently to
tissue macromolecules or conjugate with cysteine and glutathione. Covalent
1-3
-------
binding of the reactive intermediates to macromolecules is considered to be
responsible for the hepato- and nephrotoxicity of chloroform. While the
liver is the primary site for chloroform metabolism, other tissues, including
the kidney, can also metabolize chloroform to CO^.
There is no evidence to suggest any qualitative difference for
chloroform metabolic pathways in mice, rats, and humans. Interspecies
comparisons of the magnitude of chloroform metabolism have been made only for
the oral route. Metabolism of chloroform across species, including mice,
rats, squirrel monkeys, and humans is proportional to the surface area of the
species. The end metabolite, C02, is excreted in expired air. Dose
dependent pulmonary exhalation is the principal route of excretion for
unmetabolized chloroform. Small amounts of chloroform metabolites are
excreted in the urine and feces. Results from observations in humans suggest
that chloroform metabolism is rate limited.
Regardless of the route of entry into the body, chloroform is excreted
unchanged through the lungs and eliminated via metabolism, with the primary
stable metabolite, CC^, also being excreted through the lungs. High
concentrations of unchanged chloroform have been found in the bile of
squirrel monkeys after oral administration, but not in the urine or feces.
The inorganic chloride generated f^om chloroform metabolism is excreted via
the urine.
Decay curves for the pulmonary excretion of unchanged chloroform in
humans appear to consist of three exponential components. The terminal
component, thought to correspond to elimination from adipose tissue, had a
half-time of 36 hours. This long half-time (36 hr) of chloroform residence
in the human fat compartment indicates that fatty tissue concentrations of
chloroform will not achieve steady-state equilibrium conditions with exposure
1-4
-------
concentrations until 6 to 7 days of continuous exposure to ambient
concentrations, or longer for repetitive daily exposures in the workplace.
Conversely, the long residence time of chloroform in the fat compartments of
humans indicates that complete desorption of chloroform from these
compartments requires 6 to 7 days in chloroform-free environs.
1.4 HEALTH EFFECTS OVERVIEW
Neurological, hepatic, renal, and cardiac effects have been associated
with exposure to chloroform. These effects have been documented in humans as
well as in experimental animals. In addition, studies with animals indicate
that chloroform is carcinogenic and may be-teratogenic.
1.4.1 Toxicity
Evidence of chloroform's effects on humans has been obtained -primarily
during the use of this chemical as an inhalation anesthetic. In addition to
depression of the central nervous system, chloroform anesthesia was
associated with cardiac arrhythmias (and some cases of cardiac arrest),
hepatic necrosis and fatty degeneration, polyuria, albuminuria, and in cases
of severe poisoning, renal tubular necrosis. When used for obstetrical
anesthesia, chloroform was likely to produce respiratory depression in the
infant. Experimental exposures of humans to chloroform have focused only on
subjective responses. Humans exposed experimentally to chloroform for 20 to
30 minutes have reported dizziness, headache, giddiness, and tiredness at
concentrations >1000 ppm, and light intoxication at concentrations above
4000 ppm.
Similar symptoms occurred in workers employed in the manufacture of
lozenges containing chloroform; exposure concentrations ranged from 20 to
237 ppm, with occasional brief exposure to =1000 ppm. Additional complaints
were of gastrointestinal distress, and frequent and scalding urination. The
1-5
-------
only other report of adverse effects stemming from occupational exposure to
chloroform was of enlargement of the liver; this report was compromised by
the apparent lack of suitable controls.
Acute inhalation experiments with animals revealed that single exposures
to 100 ppm were sufficient to produce mild hepatic effects in mice. The
exposure level that would produce mild renal effects is not known, but frank
toxic effects occurred in the kidneys of male mice exposed to 5 mg/L
(1025 ppm). In subchronic inhalation experiments, histological evidence of
mild hepato- and nephrotoxicity occurred in rats with exposures to as low as
25 ppm, 7 hoursVday for 6 months. The effects were reversible if exposure was
terminated, and did not occur when exposure was limited to 4 hours/day.
Information on the effects of acute and long-term oral exposure to
chloroform is available primarily from experiments with animals. Human data
are mainly in the form of case reports and involve the abuse of medications
containing not only chloroform, but other potentially toxic ingredients as
well; however, a fatal dose of as little as 1/3 ounce was reported. As with
inhalation exposure, the primary effects of oral exposure were hepatic and
renal damage. Narcosis also occurred with high doses, but this effect was not
usually a focus of concern in these experiments. Subchronic and chronic
toxicity experiments with rats, mice, and dogs did not clearly establish a
no-effect level of exposure for systemic toxicity. Although a dose level of
17 mg/kg/day of chloroform produced no adverse effect in four strains of
mice, the lowest dosage tested, 15 mg/kg/day, elevated some clinical
chemistry indices of hepatic damage in dogs and appeared to affect a
component of the reticuloendothelial system (histiocytes) in their livers.
No controlled studies have been performed to define dose-response
thresholds for neurological or cardiac effects of ingested or inhaled
1-6
-------
chloroform. It is not known whether subtle impairment of neurological or
cardiac function might occur at levels as low as or lower than those which
affect the liver.
Several suostances that are of interest because of accidental or
intentional human exposure have been shown to modify the systemic toxicity of
chloroform, usually -by modifying the metabolism of chloroform to the reactive
intermediate. Examples of substances that potentiate chloroform-induced
toxicity are ethanol, PBBs, ketones and steroids. Factors that appear to
protect against toxicity include disulfiram and high carbohydrate diets.
1.4.2 Reproductive Effects
Observations reported in four articles and two abstracts indicate-that
chloroform at the concentrations used in these studies has the potential for
causing adverse effects in pregnancy maintenance, delays in fetal
development, and the production of terata in laboratory animals. The studies
which administered chloroform by inhalation 7 hr/day reported more severe
outcomes than other studies that administered chloroform by intubation, once
or twice a day. The adverse effects produced in the conceptus were observed
in association with maternal toxicity, however, the type and severity of
effects appeared to be specific to the conceptus, affecting development to a
much greater degree than the occurrence of maternal toxicity. It is
concluded that chloroform is a potential developmental toxicant. The results
of a preliminary study indicate that chloroform has no significant adverse
behavioral effect on the fetus or produces embryotoxic effects only at
maternally toxic levels.
1.4.3 Mutaqenicity
It has been demonstrated that chloroform can be metabolized in vivo and
i_n vitro to a substance (presumably phosgene) that interacts with protein and
1-7
-------
lipid. The potential for metabolically activated chloroform to bind to DNA
cannot be determined from the available studies, but, if binding to DNA does
occur, it would be at a very low level.
The majority of the assays for genotoxicity have yielded negative
results. However, many of these results are inconclusive because of
inadequacies in the experimental protocols. The major problem has been the
use of exogenous activation systems (i.e., S9 mix). In none of the studies
was it shown that chloroform was activated or metabolized by the activation
system used. Metabolism of 2-aminoanthracene or vinyl compounds (used as
positive controls) is an inadequate indication that the activation system can
metabolize chloroform because these substances are not halogenated alkanes.
A better indication that the activation system is sufficient for metabolism
of chloroform may be to show that it metabolizes ^CHC^ to intermediates
that bind to macromolecules. A second problem in the use of exogenous
activation systems is the possibility that highly reactive metabolites may
react with microsomal or membrane lipid or protein before reaching the DNA of
the test organism. Another problem in jm vitro tests is that adequate
precautions are sometimes not taken to prevent the escape of volatilized
chloroform.
On the basis of presently available data, no definitive conclusion can
be reached concerning the mutagenicity of chloroform. However, evidence from
studies measuring binding to macromolecules, DNA damage, and mitotic arrest
suggest that chloroform may be mutagenic. Alternatively, because recent
studies on the mechanism of action of tumor promoters suggest that promoters
can damage DNA, chloroform may promote carcinogenesis rather than initiate
it.
1-8
-------
1.4.4 CareInoqenicity
The carcinogenic potential of chloroform has been experimentally
evaluated in several animal species and by epidemiologic surveys. Chronic
animal studies have been conducted in eight strains of mice, two strains of
rats, and beagle dogs. In all of these studies chloroform was administered
by the oral route and not by inhalation, an important route of chloroform
exposure for humans. However, a carcinogenic response from chloroform
exposure is not expected to be dependent upon the route of assimilation into
the body.
Evidence for the carcinogenicity of chloroform in experimental animals
includes: statistically significant increases in renal epithelial tumors in
male Osborne-Mendel rats (two studies); hepatocellular carcinomas in male and
female BeCsFi mice; kidney tumors in male ICI mice; and hepatomas in female
Strain A mice and NIC mice. Chloroform has also been shown to promote growth
and metastasis of murine tumors. In these cancer studies the carcinogenicity
of chloroform is organ specific, primarily liver and kidney, the target
organs of acute chloroform toxicity and covalent binding as well.
The carcinogenicity of chloroform was first investigated in 1945.
Although the number of-animals in eath test group was small and the mortality
was high at the higher doses, an increased incidence of hepatomas was
observed in the Strain A mice. Induction of hepatomas was confirmed in 1967
in a very limited study in NLC mice. Chloroform was administered in oil by
gavage in both studies.
In 1976, male and female B6C3F1 mice, chloroform-treated by corn oil
gavage, showed highly significant dose-dependent increases in hepatocellular
carcinomas, with metastases to the lungs in some mice. In a similar study,
1-9
-------
statistically significant dose-dependent increases of kidney epithelial
tumors were found in male Osborne-Mendel rats.
In another study (1979), kidney tumors were observed in male ICI mice
administered chloroform in either toothpaste or arachis oil vehicle.
In the most recently published study (1985), chloroform administered in
the drinking water of male Osborne-Mendel rats induced a statistically
significant increase in the incidence of renal tumors, thus supporting the
findings from the earlier study in which chloroform was administered in corn
oil by gavage. Female BeCsFi mice, however, did not show an increase in the
incidence of liver tumors when chloroform-was administered in the drinking
water. This was inconsistent with the positive findings reported in previous
investigations of chloroform oil gavage treatment of mice. The lack of
response of the mice in the drinking water study versus the highly
significant response of these mice when chloroform was given in corn oil
vehicle as a single bolus, suggests that chloroform-induced heptatocellular
carcinomas in this strain of mice may be related to absorption patterns, the
dosing regimen, peak blood levels of, chloroform, and target tissue levels of
its reactive intermediate metabolites. The corn oil carrier has not been
shown to induce an increase in the incidence of liver tumors ~rn mice.
Other studies of chloroform carcinogenicity have shown negative results.
Treatment with a gavage dose of chloroform in toothpaste did not produce a
carcinogenic response in female ICI mice or in male mice of the CBA, C57BL,
and CF/1 strains, nor was a carcinogenic response observed in male or female
Sprague-Dawley rats given chloroform in toothpaste by gavage, but early
mortality was high in control and treatment groups. Gavage doses of
chloroform in toothpaste did not cause a carcinogenic response in male and
female beagle dogs treated for over 7 years, although there was an increased
1-10
-------
incidence of hepatic nodular hyperplasia. The daily chloroform doses given
to mice and rats in toothpaste or arachis oil were lower than those given in
corn oil or drinking water in studies showing a positive carcinogenic
response. In newborn (C57 x DBA2-F1) mice given subcutaneous doses during
the initial 8 days of life and observed for their lifetimes, a carcinogenic
effect of chloroform was not evident. The dose levels used appeared well
below a maximum tolerated dose and the period of treatment was quite short.
In Strain A mice, chloroform was ineffective at maximally tolerated and lower
doses in a pulmonary adenoma bioassay. However, other chemicals that have
shown carcinogenic activity in different tests were ineffective in this
particular Strain A mouse pulmonary adenoma bioassay. Chloroform does-not
induce transformation of Syrian baby hamster kidney cells (BHK-21/C1 13)
i_n vitro.
While no epidemiological studies have evaluated chloroform by itself,
several studies have been made of populations with chlorinated drinking
water, in which chloroform is the predominant chlorinated hydrocarbon
compound. Small increases in rectal, bladder, and colon cancer were
consistently observed by several case-control and ecological studies, several
of which are statistically significant. Because other possible carcinogens
were present along with chloroform, it is impossible to identify chloroform
as the sole carcinogenic agent. Therefore, the epidemiologic evidence for
chloroform's carcinogenicity must be termed inadequate.
It is generally accepted that the carcinogenic activity of chloroform
resides in its highly reactive intermediate metabolites such as phosgene.
Irreversible binding of chloroform metabolites to cellular macromolecules
supports several theoretical concepts of the mechanism(s) for its
carcinogenicity. Available data on chloroform metabolism and
1-11
-------
pharmacokinetics pertinent to the conditions of the carcinogenicity bioassays
are used in the extrapolation of the dose-carcinogenic response relationships
of laboratory animals to humans. There is no difference in absorption of
chloroform across species. Also, there is no evidence to suggest any
qualitative difference in the metabolic pathways or metabolite profiles of
mice, rats, and humans for chloroform. An experimental basis exists for
determining relative amounts of chloroform metabolized in various species,
including man, and this information has been used in the risk assessment for
chloroform.
1.4.5 Quantitative Risk_Assessment
The quantitative risk assessment is based on the assumption of a non-
threshold mechanism, and- consequently .mathematical extrapolation models
consistent with this assumption were evaluated. Although the nonthreshold
mathematical risk extrapolation model is conservative based upon a public
health point of view, the correction used in the calculation of a human
equivalent dose is scientifically conservative and may lead to an
overestimate of the amount of chloroform metabolized in the test animals, and
hence underestimate the risk. In addition, experimental data that include
covaleni- binding in. human tissues suggest that humans may have a greater than
expected capacity to metabolize chloroform when compared to rodents, again
indicating the possibility of a higher risk for humans than estimated in the
assessment. Using the linearized multistage model, the geometric mean, 8.1 x
10~2 per mg/kg/day, of the slope estimates, q^*, calculated from chloroform-
induced liver tumors in male and female mice treated by gavage, is the value
used to compare the relative potency of chloroform to other carcinogens and
to calculate the unit risk for drinking water and air. The upper-bound
1-12
-------
estimate of cancer risk due to 1 pg/L of chloroform in water is consistent
with the limited epidemiologic data available for humans.
Five data sets are used to estimate the carcinogenic risk of chloroform.
The end points include liver tumors in female mice (NCI, 1976), liver tumors
in male mice (NCI, 1976), kidney tumors in male rats (NCI, 1976, and
Jorgenson et al., 1985) and kidney tumors in male mice (Roe et al., 1979).
The unit risk values at 1 mg/kg/day, calculated by the linearized multistage
model on the basis of these data sets, are comparable. The risk value is
useful for estimating the possible magnitude of the public health impact.
The upper-bound incremental cancer risk derived from the geometric mean of 4
data sets, chloroform gavage studies which showed a statistically significant
increase of hepatocellular carcinomas in mice, is 8.1 x 10~^ per mg/kg/day.
The CAG potency index for chloroform (defined as the slope x molecular
weight) is 1 x 10 , ranking it in the lowest quartile of 55 chemicals that
the CAG has evaluated as suspect carcinogens. The upper-bound estimate of
the incremental cancer risk due to ingesting 1 ^g/L of chloroform in drinking
water is 2.3 x 10 . The upper-bound estimate of the incremental cancer risk
due to inhaling 1 pg/nr of chloroform in air based upon positive gavage
carcinogenicity studies is 2.3 x 10~5. The upper-bound nature of these
estimates is such that the true risk is not likely to exceed this value and
may be lower.
Based on EPA's propdsed Carcinogen Risk Assessment Guidelines,
chloroform is classified as having sufficient animal evidence and inadequate
epidemiologic evidence. The overall weight-of-evidence classification is
group B2, meaning that chloroform is probably carcinogenic in humans.
Applying the International Agency for Research on Cancer (IARC) criteria, the
level of animal evidence for carcinogenicity is sufficient, and the overall
1-13
-------
IARC classification is group 2B, meaning that chloroform should be considered
to be a probable human carcinogen.
1-14
-------
2. INTRODUCTION
The U.S. Environmental Protection Agency is responsible under the
authority of various laws for the identification, comprehensive assessment,
and as appropriate regulation, of environmental substances which may be of
concern to the public health. For example, under Section 112 of the Clean
Air Act the EPA Administrator is directed to establish standards for any air
pollutant (other than those for which national ambient air quality standards
are applicable) which, in his judgment, "causes, or contributes to, air
pollution which may reasonably be anticipated to result in an increase in
mortality or an increase in serious irreversible, or incapacitating
reversible, illness."
Within EPA, the Office of Health and Environmental Assessment is
responsible for providing scientific assessments of health effects for
potentially hazardous air pollutants such as chloroform. These health
assessment documents form the scientific basis for subsequent agency actions,
including the Administrator's judgment as to whether regulations or standards
may be appropriate.
This Health Assessment Document for Chloroform represents a
comprehensive data base that considers all sources of chloroform in the
environment, the likelihood for human exposures, and the possible
consequences to man and lower organisms from its absorption. This
information is integrated into a format that can serve as the basis for
qualitative and quantitative risk assessments, while at the same time
identifying gaps in our knowledge that limit present evaluative capabilities.
Accordingly, it is expected that this document may serve the information
needs of many government agencies and private groups that may be involved in
2-1
-------
decision making and regulatory activities. (As with all such EPA documents,
a preliminary draft was made available to the scientific community and the
general public so that comments of interested individuals and organizations
could be considered, and the latest scientific evidence incorporated, in the
final draft. The preliminary draft was also reviewed by the Environmental
Health Advisory Committee of EPA's Science Advisory Board at a public meeting
(49 Federal Register 9609).
2-2
-------
3. BACKGROUND INFORMATION
3.1. INTRODUCTION
This section provides background information supportive of the human
health effects data presented in subsequent sections. It is not intended to
be a comprehensive review of analytical methodology, sources, emissions, air
concentrations, or environmental transport and fate information. In order to
fulfill this purpose and since the literature concerning chloroform is vast,
only a portion of the available literature was included. Those articles
included were chosen because of their relevance to the topic at hand and
because they were representative of the literature as a whole.
To provide the most complete overview possible, some non-peer reviewed
information has been added, from the Chloroform Materials Balance Draft
Report by Rehm et al. (1982). This is an updated version of the original
Level I Materials Balance: Chloroform (Wagner et al., 1980). As described by
Wagner et al. (1980):
A Level I Materials Balance requires the lowest level of
effort and involves a survey of readily available information for
constructing the materials balance. Ordinarily, many assumptions
must be made in accounting for gaps in information; however, all
are substantiated to the greatest degree possible. Where possible,
the uncertainties in numerical values are given, otherwise they are
estimated. Data gaps are identified and recommendations are made
for filling them. A Level I Materials Balance relies heavily on
the EPA's Chemical Information Division (CID) as a source of data
and references involving readily available information. Most Level
I Materials Balance are completed within a 3-6 week period; CID
literature searches generally require a 2 week period to complete.
Thus, the total time required for completion of a Level I materials
balance ranges from 5-7 weeks.
Because a greater level of effort went into the 1980 and 1982 Materials
Balance reports on chloroform than would normally be devoted to background
(non-health effects) information in an EPA health assessment document,
3-1
-------
information in this chapter is drawn from these reports. However, because
such information has not been peer-reviewed and includes some major
assumptions, it should not be used to support any regulations or standards
regarding risks to public health.
3.2. PHYSICAL AND CHEMICAL PROPERTIES
Chloroform (CHC13) (CAS registry number 67-66-3) is a member of a family
of halogenated saturated aliphatic compounds. Synonyms for chloroform
include the following:
Chloroforme (French) Methenyl trichloride
Chloroformio (Italian) Methyl trichloride
Formyl trichloride NCI-C02686
Methane trichloride Trichloromethaan (Dutch)
Methane, trichloror Trichloroform
Methenyl chloride Trichloromethane
Table 3-1 lists various physical properties. Chloroform is a colorless,
clear, dense, volatile liquid with an ethereal non-irritating odor (DeShon,
1979). Chloroform is nonflammable; however, when hot chloroform vapors are
mixed with alcohol vapors, the mixture burns with a greenish flame. At 25°C
and 1 atmosphere, a 1 ppm concentration of chloroform in air is equal to
4.88 mg/m3.
Chloroform decomposes with prolonged exposure to sunlight regardless of
the presence of air (DeShon, 1979). It also decomposes in the dark in the
presence of air. The principal decomposition products include phosgene,
hydrogen chloride, chlorine, carbon dioxide, and water. Ozone causes
chloroform to decompose rapidly.
Chloroform forms a hydrate in water at 0°C (CHC13 • 18H20, CAS registry
number 67922-19-4); the hexagonal crystal decomposes at 1.6°C (DeShon, 1979).
3-2
-------
TABLE 3-1. PHYSICAL PROPERTIES OF CHLOROFORM
Molecular weight
Melting point (°C)
Boiling point (°C)
Water-chloroform azeotrope ('
Specific gravity (25/4°C)
Vapor density (101 kPa, 0°C,
Vapor pressure
°C kPa
-30 1.33
-20 2.61
-10 4.63
0 8.13
10 13.40
20 21.28
30 32.80
40 48.85
Solubility in water
1C q/kq H2Q
0 10.62
10 8.95
20 8.22
30 7.76
Dc)
kg/m3)
torr
10.0
19.6
34.7
61.0
100.5
159.6
246.0
366.4
119.38
-63.2
61.3
56.1
1.48069
4.36
Log octanol/water partition coefficient: 1.97a
Conversion factors at 25°C and 1 atmb
1 ppm CHCls in air equals 4.88 mg/m3
1 mg/m3 CHCls in air equals 0.205 ppm
aHansch and Leo, 1979.
bCalculated.
SOURCE: DeShon, 1979.
3-3
-------
Chloroform is chemically stable to water, having a hydrolysis half-life of
3100 years in neutral water at 25°C (Mabey and Mill, 1978); the half-life of
chloroform in air from hydroxyl radical reactions is 78 days (Hampson, 1980).
The hydrogen atom in chloroform can be removed in the presence of warm
alkali metal hydroxide to form a trichloromethyl anion (DeShon, 1979). This
anion can condense with carbonyl compounds. Both wet and dry chloroform will
react with aluminum, zinc, and iron.
Small amounts of ethanol are used to stabilize chloroform from oxidation
during storage (DeShon, 1979).
3.3. SAMPLING AND ANALYSIS
3.3.1. Chloroform in Air
Chlorofom in air can be-analyzed by a-number of methods; however, the
method of Singh et al. (1980) appears to be substantially free of artifact
problems and completely quantitative. In this method, an air sample in a
stainless steel canister at 32 psig is connected to a preconcentration trap
consisting of a 4" x 1/16" ID stainless steel tube containing glass beads,
glass wool, or 3% SE-30 on acid washed 100/120 mesh chromosorb W. The
sampling line and trap, maintained at 90°C, are flushed with air from the
canister; then the trap is immersed in liquid 02 and air is passed through
the trap, the initial and final pressure being noted (usually between 30 and
20 psig) on a high-precision pressure gauge. The ideal gas law can be used
to estimate the volume of air passed through the trap. The contents of the
trap are desorbed onto a chromatography column by backflushing it with an
inert gas while holding the trap at boiling water temperature. An Ascarite
trap may be inserted before the chromatography column to remove water.
Suitable columns include 20% SP-2100 and 0.1% CW-1500 on Supelcoport (100/120
mesh, 6' x 1/8" stainless steel) and 20% DC-200 on Supelcoport (80/100 mesh,
3-4
-------
33' x 1/8" Ni). Both columns can be operated at 45°C with a carrier gas flow
of 40 ml/min on the former column and 25 ml/min on the latter. An electron
capture detector operating at 325°C was found to be optimum. It should be
noted that the above authors found Tenax to be unsuitable for air analyses
because of the presence of artifacts in the spectrum from oxidation of the
Tenax monomer. In addition, when Tenax is used as a sorbent, safe sampling
volumes (i.e., that volume of air which, if sampled over a variety of
circumstances, will not cause significant breakthrough) should be adhered to.
Brown and Purnell (1979) determined the safe volume for chloroform per gram
Tenax to be 9*3 L (flow rate 5-600 ml/min; CHC13 cone. <250 mg/m^; temp, up
to 20°C) with a safe desorption temperature of 90°C.
The .detection limits of-=this method were,not ..specified and are dependent
on the volume of air sampled. Analyses as low as 16 ppt have been reported
using this method (Singh et al., 1980).
3.3.2. Chloroform in Hater
Chloroform in water can be analyzed by the purge-and-trap method
(Method 502.1) as recommended by the Environmental Monitoring and Support
Laboratory of the U.S. EPA (1981a). In this method, an inert gas is bubbled
through 5 ml of water at a rate of 40 ml/minute for 11 minutes, allowing the
purgeable organic compounds to partition into the gas. The gas is passed
through a column containing Tenax GC at 22°C, which traps most of the
organics removed from the water. The Tenax column is then heated rapidly to
130°C and backflushed with helium (20-60 ml/minute, 4 minutes) to desorb the
trapped organics. The effluent of the Tenax column is passed into an
analytical gas chromatography column packed with 1% SP-1000 on Carbopack-B
(60/80 mesh, 8'x 0.1" ID) maintained at 40°C. The column is then temperature
programmed starting at 45°C for 3 minutes and increasing at 8°C/minute until
3-5
-------
220°C is reached; it is then held there for 15 minutes or until all compounds
have eluted. A halogen-specific detector (or GC-MS) having a sensitivity of
0.10 pg/L with a relative standard deviation of <10% must be used.
3.3.3 Chloroform in Blood
Chloroform in blood can be analyzed by using a modified purge-and-trap
method (Pellizzari et al., 1979). This method involves diluting an aliquot
of whole blood (with anticoagulant) to =50 ml with prepurged, distilled
water. The mixture is placed in a 100 ml 3-neck round bottom flask along
with a Teflon-lined magnetic stirring bar. The necks of the flask are
equipped with a helium inlet, a Tenax trap, and a thermometer. The Tenax
trap is a 10 cm x 1.5 cm ID glass tube containing pre-extracted (Soxhlet,
methanol, 24 hr) and conditioned (270°C, 30 ml/min helium flow, 20 min) 35/60
mesh Tenax (-1.6 g, 6 cm). The sample is then heated to 50°C and purged with
a helium flow rate of 25 ml/min for 90 min. Analysis can be performed as
indicated in Section 3.3.2.
3.3.4. Chloroform in Urine
Chloroform in urine can be analyzed by using an apparatus identical to
the one described in Section 3.3.3, using 25 ml of urine diluted to 50 ml
instead of blood.
3.3.5. Chloroform in Tissue
Chloroform in tissue can be analyzed by using an apparatus identical to
the one described in Section 3.3.3, using 5 g of tissue diluted to 50 ml and
macerated in an ice bath instead of blood. The purge time is reduced to
30 minutes.
3-6
-------
3.4. EMISSIONS FROM PRODUCTION AND USE
3.4.1. Emissions from Production
3.4.1.1. Direct Production—Chloroform is produced commercially in the
United States by two methods, chlorination of methane and chlorination of
methyl chloride produced from methanol and hydrogen chloride (Wagner et al.,
1980; DeShon, 1979). The chemistry is summarized by the following reactions:
Methane Chlorination
4CH4 + 10C12 - CH3C1 + CH2C12 + CHC13 + CC14 + 10HC1
Methanol Hydroch1 orination
Catalyst
HC1 + CH3OH - CHoCl + H20
280 - 350°C
3CH3C1 + 6C12 •* CH2C12 + CHC13 + CC14 + 6HC1
The methanol process has been reported to account for 74% of capacity, with
methane accounting for only 26% of capacity (SRI International, 1983).
Moreover, the methanol process is believed to account for an ever increasing
proportion of capacity.
In the chlorination of methane, natural gas is directly chlorinated in
the gas phase with chlorine at 485-510°C (Anthony, 1979). The product
mixture contains all chlorinated methanes, which are removed by scrubbing and
separated by fractional distillation.
In the second process, gaseous methanol and HC1 are combined over a hot
catalyst to form methyl chloride (Ahlstrom and Steele, 1979). The methyl
chloride is then chlorinated with chlorine to produce CH2C12, CHC13, and CC14
(DeShon, 1979). The chlorination conditions for both processes can be
adjusted to optimize chloroform production.
3-7
-------
United States production is carried out by four manufacturers at six
sites summarized in Table 3-2. The annual production of chloroform in the
United States has risen from 35 million kg (77 million lb) in 1960 (DeShon,
1979) to >184 million kg (405 million Ibs) in 1981 (USITC, 1982) with few
declines. The production and capacity of the industry has recently declined,
however (see Table 3-2.)
3.4.1.1.1. Chloroform emissions from the methane chlorination process.
Rem et al. (1982) reported that air emissions could come from process vents,
in-process and product storage tanks, liquid waste streams, secondary
emissions (handling and disposal of process wastes), and fugitive emissions
from leaks in process valves, pumps,-compressors and pressure relief valves.
The emission factors they calculated were based on a typical methane
chlorination facility as reported by U.S. EPA (1980a) having a total chloro-
methane capacity of 2 x 10^ metric tons (441 x 10" lb), operating
continuously (8760 hr/year), and having a product mix of 20% CH3C1, 45%
CH2C12, 25% CHC^, and 10% CCl^. The emission factor for the uncontrolled
recycle methane inert gas purge vent for the above plant (0.014 kg/metric
ton) was calculated from an hourly CHC^ emission rate of 0.071 kg/hr
reported by Dow Chemical Company for a 46,000 metric ton/year facility
assuming continuous (8760 hr/year) operation (Letter from J. Beale, n.d., Dow
Chemical U.S.A., Midland, Michigan, to L. Evans, Emissions Standards and
Engineering Division, U.S. EPA, concerning Dow facility at Freeport, TX;
cited in Rehm et al., 1982). The uncontrolled emission factor for the
distillation area emergency inert gas vent (0.032 kg/metric ton) was
calculated from an emission factor for volatile organic compounds (VOC) of
0.20 kg/metric ton of total chloromethane production and composition data
showing chloroform to be 4% of VOC (U.S. EPA, 1980a). In-process and product
3-8
-------
storage emissions (0.91 to 0.80 kg/metric ton, depending on controls) were
calculated from emission equations for breathing and working losses from
AP-61 (U.S. EPA, 1981b) assuming tanks to be half-full, have 95% emission
TABLE 3-2. CHLOROFORM PRODUCERS, PRODUCTION SITES, AND CAPACITIESa
Producer
Production
site
Capacity
Millions of kg
(Millions of Ib)
Process
Diamond
Shamrock
Dow Chemical
Linden Chemicals
and Plastics, Inc.
Vulcan
Materials Co.
Belle, WV
Freeport, TX
Plaquemine, LA
Moundsville, WV
Geismar, LA
Wichita, KS
18 (40)
33 (74)
27 (60)
14 (30)
27 (60)
50 (110)
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol and
Methane
TOTAL
169 (374)
producer, Stauffer Chemical Co., which was included in the SRI study, has
been deleted from this table because it reportedly is no longer producing. The
capacities and processes for Dow Chemical have also been revised in the table,
based on information received from Dow Chemical U.S.A. The totals above reflect
these changes.
SOURCE: SRI International, 1983.
controls when present, and a 12°C diurnal temperature variation (U.S. EPA,
1980b). Rehm et al. (1982) calculated the total chloroform emissions to air
to be 70.2 metric tons (155 x 10^ Ib) by multiplying the appropriate factors
by plant capacity use after including secondary emissions (0.21 kg/metric
ton) and fugitive emissions (5.5 kg/hr).
3-9
-------
Releases of chloroform to water come from scrubbers, neutralizes, and
cooling water (Rem et al., 1982). Based on a 300 ppm CHC13 content in total
wastewater discharges averaging 68 L/min and assuming that 90% would
volatilize, Rehm et al. (1982) calculated a release factor of 0.023 kg/metric
ton. They then added to this the emission from indirect contact cooling
water (100 ppm, 5800 L cooling water/metric ton CHC^, 90% evaporation) to
calculate a release rate of 3.3 metric tons of CHC^ (7.3 x 10^ Ib) per year
to water.
No quantifiable data were available to Rehm et al. (1982) regarding
release of chloroform to land from methane chlorination.
3.4.1.1.2. Chlorination emissions from the methanol hydrochlorination-meth.yl
chloride chlorination process; • Chloroform emissions to-air come from process
vents, in-process and product storage tank emissions, and fugitive emissions
from leaks in valves, pumps, compressors, and pressure relief valves (Rem et
al., 1982). Rehm et al. (1982) used an uncontrolled emission factor reported
by Vulcan Materials Company for process vents (Hobbs, 1978), and assumed
continuous (8760 hr/year) operation, and controls sufficient to reduce
emissions 80% to obtain the emission factor for controlled process vents
(0.003 kg/metric ton for controlled; 0.015 kg/metric ton for uncontrolled).
Storage emissions (0.176 kg/metric ton for controlled, 0.88 kg/metric ton for
uncontrolled) were calculated from Hobbs (1978) and from emission equations
for breathing and working losses from AP-42 (U.S. EPA, 1981b), assuming tanks
to be half-full, have 80% emission reduction controls, and a 12°C diurnal
temperature variation (U.S. EPA, 1980b). Fugitive emission factors
(3.32 kg/hr for uncontrolled; 1.08 kg/hr for controlled) for volatile organic
compounds were used (U.S. EPA, 1980c) along with a control factor of 67.5%
based on leak detection and repair (U.S. EPA, 1980b). Emission rates were
3-10
-------
then calculated to be 196 metric tons (432 x 103 Ib) based on plant capacity,
capacity use (66%), the level of emission controls used at each plant, and
continuous operation (8760 hr/year).
Rem et al. (1982) assumed that releases of chloroform to water from
methyl chloride chlorination resulted from contamination of cooling water and
from spent acid and spent caustic streams. For cooling water contamination,
they assumed minor spills and leaks resulted in contamination of 100 mg
chloroform per liter of cooling water, that 5800 L of cooling water per
metric ton of chloroform produced was consumed, and that 90% of the
chloroform evaporated. For spent acid and spent caustic streams they assumed
that 0.04 kg of chloroform is released for every metric ton of chloromethane
produced. They further stated that this release factor was not considered to
be very reliable (and in fact industry commenters have suggested that no
measurable losses of chloroform are observed under normal operating
conditions). However, in the absence of better data Rehm et al. (1982) used
this factor to approximate emissions. In addition, they assumed that one
third of the chloromethane production consists of chloroform and that 90% of
the released chloroform evaporates. Using these assumptions, Rehm et al.
(1982) calculated that 0.070 kg of chloroform is released to water per metric
ton of chloroform produced, or 8.0 metric tons (17.6 x 10^ Ib) of chloroform
were released to water based on 1980 production levels.
Rem et al. (1982) reported that the bottoms from chloroform distillation
in the methyl chloride chlorination process are the feed for carbon
tetrachloride and perchloroethylene production, and that during their
production a residue is formed that contains chloroform. This residue is
landfilled or deep-well injected. This represents the only known release of
chloroform from carbon-tetrachloride/perchloroethylene production. (Dow
3-11
-------
Chemical U.S.A. has, to the contrary, reported that the residue is instead
recycled as a raw material feed to production, with no measurable amount of
chloroform exiting from the plant.)
Rem et al. (1982) assumed that 1.02 kg of residue is produced per metric
ton of chloroform from methyl chloride production (see Wagner et al., 1980).
They further assumed (as did Wagner et al., 1980) that 18.4% of the residue
is chloroform and that 25% is landfilled (again, these figures have been
questioned by Dow Chemical U.S.A.). This results in a release factor of
0.047 kg chloroform per metric ton of chloroform produced, or 5.4 metric tons
(12 x 103 Ib) based on 1980 production.
3.4.1.1.3. Summary of direct production. Direct production emits some
283 metric tons into the environment on an annual basis. Greater than 95%
(266.2 metric tons) of this is emitted into the air. Direct production
accounts for =3% of all environmentally released chloroform, and =3.6% of
all chloroform released to air. Table 3-3 summarizes chloroform discharges
from direct production.
3.4.1.2. Indirect Production
3.4.1.2.1. Chloroform formation during ethylene dichloride production.
Ethylene dichloride (EDC) is produced by two methods, direct chlorination and
oxychlorination, and is used principally for vinyl chloride monomer (VCM)
production. (See EPA Health Assessment Document for EDC, EPA 600/8-84-006F.)
A combination of the two methods is used by most VCM production facilities in
a process known as the balanced process since the HC1 from the
dehydrochlorination of EDC is used to produce more EDC from ethylene, the
major products of the overall reaction being VCM and H20.
3-12
-------
TABLE 3-3. ESTIMATED CHLOROFORM DISCHARGES FROM DIRECT SOURCES
Environment release (metric tons/year)a
Source Air Water
Methyl chloride
chlorination 196 8
Methane chlorination 70.2 3.3
Total: 266.2 11.3
Land Total
5.4 209.4
73.5
5.4 282.9
figures may now be high due to reduced industry capacity, and
conversion to the methanol process by all but one production site.
SOURCE: Rehm et al. (1982).
Direct Chlorination
CH2 = CH2 + C12 + CH2C1CH2C1
Oxych1 orination
catalyst
2 HC1 + 1/2 02 C12 + H20
CH2 = CH2 + C12 CH2C1CH2C1
Balanced Process
CH2 = CH2 + C12 - CH2C1CH2C1
CH2 = CH2 + 2HC1 + 1/2 02 CH2C1CH2C1 + H20
A
HC1 + CH2 = CHC1 «-
3-13
-------
Chloroform is formed as a by-product during EDC manufacture. Rehm et
al. (1982) estimated chloroform emissions to air from EDC production based on
emission factors developed by the EPA during field studies of domestic EDC
production facilities (see U.S. EPA, 1980d). Domestic production facilities
are listed in Table 3-4. Emission sources (e.g., process vents, fugitive
emissions, storage) for each plant, plant capacity, capacity utilization
(56%), and control technology were all combined with the emission factors to
determine the overall chloroform emissions at the current level of control.
According to their calculations, chloroform emissions to the atmosphere are
=760 metric tons/year (1675 x 103 Ib/year).
Chloroform releases to water may occur during the discharge of
wastewater from the process; however, the amount of chloroform present is
unknown.
Chloroform releases to land from EDC production reportedly occur when
the light ends from EDC distillation are landfilled (Rem et al., 1982). An
estimated 217 metric tons (478 x 103 Ib) were landfilled in 1980. (This
estimate may be high; Dow Chemical U.S.A. has commented, for example, that
the landfilling of both light and heavy ends from EDC distillation has not
occurred since the 1979s.)
3.4.1.2.2. Chlorination of drinking water. Chloroform in drinking water
arises when humic substances or methyl ketones (e.g., acetone) in water react
with a hypochlorite anion (Stevens et al., 1976; NAS, 1978). Hypochlorite is
the principal reactant in chlorinated water above pH 5. Chloroform is
produced by the haloform reaction outlined below.
R-COCH3 + 30Cr RCOCC13 + 30H~
R-COCC13 + OH' RCOQ- + CHC13
3-14
-------
TABLE 3-4. ETHYLENE DICHLORIDE PRODUCERS, PRODUCTION SITES,
AND CAPACITIES
Capacity
Producer
Alantic Richfield Co.
ARCO Chem. Co. Div.
Borden, Inc.
Borden Chem. Div.
Petrochems. Div.
Diamond Shamrock
Dow Chemical . U.S.A.
E.I. du Pont de Nemours and Co.,
Conoco Inc., subsid.
Conoco Chems. Co. Div.
Ethyl Corp.
Chems. Group
Formosa Plastics Corp. U.S.A.
Georgia-Pacific Corp.
Chem. Div.
The BF Goodrich Co.
BF Goodrich Chem. Group
Convent Chem. Corp., subsid.
Production
site
Port Arthur, TX
Geismar, LA
Deer Park, TX
Freeport, TX
Oyster Creek, TX
Plaquemine, LA
Inc.
Lake Charles, LA
Baton Rouge, LA
Pasadena, TX
Baton Rouge, LA
Point Comfort, TX
Plaquemine, LA
La Porte, TX
Calvert City, KY
Convent, LA
millions of kg
(millions of Ib)
204
231
145
726
476
862
524
318
102
249
386
748
719
454
363
(450)
(510)
(320)
(1600)
(1050)
(1900)
(1155)
(700)
(225)
(550)
(850)
(1650)
(1585)
(1000)
(800)
PPG Industries, Inc.
Chems. Group Chem. Div.
Lake Charles, LA 1225 (2700)
3-15
-------
TABLE 3-4. (continued)
Producer
Production
site
Capacity,
millions of kg
(millions of lb)
Shell Chem. Co.
Union Carbide Corp.
Ethylene Oxide Derivatives Div.
Vulcan Materials Co.
Vulcan Chems. Div.
Deer Park, TX
Norco, LAb
Taft, LAb
Texas City, TXb
Geismar, LA
TOTAL
635
544
68a
68
159
(1400)
(1200)
(150)
(150)
(350)
9,206 (20,295)
aCaptive use only.
bReportedly now shut down.
SOURCE: SRI International, 1983.
Rem et al. (1982) used the data from the National Organics
Reconnaissance Survey (NORS) (Symons et al., 1975) and the National Organics
Monitoring Survey (NOMS) (U.S. EPA, n.d.) to estimate the concentration of
chloroform in drinking water. These surveys provided information on
chloroform concentrations in 137 U.S. cities. To determine,the amount of
chloroform generated, the authors multiplied the volume of water treated by
each city by the chloroform concentration in the drinking water. The amount
of chloroform generated by each city was then summed and divided by the
volume of water generated to give a weighted average concentration of
41 ng/L. This was then multiplied by the estimated volume of water
chlorinated annually (4.6 x 1013 L/year) to yield the amount in the U.S.
(1900 metric tons, 4.2 x 106 lb). Rehm et al. (1982) stressed that this
3-16
-------
value probably represents a minimum since NORS was conducted during the
winter (hence, chloroform levels were low) and NOMS samples were iced when
taken (hence, may be lower than if allowed full contact time). Thus, the
actual value may be higher than this estimate.
3.4.1.2.3. Chlorination of municipal sewage. Chlorination of municipal
sewage results in increased chloroform concentrations (MAS, 1978). Municipal
wastewater generally contains a lower concentration of chloroform precursors
(humic materials) than do ambient waters; therefore, the amount of chloroform
generated from wastewater Chlorination is smaller (Wagner et al., 1980).
Rehm et-al. (1982) calculated chloroform production from wastewater treatment
based on analyses of the seoendary effluent from 28 municipal plants
published by the EPA (U.S. EPA, 1979a). These analyses showed that the
average chloroform concentration increased by 9 pg/L, from 5 to 14 yg/L.
Rehm et al. (1982) assumed that all municipal wastewater was chlorinated and
multiplied the average concentration increase by the municipal wastewater
flow (9.7 x 1010 L/day) listed in U.S. EPA (1981c). By this method, 320
metric tons (0.7 x 10^ Ib) were calculated to be produced annually.
3.4.1.2.4. Chlorination of cooling waters. Cooling water used in electric
power generating plants is treated with chlorine as a blocide to prevent
fbuling intake screens and condensers in both once-through and closed cycle
systems (U.S. EPA, 1980e). Rehm et al. (1982) calculated chloroform
production based on the size of the average power plant (hence, the volume of
water required), the type of cooling system used (once-through or
recirculating), and the fact that 65% of all power plants chlorinate cooling
water. They calculate that 72 metric tons of chloroform (160 x 103 Ib) are
discharged directly into water by once-through systems and .190 metric tons of
chloroform (420 x 103 Ib) are emitted into the air by recirculating systems.
3-17
-------
A total of 262 metric tons of chloroform (580 x 103 Ib) are produced annually
from cooling water chlorination.
3.4.1.2.5. Chlorination in the pulp and paper industry. Pulp and paper
mills emit more chloroform to the environment than any other single source.
Chloroform is produced during the bleaching of wood pulp, a process that
whitens the final paper product. Rehm et al. (1982) based their estimation
on information contained in a number of documents concerned with the pulp and
paper industry (U.S. EPA, 1980f; NCASI, 1977; Metcalf and Eddy, Inc., 1972;
TAPPI, 1963). From the operating conditions, analytical data, and production
steps detailed in these documents, Rehm et al. (1982) determined the quantity
of chloroform produced for each of nine different types of mills for which
monitoring data existed and applied these values to all mills for which no
values from sampling existed. The authors determined that chloroform is
emitted at three different stages: into the air during the bleaching process
itself; into the air during the detention time in wastewater treatment
plants; and into the water from treatment plant effluent. The amount of
chloroform produced annually was calculated to be 128 metric tons (282 x
103 Ib) released to air during bleaching operations, 3985 metric tons (8.78 x
10° Ib) released to air-from wastewater detention-(4113-jnetric tons to air),
and 298 metric tons (657 x 103 Ib) discharged into water from treatment
plants (4411 metric tons or 9.72 x 106 Ib total).
3.4.1.2.6. Chloroform production from combustion of leaded gasoline.
Chloroform has been reported to be a component of automobile exhaust (Harsch
et al., 1977). Its presence is reportedly the result of using ethylene
dichloride and ethylene dibromide as lead scavengers in leaded gasoline
(Lowenbach and Schlesinger Associates, 1979). Rehm et al. (1982) cite other
sources which state that chloroform is not formed during the combustion of
3-18
-------
leaded fuel containing ethylene dichloride. The Emissions Testing and
Characterization Branch of EPA's Environmental Sciences Research Laboratory
measured chlorocarbon emissions from automobiles using leaded gasoline and
found no chloroform. Rehm et al. (1982) then reason that even if chloroform
were present in automobile exhaust, the decrease in the usage of leaded
gasoline will decrease the amount of chloroform produced.
The authors cite an EPA report (U.S. EPA, 1982a) which states that
leaded gasoline consumption is expected to drop by >75%, from 34 x 10^
gallons to 8.3 x 109 gallons/year. Rehm et al. (1982) used the estimate of
Wagner et al. (1980) that 1% of the ethylene dichloride in gasoline would be
converted to ch^roform, and the new lead phase-down regulations (U.S. EPA,
1982b) to calculate an annual emission rate of 180 metric tons (397 x 103 Ib)
in 1983 and 44 metric tons (97 x 103 Ib) by 1990.
3.4.1.2.7. Chloroform formation during atmospheric trichloroethylene
decomposition. Trichloroethylene is a major industrial solvent used
principally for vapor degreasing of fabricated metal parts (66%) (Chemical
Marketing Reporter, 1981), and the majority of each year's production is used
for replacement of evaporative loss to the environment (Rem et al., 1982).
The postulated formation of chloroform during the atmospheric decomposition
of trichloroethylene is based on laboratory experiments in which
trichloroethylene, N0Ł, h^O, and a hydrocarbon mixture were irradiated with
light having the intensity and spectral distribution of the lower troposphere
(U.S. EPA, 1976). Dichloroacetylchloride, phosgene, chloroform, and HC1 were
detected as products. Rehm et al. (1982) did not describe the method they
used to determine the amount of chloroform produced from this reaction;
however, they estimated 780 metric tons of chloroform are produced annually.
3-19
-------
3.4.1.2.8. Miscellaneous indirect source. Wagner et al. (1980) have listed
a number of other indirect sources of chloroform that defy quantification.
These sources are: chlorination wastewaters from the textile industry, the
food processing industry, breweries, combustion of tobacco products treated
with chlorination pesticides, thermal decomposition of plastics, biological
production in red marine algae, and the reaction of chlorinated pollutants
with humic substances in natural waters.
3.4.1.2.9. Summary of indirect production. Chloroform is produced and
emitted into the environment from a variety of indirect sources. These
sources account for -84% of all chloroform air emissions, 98% of water
discharges, and 36% of all land-discharges (84.6% of all environmental
releases). The environmental discharges of chloroform from indirect sources
are summarized in Table 3-5.
3.4.2. Emissions from Use
Chloroform is used consumptively for the production of
chlorodifluoromethane or Fluorocarbon-22 (accounts for 90% of domestic 1982
production, 60% for refrigerants, 30% for fluoropolymer) and for exports (3%
in 1982) (Chemical Marketing Reporter, 1983). It is used nonconsumptively as
an extraction solvent; as a solvent for penicillin, alkaloids, vitamins,
flavors, lacquers, floor polishes, artificial silk manufacture, resins, fats,
greases, gums, waxes, adhesives, oils, and rubber; as a dry cleaning agent;
as an intermediate in pesticide and dye manufacture; and as a fumigant
ingredient (Rem et al., 1982; DeShon, 1979; Merck Index, 1976). The great
majority of chloroform used nonconsumptively is emitted into the environment
since (except for expansions) the chloroform purchased for these uses is
make-up solvent used to replace that amount not recovered from processes
(Wagner et al., 1980).
3-20
-------
Source
TABLE 3-5. CHLOROFORM DISCHARGES FROM INDIRECT SOURCES
Environmental release (metric tons/year)
Air Water Land Total
Pulp and paper mills
Drinking water cMorination
Ethylene dichloride
manufacture
Trichloroethylene
photodegradation
Municipal wastewater
4113
0
298
1900
0
0
217
0
4411
1900
977
780
chlorination
Cooling water chlorination
Automobile exhaust
TOTAL
0
190
180
6023
320
72
0
2590
0
0
0
217
320
262
180
8830
aMinor releases possible.
SOURCE: Rehm et al., 1982.
3.4.2.1. Emissions from Pharmaceutical Manufacturing—Chloroform is used as
an extraction solvent during the manufacture of some antibiotics and
steroids, and during the manufacture of certain other biological and natural
Pharmaceuticals (Rem et al., 1982). It is also used as a chemical
intermediate. Based on a Pharmaceutical Manufacturing Association (PMA)
Survey,. Rehm et al. (1982) reported that =1000 metric tons of chloroform
are released into the environment (no year was specified; the PMA report was
dated 1978). The distribution was as follows: 57.0% (570 metric tons) to
air, 4.6% (46 metric tons) to water, and 38.4% (384 metric tons) to land.
3-21
-------
3.4.2.2. Emissions from F1uorocarbon-22 Production—The single largest use
of chloroform is for Fluorocarbon-22 production (Chlorodifluoromethane,
CHC1F2). Fluorocarbon-22 producers and production sites are listed in
Table 3-6. Chloroform release to the environment can occur from process
emissions, fugitive emissions, and storage emissions (Rem et al., 1982).
Rehm et al. (1982) reported that the first source listed does not represent a
significant source of chloroform emissions based on the design of
Fluorocarbon-22 production facilities and the process description.
Storage emission.estimates were based on U.S. EPA (1980g) for chloroform
feedstock storage in fixed roof tanks. Rehm et al. (1982) reported that the
Allied Chemical facility at El Segundo, California, uses a control system
that results in complete capture of chloroform vapors. Du Pont (and all
others by assumption) use refrigerated condensers that reduce the
uncontrolled emission factor of 2.5 kg/metric ton by 66%. Fugitive emissions
from leaks in valves, pumps, compressors, and relief valves were estimated to
result in an uncontrolled emission rate of 0.75 kg/metric ton. Total
emissions to air were calculated by Rehm et al. (1982) to be 139 metric
tons/year by multiplying the emission factors by 1980 Fluorocarbon-22
production (97,500 metric tons).
No estimate was made for emissions to wastewater because of a lack of
data. Emissions to land were based on the reported practice of landfilling
spent catalyst by Allied. Rehm et al. (1982) assumed a catalyst
contamination level of 10% and a total emission of 2.0 metric tons/year.
3.4.2.3. Emissions from Hypalon® Manufacture—Hypalon® is a chemically
resistant synthetic rubber made by substituting chloride and sulfonyl groups
onto polyethylene. The process involves dissolving polyethylene in
chloroform followed by reaction with chlorine and sulfur dioxide. Based on a
3-22
-------
TABLE 3-6. CHLORODIFLUOROMETHANE PRODUCERS AND PRODUCTION SITES
Production
Producer site
Allied Corp.
Allied Chem. Co. Baton Rouge, LA
Danville, IL
El Segundo, CA
E.I. du Pont de Nemours and Co., Inc.
Petrochems. Dept.
Freon® Products Div. Deepwater, NJ
Louisville, KY
Montague, MI
Pennwalt Corp.
Chems. Group
Fluorochemicals Div. Calvert City, KY
SOURCE: SRI International, 1983.
Du Pont report, Rehm et al. (1982) estimated that 54.9 metric tons of
chloroform were emitted into the air from Hypalon® manufacture in 1980, based
on an emission inventory conducted by the Texas Air Control Board, Austin,
Texas. No information was available for water or land emissions.
3.4.2.4. Chloroform Emissions from Grain Fumigation—Chloroform is a
registered pesticide for use on certain insects that commonly infest stored
raw bulk grains and is present in only one product (Rem et al., 1982). This
product, Chlorofume® FC.30 Grain Fumigant (Reg. No. 5382-15), marketed by
Vulcan Materials Company, contains 72.2% chloroform, 20.4% carbon disulfide,
and 7.4% ethylene dibromide. Originally registered in 1968, the EPA issued a
"Notice of Presumption Against Continued Registration of a Pesticide Product
— Chloroform (Trichloromethane)" in 1976 because of oncogenic effects in
rats and mice (U.S. EPA 1982c). Continued study resulted in returning it to
3-23
-------
the normal registration process (U.S. EPA, 1982c). Based on a personal
communication with D. Lindsay of Vulcan Materials, Rehm et al. (1982)
estimated that between 10,000 and 12,000 gallons of chloroform per year were
used in grain fumigants in the United States. Vulcan reported 1981 sales of
7000 gallons of Chlorofume® in 1981 or 5054 gallons (19,131 L). Based on its
density, 28,400 kg (28.4 metric tons) was released to the environment (air)
in this way.
3.4.2.5. Chloroform Losses from Loading and Transportation—Rem et al.
(1982) estimated chloroform .losses from loading ships, barges, tank cars, and
tank trucks. The method was based on the degree of chloroform saturation of
the air expelled from tanks during filling, temperature, vapor pressure,
control efficiency, and fi1 Ting methods as described by U.S. EPA (1979b) and
Environment Reporter (1982). The U.S. mode of transportation was taken from
Sax (1981) as follows: rail, 40.3%; barge, 47.8%; and truck, 11.9%. Loading
losses were calculated to be 40.9 metric tons (90,200 Ib).
Transit losses result from temperature and barometric pressure changes.
The losses were assumed to be the same for barges, tank trucks, and rail cars
and were estimated from the following equation:
LT = 0.1 PW
where LT = transit loss, Ib/week - 103 gal transported
P = true vapor pressure of transported liquid, psia
W = density of condensed vapors Ib/gal
No reference or justification for the use of the formula was presented (and
commenters have noted the calculation for losses may be high because the
method assumes a constant concentration throughout the vapor space). By this
method, and assuming 1 week transit time, Rehm et al. (1982) calculated that
3-24
-------
49.2 metric tons (0.11 x 106 Ib) chloroform were lost to the air using 1980
production values.
3.4.2.6. Miscellaneous Use Emissions--Rem et al. (1982) cited the previous
materials balance (Wagner et al., 1980) in predicting the emissions from
chloroform contamination of methyl chloride, methylene chloride, and carbon
tetrachloride. Chloroform is present to some extent in these products since
they are all made by the same process. Assuming a contamination level of 7.5
ppm, 17.5 ppm and 150 ppm for methyl chloride, methylene chloride, and carbon
tetrachloride, respectively, Rehm et al. (1982) estimate that releases to
air, land and water would be 9.8, 0.6, and 0.15 metric tons, respectively.
Chloroform is also used in a variety of products (see Section 3.4.2) and
as a general solvent. Rehm et al- (1982) estimate.that, while these uses are
generally declining, laboratory uses in particular may account for 8.5% of
production or 14,200 metric tons of chloroform. Rehm et al. (1982), however,
estimated the range of uncertainty to be + 50%, and industry commenters have
stated that laboratory use of chloroform is very minimal.
3.4.2.7. Summary of Chloroform Discharges from Use—Chloroform discharges
from manufacturing facilities that use chloroform as a process ingredient
account for 12% of all chloroform emissions to air, 1.8% of water discharges,
and 63% of all land discharges, or 12.7% of chloroform discharges to all
media. Table 3-7 summarizes chloroform discharges to all media.
3.4.3. Summary
Chloroform is produced by direct and indirect processes. Direct
production accounts for 184 million kg annually, while indirect production
accounts for =8.8 million kg annually.
Direct production of chloroform and processes associated with its use
(i.e., Fluorocarbon-22 production, Hypalon® manufacture, loading and transit
3-25
-------
TABLE 3-7. CHLOROFORM DISCHARGES FROM USE
Environmental Release (metric
Source
Pharmaceuticals
Chi orodif luorome thane
manufacture
Loading and transit losses
HypalorV9 manufacture
Grain fumigation
Secondary product
contamination
Total
Air
570
139
90.1
54.9
28.4
9.8
892.2
Water
46
__b
0
a
0
0.6
46.6
Land
384
2
0
__a
0
0.2
386.2
tons/year)
Total
1000
141
90.1
54.9
28.4
10.6
1325.0
^Minor releases possible
SOURCE: Rehm et al., 1982.
losses, grain fumigation, pharmaceutical use) emit some 1.6 million kg to the
environment. Virtually all of the indirectly produced chloroform may be
emitted into the environment; assuming this to be so, the total amount of
chloroform emitted is =10.4 million kg. This represents =5.6% of direct
production. The relative source contributions from all quantifiable sources
are listed in Table 3-8.
3.5. AMBIENT AIR CONCENTRATIONS
Monitoring data for a number of U.S. and world locations are presented
in Table 3-9. For the most part, ambient concentrations remain <1000 ppt
(1 ppt = lO"1-2, v/v), and some <10 ppt. There are notable exceptions,
however, although the reasons for this are not readily apparent.
3-26
-------
TABLE 3-8. RELATIVE SOURCE CONTRIBUTION FOR CHLOROFORM
oo
I
Environmental Release (metric tons/year)
Source
Pulp and paper mills
Drinking water chlorination
Pharmaceuticals
Ethylene dichloride manufacuture
Trichloroethylene photodegradation
Municipal wastewater chlorination
Cooling water chlorination
Methyl chloride chlorination
Automobile exhaust
Chlorodifluoromethane manufacture
Loading and transit losses
Methane chlorination
Hypalon® manufacture
Grain fumigation
Secondary product contamination
Laboratory usage0
TOTAL
Air
4113
0
570
760
780
0
190
196
180
139
90.1
70.2
54.9
28.4
9.8
7181.4
% of
Totala
39
0
5.5
7.3
7.5
0
1.8
1.9
1.7
1.3
0.9
0.7
0.5
0.3
0.1
68.8
Water
298
1900
46
___b
0
320
72
8
0
0
3.3
0
0.6
2647.9
% of
Totala
2.9
18
0.4
0
3.1
0.7
0.1
0
0
0.03
0
0.006
25.4
Land
0
0
384
217
0
0
0
5.4
0
2
0
0
0.2
608.6
% of
Total a
0
0
3.7
2.1
0
0
0
0.1
0
0.02
0
0
0.002
5.8
Total
4411
1900
1000
977
780
320
262
209.4
180
141
90.1
73.5
54.9
28.4
10.6
10,438a
% of
Total a
42
18
10
9.4
7.5
3.1
2.5
2.0
1.7
1.4
0.9
0.7
0.5
0.3
0.1
^Values are rounded.
bDashed lines indicate minor releases possible,
CNot included because of uncertainty-
SOURCE: Rehm et al. (1982).
-------
TABLE 3-9. AMBIENT LEVELS OF CHLOROFORM
Location
Alabama
Tuscaloosa
Talladega Forest
Arizona
Phoenix
Cal ifornia
Stanford Hills
Point Reyes
Los Angeles
Palm Springs
CO
r^o Yosemite
00
Mill Valley
Riverside
Badger Pass
Point Arena
Point Arena
Los Angeles
Oakland
Type of Site
urban
rural
urban
clean9
clean marine
urban
urban-
suburban
remote-
high altitude
clean marines
urban-
suburban
remote-
high altitude
clean marine
clean marine
urban
urban
Date
2/77
2/77
4-5/79
11/75
12/75
4-5/76
5/76
5/76
1/77
4-5/77
5/77
5/77
8-9/78
4/79
6-7/79
Analytical
Method
GC-ECD
GC-ECD
GC-
coulometry
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
Concentration (ppt, v/v)
Max.
3000
200
514.0
217
114
724
616
24
36
310
38
42
48
223.5
60.1
Min.
100
NR
27.1
12
15
23
20
12
4
24
2
8
12
24.3
13.1
Average
800
100
111.4
33
37
102
99
17
25
25
16
20
18
88.2
32.1
Reference
Holzer et al . ,
Holzer et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
Singh et al . ,
1977
1977
1981
1979
1979
1979
1979
1979
1979
1979
1979
1979
1979
1981
1981
Delaware
Delaware City
Kansas
Jetmore
NS
remote-
continental
7/74
6/78
GC-ECD
GC-ECD
34
<10 NR
16
Lillian et al., 1975a
Singh et al., 1979
(continued on next page)
-------
TABLE 3-9. (continued)
Location
Type of Site Date
Analytical Concentration (ppt, v/v) Reference
Method Max. Min. Average
CO
I
IN)
Maryland
Baltimore
Montana
Western Montana
Nebraska
Reese River
urban
remote
7/74
3/76
remote-
high altitude 5/77
New Jersey
Rutherford
Newark
Piscataway
Somerset (county)
Bridgewater
township
Bound Brook
Patterson
Clifton
Fords
Newark
Passaic
Hoboken
Seagrit
Seagrit
Sandy Hook
Sandy Hook
urban
urban
urban
urban
rural
urban
urban
urban
urban
urban
urban
urban
urban
urban
urban
urban
GC-ECD
GC-MS
GC-ECD
<10 <10 NR
NR NR
19
13
1978
1978
1978
1978
1978
3/76
3/76
3/76
3/76
3/76
3/76
3/76
6/74
6/75
7/74
7/75
GC-MS
GC-MS
GC-MS
GC-MS
GC-MS
GC-MS
GC-MS
GC-MS
GC-MS
GC-MS
GC-MS
GC-MS
GC-ECD
GC-ECD
GC-ECD
GC-ECD
31,000
7500
2900
11,000
NR
NR
NR
NR
NR
NR
NR
NR
60
50
63
55
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
<10
5
<10
10
4600
3900
2200
5000
NDb
854
768
1700
3422
7582
854
427
40
35
30
25
Lillian et al., 1975a
Cronn and Harsch, 1979
Singh et al., 1979
Bozzelli
and Kebbekus,
Bozzelli and
Kebbekus, 1979
Bozzelli and
Kebbekus
Bozzelli
Kebbekus
Bozzelli
Kebbekus,
Pellizzari
Pellizzari
Pellizzari
Pellizzari
Pellizzari
Pellizzari
Pellizzari
Lillian et
Lillian et
Lillian et
Lillian et
1979
1979
and
1979
and
1979
1977
1977
1977
1977
1977
1977
1977
al., 1975a
al., 1975b
al., 1975a
al., 1975b
(continued on next page)
-------
TABLE 3-9. (continued)
CO
I
Location
Bayonne
New York
Staten Island
New York City
New York City
White Face
Mountain
White Face
Mountain
Niagara Falls
Ohio
Wi Imington
Wilmington
Texas
Houston
Washington
Pullman
Pullman
England
Liverpool
/Manchester
Type of Site Date
urban
urban
urban
urban
remote
remote
urban
Air Force
Air Force
urban
rural
rural
suburban
7/75
3/76
6/74
6/75
9/74
9/75
NS
Base 7/74
Base 7/75
6-7/77
12/74
to 2/75
11/75
NS
Analytical Concentration
Method Max. Min.
GC-ECD
GC-MS
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-MS
GC-ECD
GC-ECD
GC-MS
GC-MS
GC-ECD
GC-ECD
15,000 <10
NR NR
480 <10
450 10
250 <10
350 6
21,611 215
4800 <10
5000 20
11,034 Trace
NR NR
I
NR NR
8C 3C
(ppt, v/v) Reference
Average
1030
4268
160
200
9
8
NR
340
480
NR
20
43
NR
Lillian et al . ,
Pellizzari, 1977
Li 1 1 ian et al . ,
Lillian et al . ,
Lil lian et al . ,
Lillian et al . ,
Pel 1 izzari
et al., 1979
Lillian et al . ,
Lillian et al . ,
Pellizzari
et al., 1979
Grimsrud and
Rasmussen, 1975
Rasmussen et al .
1975a
1975a
1975b
1975a
1975b
1975a
1975a
,1977
Pearson and McConnell,
1975
(continued on next page)
-------
TABLE 3-9. (continued)
Analytical Concentration (ppt, v/v) Reference
Location Type of Site Date Method Max. Min. Average
00
1
CO
1— »
Organochlorine urban
manufacturer
Moel Faman, urban
Flintshire
Rannoch Moor, urban
Argyllshire
Rural areas rural
Ireland
Cork urban
Japan
Kobe NS
Atlantic Ocean
Northeast Atlantic
(Cape Blanc to
Lands End)
34°19'N 13°32'W to
49°54'N 05°54'W
NS GC-ECD 40C <10.1C NR Pearson and McConnell,
1975
NS GC-ECD 0.4C <0.1C NR Pearson and McConnell,
1975
NS GC 0.5C O.lc NR Murray and
NS GC 1.2 0.82 0.82 Murray and
1974 GC-ECD NR NR 26.5 Cox et al . ,
NS GC-ECD 9400 300 Okuno et al
7-8/72 GC 0.96 0.14 0.35 Murray and
Riley,
Riley,
1976
., 1974
Riley,
1973
1973
1973
aSubject Urban Transport.
bDetection Limits 10 ppt.
Cppb by mass.
NR = not reported; NS = not specified; ND = not detected.
-------
Singh (1977) and Singh et al. (1979) have determined northern and
southern hemisphere background concentrations as well as a global average.
The hemispheric values they reported are 14 ppt for the northern hemisphere
and <3 ppt for the southern hemisphere. This difference in hemispheric
values suggests that the oceans are not a significant source of chloroform,
but rather, that chloroform, for the most part, is anthropogenic. More
recent research, however, suggests that just the opposite may be true.
Khali 1 et al. (1983) have analyzed global concentrations of chloroform in the
air and seawater, and concluded that the tropical oceans, at least, are a
sizeable source of chloroform to the atmosphere. The reported atmospheric
concentrations of chloroform .tanged from 16 ppt at the South Pole to 45 ppt
at Cape Meares, Oregon, with a ratio of concentrations between the northern
and southern hemispheres of 1.6. Khali 1 et al. (1983) discuss the sources
and sinks of chloroform in the atmosphere, construct a global mass balance
equation, and conclude that chloroform in the atmosphere may be largely
natural in origin, rather than mostly anthropogenic as previously thought.
An interesting point not presented in Table 3-9 is that chloroform
concentrations above an inversion layer are significantly lower than
concentrations below it. In Wilmington, OH,.above an inversion layer, the
chloroform concentration was <10 ppt, whereas below it the concentration was
120 ppt (Lillian et al., 1975a).
3.6. ATMOSPHERIC REACTIVITY
The principal atmospheric reactant responsible for the removal of
chloroform is probably the hyaroxyl radical (Atkinson et al., 1979; Graedel,
1978; Altshuller, 1980; Singh, 1977; Crutzen and Fishman, 1977). Hydroxyl
radicals are formed in the lower atmosphere in at least two ways, first, by
the photodissociation of ozone (\ <310) into 0 (^-D) atoms (Atkinson et al.,
3-32
-------
1979).These go on to react with either water, hydrogen or methane to form
hydroxyl radicals. The second important source of hydroxyl radicals is the
reaction of hydroperoxyl radicals with nitric oxide.
Hydroxyl radical reactions probably follow the course outlined below
(Graedel, 1978):
CHC13 + HO -CC13 + H20
•CC13 + 02 *02CC13
•02CC13 + NO -OCC13 + N02
•OCC13 COC12 (phosgene) + Cl-
COC12 + H20 C02 + 2 HC1
Pearson and McConnell (1975) found HC1 and C02 ,as. .the only products of
chloroform irradiation with UV (\ >290 nm) light. The half-life reported by
these workers (23 weeks) was of the same order of magnitude as that
calculated from the hydroxyl radical rate constant (11.5 weeks) (Singh et
al., 1981).
Chloroform will not react photolytically in the troposphere; the UV
cutoff for chloroform is 175 nm (i.e., it will not absorb light >175 nm).
Callahan et al. (1979) calculated .that roughly.1% of the tropospheric
chloroform would diffuse eventually into the stratosphere, based on a
lifetime of 0.2-0.3 years and a troposphere-to-stratosphere turnover time of
30 years.
3.7. ECOLOGICAL EFFECTS/ENVIRONMENTAL PERSISTENCE
3.7.1. Ecological Effects
3.7.1.1. Terrestrial--Oata on the terrestrial ecological effects of
chloroform are not available. Significant effects are not expected because
chloroform is quite volatile and does not accumulate in terrestrial (or
3-33
-------
aquatic) environments, and is diluted rapidly and degraded to low
concentrations in the troposphere (HAS, 1978). Conceivably, acute effects on
wildlife can occur in the vicinity of major chloroform spills, but
significant chronic effects from long-term exposure to low ambient levels is
unlikely.
3.7.1.2. Aquatic--The toxicity of chloroform to aquatic organisms has
been reviewed by the U.S. EPA (1980h). As summarized in Table 3-10, two
freshwater fish (rainbow trout, bluegill) and one invertebrate (Daphnia
magna) species have been acutely tested under standard conditions; LCgQ
concentrations ranged from 28,900 to 115,000 ng/L (Bentley et al., 1975;
U.S. EPA, 1978a), and the trout was more sensitive than the bluegill. With
stickleback, goldfish, and orange-spotted-sunfish, anesthetization or death
occurred after exposure to 97,000 to 207,000 pg/L chloroform for 30 to 90
minutes (Clayberg, 1917; Jones, 1947a; Gherkin and Catchpool, 1964). Only
one test has been conducted with chloroform and saltwater organisms; the 96-
hour LC50 for the pink shrimp was 81,500 p.g/L (Bentley et al., 1975).
Embryo-larval tests with rainbow trout at 2 levels of hardness provided
27-day LC50 values of 2030 and 1240 ng/L (Birge et al., 1979). There was a
40% incidence of teratogenesis in the embryos at hatching (23 day exposure at
10,600 pg/L).
Bluegills bioconcentrated radiolabeled chloroform by a factor of 6 after
a 14-day exposure, and the tissue half-life was <1 day (U.S. EPA, 1978a).
This degree of bioconcentration and short biological half-life suggest that
chloroform residues would not be an environmental hazard to consumers of
aquatic life (U.S. EPA, 1980h).
3-34
-------
TABLE 3-10. ACUTE AND CHRONIC EFFECTS OF CHLOROFORM ON AQUATIC ORGANISMS
CO
I
CO
en
Species
Cladoceran
Daphm'a magna
Rainbow trout
Salmo qairdneri
Rainbow trout
Salmo qairdneri
Bluegill
Lepomis macrochirus
Bluegill
Lepomis macrochirus
Pink shrimp
Penaeus duorarurtr
Orangespotted sunfish
Lepomis humilis
Goldfish
Carassius auratus
Threespine stickleback
Gasterosteus aculeatus
Duration
48-hr
96-hr
96-hr
96-hr
96-hr
96-hr
1-hr
30-60 min
90-min
Concentration
(yg/L)
28,900
66,800
43,800
115,000
100,000
81,500
106,890 to
152,700
97,000 to
167,000
207,648°
Method
S,Ua
S,U
S,U
s,u
s,u
s,u
NS
NS
NS
Effect Reference
LCso U.S. EPA, 1978a
LC50 Bentley et al.,
LC50 Bentley et al.,
LCtjQ Bentley et al . ,
LC50 Bentley et al.,
LC^Q Bentley et al . ,
death Clayberg, 1917
50% Cherkin and
anesthetized Catchpool, 1964
anesthesia Jones, 1947a
with recovery
1975
1975
1975
1975
1975
Ninespine stickleback NS
Punqitius punqitius
148,320 to NS
296,640
Avoidance Jones, 1947b
-------
TABLE 3-10 (continued)
CO
en
Species
Rainbow trout
(embryo-larval)
Salmo qairdneri
Rainbow trout
(embryo-larval)
Salmo qairdneri
Rainbow trout
(embryo)
Salmo qairdneri
Duration
27 daysd
27 daysd
23 days
Concentration
(H9/L)
2,030
1,240
10,600
Method
F,Me
F,Me
F,Me
Effect
Reference
LCgQ at Birge et al . ,
50 mg/L hardness
LC.0 at
200 mg/L
hardness
JQ%
teratogenesis
Birge et al . ,
Birge et al . ,
1979
1979
1979
aStatic test, unmeasured concentration.
^Saltwater species.
^Corrected from vol/vol to pg/L.
dExposures began within 20 minutes of fertilization and ended 8 days after hatching.
eFlow-through test, measured concentration.
hr = hour; min = minutes; NS = Not stated.
SOURCE: U.S. EPA, 1980h
-------
3.7.2. Environmental Persistence
A number of researchers have reported the dominance of hydroxyl radical
oxidations in the fate of chloroform in the atmosphere (see Section 3.6).
Singh et al. (1981) calculated an atmospheric residence time for chloroform
based on the NASA reviewed rate constant reported by Hampson (1980). They
reported a 116-day (16.6-week) residence time for a hydroxyl radical
concentration of 106 molecules/cm3. This compares well to the observed 33-
week lifetime of chloroform in a sunlit flask (Pearson and McConnell, 1975).
This lifetime was based on experiments conducted in northwest England, which
receives less Intense sunlight than most of the U.S., and may account for its
longevity.
According to recent hydroxyl radical measurements, tropospheric ambient
air concentrations range from = 10^ to lO'7 molecules/ml (Atkinson et al.,
1979); models of the troposphere have suggested a concentration ranging
between 2 to 6 x 105 molecules/m. (Crutzen and Fishman, 1977; Singh, 1977).
Table 3-11 summarizes the literature values for kg^, the temperature of
measurement, and the calculated lifetime based on the indicated hydroxyl
radical concentration. If a hydroxyl radical concentration of 2 x 106
molecules/cm^ (typical for summer,months; winter concentrations are lower)
(Singh et al., 1981) is assumed, most of the lifetimes calculated from the
rate constants range between 0.2 to 0.5 years (69 to 181 days, 122 average).
Mabey and Mill (1978) critically reviewed hydrolysis data available in
the literature. They determined that chloroform had a hydrolysis half-life
of >3,000 years at pH 7 and 298 K. This is based on a base hydrolysis rate
of 0.602 x 10~4 and a OH~ concentration of 10~7 in neutral water.
Dilling et al. (1975) and Dilling (1977) determined the volatilization
half-life of chloroform from water. For a 1 ppm chloroform solution stirred
3-37
-------
TABLE 3-11. VALUES FOR k
OH
CO
1
CO
00
kOH x 1014
cm3 molecule-1 sec-1
6.51
10.1
16.8
11.4
6.4
7.4
10
K
265
296
298
298
265
273
298
[OH] Lifetime
x 10-5 (yeart)
4 1.2
10 0.19
9 0.56
Reference
Singh et al . ,
Howard and
Evenson, 1976
1979
Cox et al., 1976
Davis et al . ,
Hampson, 1980
1976a
b
a(4.69 + 0.71) x ID'12 exp - (2254 + 214/RT).
b Evaluated by NASA.
-------
at 200 rpm, the time for 50% removal was 21.5 minutes (average); 90% removal
was accomplished in 71 minutes. The addition of dry granular bentonite clay,
dolomitic limestone, or peat moss had little effect on the evaporation rate.
The rapid volatilization of chloroform was seen also by Jensen and Rosenberg
(1975), who reported that 0.1-1.0 ppm solutions of chloroform in partly open
sunlit aquaria lost 50-60% of the chloroform in 8 days as opposed to only 5%
in closed aquaria. Pearson and McConnell (1975) suggested that the presence
of chloroform in ambient waters may be from aerial transport and washout.
The potential for biodegradation of chloroform in the aquatic
environment was examined by Pearson and McConnell (1975), who observed no
aerobic biodegradation. Tabak et al. (1981) report "significant degradation
with gradual adaptation" of chloroform in a static flask test. Under
anaerobic conditions similar to those which may be present in lake sediments
or groundwater, the reduction of chloroform without bacteria has been
reported by Klecka and Gonsior (1984). In the presence of methanogenic
bacteria, Bouwer and McCarty (1983) and Bouwer et al. (1981) report the
degradation of chloroform with up to a 97% yield of CO?. Similar
observations by Wood et al. (1981) confirm these results.
An EXAMS model of the fate of chloroform in a pond, a river, and an
oligotrophic lake and eutrophic lake revealed the dominant process in all
cases to be volatilization. Input parameters included hydrolysis (2^,5 x
10~9 hr~l), octanol/water partition coefficient (91). vapor pressure
(150.5 torr at 20°C), solubility (8200 ppm), Henry's Law Constant (2.88 x
10~3), reaeration rate ratio (0.583), alkoxy radical rate constant (0.7 NT*
hr-1) (RO- = 1014 M), and a stream loading of 1 g/hr. No photochemical or
bacterial degradation parameters were entered since chloroform has no UV
absorbance >175 nm, and virtually no bacterial degradation occurs with
3-39
-------
chloroform (Pearson and McConnell, 1975; Bouwer et al., 1981). Table 3-12
summarizes the EXAMS model generated fate of chloroform. Note that the EXAMS
model does not include biodegradation rates.
3.8. EXISTING CRITERIA, STANDARDS, AND GUIDELINES
3.8.1. Air
The Occupational Safety and Health Administration (OSHA) currently
limits occupational exposure to chloroform to a ceiling level of 50 ppm
(29 CFR 1982a). This ceiling level is not to be exceeded in the workplace at
any time. To protect against mild central nervous system depression,
irritant effects, and fetal abnormalities (which were considered to occur at
lower exposure levels than those causing liver injury), the National
Institute-for-Occupational Safety and Health (NIOSH) recommended in 1974 that
exposure to chloroform be limited to 10 ppm as a Time-Weighted Average (TWA)
exposure for up to a 10-hour workday, 40 hour workweek. A ceiling level of
50 ppm was proposed for any 10-minute period (NIOSH, 1974). NIOSH lowered
the recommended criterion to 2 ppm TWA in 1976 (NIOSH, 1977) in response to a
positive NCI carcinogenesis bioassay (NCI, 1976). NIOSH recommended that
exposure to halogenated anesthetic agents, including chloroform, be limited
to 2 ppm because this is the lowest detectable level using the recommended
sampling and analysis techniques, and not because a safe level of airborne
exposure could be defined.
On the basis of recent reports of carcinogenicity and embryotoxicity.
the American Conference of Governmental Industrial Hygienists (ACGIH)
currently classifies chloroform as an Industrial Substance Suspect of
Carcinogenic Potential for Man (ACGIH, 1981). The ACGIH recommends a
Threshold Limit Value (TLV) of 10 ppm and a 15-minute Short-Term Exposure
Limit (STEL) TWA of 50 ppm for chloroform.
3-40
-------
TABLE 3-12. SUMMARY OF EXAMSa MODELS OF THE FATE OF CHLOROFORMb
Maximum total concentrations in water
column (mg/L)
Maximum concentration in sediments
(dissolved in pore water, mg/L)
Maximum concentration in Bios
Plankton (pg/g)
Benthos (pg/g)
Maximum total concentration in sediment
(mg/kg, dry weight)
Total steady accumulation (kg)
% in water column
% in sediments
Disposition
chemical transformation (%)
biotransformation (%)
volatilization (%)
Volatilization half-life
exported (%)
export half-life
Mass flux from volatilization (kg/hr)
Self-purification time
River
9.92 x 10-7
9.85 x ID'7
2.58 x 10'5
2.56 x 10~5
3.19 x 10~6
9.15 x ID'4
96.93
3.07
0.00
0.00
1.74
36 hours
98.26
0.65 hours
1.74 x ID'5
37 hours
Pond
2.50 x 10-3
1.36 x 10-3
6.49 x 10-2
3.53 x ID'2
6.50 x 10~3
5.43 x 10~2
91.9
8.1
0.00
0.00
93.35
40 hours
6.65
566 hours
9.33 x 10~4
31 days
Oligotrophic
1.33 x 10-4
5.82 x 10-6
3.45 x 10~3
1.51 x ID'4
2.83 x 10~5
0.33
99.95
0.05
0.00
0.00
94.98
10 days
5.02
192 days
2.28 x 10~3
65 days
Eutrophic
Lake
1.26 x 10-4
4.57 x 10~6
3.27 x 1CT3
1.19 x 10-4
9.35 x 10-6
0.3
99.94
0.06
0.00
0.00
95.57
9 days
4.43
196 days
2.29 x 10-2
56 days
aBurns et al. 1981.
on a stream load of 1.00 g/hr.
-------
Foreign industrial air standards for chloroform include Bulgaria,
10 ppm; Czechoslovakia, 10 ppm (50 ppm for brief exposures); Finland, 50 ppm;
Hungary, 4 ppm (20 ppm for brief exposures); Japan, 50 ppm; Poland, 10 ppm;
Rumania, 10 ppm; Yugoslavia, 50 ppm; West Germany, 10 ppm (Utidjian, 1976).
3.8.2. Water
As discussed in the Ambient Water Quality Criteria Document for
chloroform (U.S. EPA, 1980h), the EPA has proposed an amendment that would
add to the National Interim Primary Drinking Water Regulations a section on
the control of organic halogenated chemical contaminants. The proposed limit
for total trihalomethanes in-drinking water, which includes chloroform as the
major constituent, is 100 ^g/L. Although some estimates of cancer rtsk were
performed, this limit was set primarily on the basis of technological and
economic feasibility, and initially will apply only to water supplies serving
>75,000 consumers. The basis and purpose of this regulation are discussed in
a report that was prepared by the Office of Drinking Water (U.S. EPA, 1978b).
The U.S. EPA (1980h) recently derived cancer-based ambient water
criteria for chloroform. Since zero level concentrations of chloroform will
never be attainable in chlorine-treated water, levels that may result in
incremental increases of cancer risk over the lifetime were estimated at
risks of 1 x 10"5, 10~6, and 10~7. The corresponding recommended criteria,
which were derived with the tumor incidence data from the NCI bioassay with
female mice (NCI, 1976), are 1.90, 0.19, and 0.019 pg/L, respectively, if
exposure is assumed to be from the consumption of drinking water and fish and
shellfish products and at 157 pg/1, 15.7 pg/L, and 1.57 ng/L, respectively,
if exposure is assumed to be from the consumption of aquatic organisms only.
3-42
-------
3.8.3. Food
Chloroform has been approved by the Food and Drug Administration (FDA)
as a component of articles intended for use in contact with food (i.e., an
indirect food additive). The use of chloroform in the food industry is
summarized as follows:
Component of adhesives CFR 1982b
Adjuvant substance required CFR 1982c
in the production of
polycarbonate resins
Chloroform also has been exempted from the requirement of tolerance when
used as a solvent in pesticide formulations that are applied to growing crops
(CFR 1982d), or when used as -a fumigant after harvest for barley, corn, oats,
popcorn, rice, rye, sorghum (milo), or wheat (CFR 1982e).
3.8.4. Drugs and Cosmetics
The positive NCI carcinogenicity bioassay of chloroform (NCI, 1976) has
prompted the FDA to restrict the use of chloroform in drug (CFR 1982f) and
cosmetic (CFR 1982g) products.
3.9. RELATIVE SOURCE CONTRIBUTIONS
The sum of all the environmental releases of chloroform from all sources
listed in Section 3.4.3 amounts to a total of 10,438 metric tons. All
sources are summarized in Table 3-8 with the percent of the total emissions.
Total emissions from all sources constitute about 5.6% of production (184,000
metric tons). Table 3-8 does not include estimated emissions from laboratory
use. Rehm et al. (1982) suggested that these are potentially large but gave
no numerical estimate.
3-43
-------
3.10 REFERENCES FOR CHAPTER 3
ACGIH (American Conference of Governmental Industrial Hygienists). (1980)
Documentation of the threshold limit values. 4th ed. Cincinnati, OH;
pp. 90-91.
ACGIH (American Conference of Governmental Industrial Hygienists). (1981)
Threshold limit values for chemical substances and physical agents in
the workroom environment with intended changes for 1981. Cincinnati,
OH; pp. 13, 42.
Ahlstrom, R.C., Jr.; Steele, J.M. (1979) Methylchloride. In: Grayson, M.;
Eckroth, D., eds. Kirk-Othmer encyclopedia of chemical technology. 3rd
ed. New York: John Wiley and Sons, Inc.; vol. 5., pp. 677-685.
Altshuller, A.P. (1980) Lifetimes of organic molecules in the troposphere
and lower stratosphere. In: Pitts, J.N., Jr.; Metcalf, R.L., eds.
Advances in environmental science and technology. New York: John Wiley
and Sons, Inc.; vol. 10, pp. 181-219.
Anthony, T. (1979) Methylene chloride. In: M. Grayson; D. Eckroth, eds.
Kirk-Othmer encyclopedia of chemical technology, 3rd ed., New York:
John Wiley and Sons, Inc. vol. 5 pp. 686-693.
Atkinson, R.; Darnal1,.K.R.; Lloyd, A.C.; Winer, A.M.; Pitts, J.N., Jr.
(1979) Kinetics and mechanisms of the reactions of the hydroxyl radical
with organic compounds in the gas phase. In: Adv. Photochem.
11:375-488.
Bentley, R.E., et al. (1975) Acute toxicity of chloroform to bluegill
(Lepomis macrochirus). rainbow trout, (Salmo qairdneri), and pink shrimp
(Penaeus durorarum). Contract No. WA-6-99-1414-B. U.S. Environmental
Protection Agency (Cited in U.S. EPA, 1980h.)
Birge, W.J.; Black, J.A.; D.M. Bruser (1979) Toxicity of organic chemicals
to embryo-larval stages of fish. EPA-560/11-79-007. U.S. Environmental
Protection Agency. (Cited in U.S. EPA, 1980h.)
Bouwer, E.J.; McCarty, P.L. (1983) Transformations of 1- and 2-carbon
halogenated aliphatic organic compounds under methanogenic conditions.
Appl. Environ. Microbial 45:1286-1294.
Bouwer, E.J.; Rittman, B.; McCarty, P.L. (1981) Anaerobic degradation of
halogenated 1- and 2-carbon organic compounds. Environ. Sci. Technol.
15: 596-599.
Bozzelli, J.W.; Kebbekus, B.B. (1979) Analysis of selected volatile
organic substances in ambient air. Trenton: State of New Jersey
Department of Environmental Protection.
Brown, R.H.; Purnell, C.J. (1979) Collection and analysis of trace
organic vapour pollutants in ambient atmospheres. The performance of a
Tenax-GC adsorbent tube. J. Chromatog. 178:79-90.
3-44
-------
Burns, L.H.; Cline, D.M., Lassiter, R.R. (1981) Exposure Analysis Modeling
System (EXAMS). Prepared by Environmental Research Lab., Office of
Research and Development, U.S. EPA, Athens, Georgia.
Callahan, M.A.; Slimak, M.W.; Gabel, N.W.; May, I.P.; Fowler, C.F.;
Freed, J.R.; Jennings, P.; Durfee, R.L.; Whitmore, F.C.; Maestri, B.;
Mabey, W.R.; Holt, B.R.; Gould, C. (1979) Water related fate of 129
priority pollutants, vol. II. EPA-440/4-79-029B. Washington, D.C.,
U.S. Environmental Protection Agency.
CFR. (1982a) Air Contaminants. 29:§1910.1000.
CFR. (1982b) Adhesives. 21:§175.105.
CFR. (1982c) Polycarbonate Resins. 21:§177.1580.
CFR. (1982) Exemption from the requirement of a tolerance. 40:§180.1001.
CFR. (1982e) Chloroform; exemption from the requirement of a tolerance.
40:§180.1009.
CFR. (1982f) Chloroform use as an ingredient (active or inactive) in drug
products.
CFR. (1982g) Chloroform as an ingredient in cosmetic products. 21:§700.18.
Chemical Marketing Reporter. (1981) Chemical Profile: Trichloroethylene.
Chemical Marketing Reporter, April 6, 1981.
Chemical Marketing Reporter. (1983) Chemical Profile. Chloroform.
Chemical Marketing Reporter, June 25, 1979, p. 54.
Cherkin, A.; Catchpool, J.F. (1964) Temperature dependence of anesthesia
in goldfish. Science 144:1460. (Cited in U.S. EPA, 1980h.)
Clayberg, H.D. (1917) Effects of ether and chloroform on certain fishes.
Biol. Bull. 32:234. (Cited in U.S, EPA, 1980h.)
Cox, R.A.; Derwent, R.G.; Eggleton, A.E.J.; Lovelock, J.E. (1976)
Photochemical oxidation of halocarbons in the troposphere. Atmos.
Environ. 10:305-308.
Cronn, D.R.; Harsch, D.E. (1979) Determination of atmospheric halocarbon
concentrations by gas chromatography-mass spectrometry. Anal. Lett.
12:1489-1496.
Crutzen, P.J.; Fishman, J. (1977) Average concentrations of OH in the
troposphere and budgets of CH4, CO, H2, and CH3CC13. Geophys. Res.
Lett. 4:321-324.
Davis, D.D.; Machado, G.; Conaway, B.; Oh, Y.; Watson, B. (1976) A
temperature dependent kinetics study of the reaction of OH with CH3C1,
CHC13, and CH3Br. J. Chem. Phys. 65:1268-1274.
3-45
-------
DeShon, H.D. (1979) Chloroform. In: Grayson, M.; Eckroth, D., eds.
Kirk-Othmer encyclopedia of chemical technology. 3rd ed. New York:
John Wiley and Sons, Inc.; vol 5., pp. 693-703.
Dilling, W.L. (1977) Interphase transfer processes. II. Evaporation rates
of chloromethanes, ethanes, ethylenes, propanes, and propylenes from
dilute aqueous solutions. Comparisons with theoretical predictions.
Environ. Sci. Technol. 11:405-409.
Billing, W.L.; Tefertiller, N.B.; Kallos, G.J. (1975) Evaporation rates
and reactivities of methylene chloride, chloroform, 1,1,1-
trichloroethane, trichloroethylene, tetrachloroethylene, and other
chlorinated compounds in dilute aqueous solutions. Environ. Sci.
Technol. 9:833-838.
Environmental Reporter. (1982) State Air Laws. Bureau of National Affairs.
Washington, DC. (Cited in Rehm et al., 1982).
Graedel, I.E. (1978) Chemical compounds in the atmosphere. New York:
Academic Press, pp. 331, 334.
Grimsrud, E.P.; Rasmussen, R.A. (1975) Survey and analysis of halocarbons
in the atmosphere by gas chromatography-mass spectometry. Atmos.
Environ. 9:1014-1017.
Hampson, R.F. (1980) Chemical, kinetic, and photochemical data sheets for
atmospheric reactions. Report FAA-EE-80-17. Washington, D.C.:
Department of Transportation.
Hansch, D.E.; Rasmussen, R.A.; Pierotti, D. (1977) Identification of a
potential source of chloroform in urban air. Chemosphere. 11:769-775.
Hansch, C.; Leo, A.J. (-1979) Substituent constants for correlation
analysis in chemistry and biology. New York: John Wiley and Sons.
pp. 172-173.
Hobbs, F.D. (1978) Trip report for Vulcan Materials Company, Geismar,
Louisiana. Prepared for U.S. Environmental Protection Agency, Emissions
Standards and Engineering Division, Research Triangle Park, NC. (Cited
in Rehm et al., 1982).
Holzer, G.; Shanfield, H.; Zlatkis, A.; Bertsch, W.; Juarez, P.;
Mayfield, M.; Liebich, H.M. (1977) Collection and analysis of trace
organic emissions from natural sources. J. Chromatog. 142:755-764.
Howard, C.J.; Evenson, K.M. (1976) Rate constants for the reactions of
hydrozyl with methane and fluorine, chlorine, and bromine substituted
methanes at 296| Kelvin. J. Chem. Phys. 64:197-202.
Jensen, S.; Rosenberg, R. (1975) Degradibility of some chlorinated
aliphatic hydrocarbons in sea water and sterilized water. Water Res.
9:659-661.
Jones, J.R.E. (1947a) The oxygen consumption of Gasterosteus aculeatus L.
to toxic solutions. Exp. Biol. 23:298. (Cited in U.S. EPA, 1980h.)
3-46
-------
Jones, J.R.E. (1947b) The reactions of Pyqosteus qungltius L. to toxic
solutions. J. Exp. Biol. 24:110. (Cited in U.S. EPA, 1980h.)
Khalil, M.A.K.; Rasmussen, R.A.; Hoyt, S.O. (1983) Atmospheric chloroform
(CHC13): ocean-air exchange and global mass balance. Tellus
358:266-274.
Klecka, G.M.; Gonsior, S.J. (1984) Reductive dechlorination of
chlorinated methanes and ethanes by reduced iron (II) porphyrins.
Chemosphere 13:391-402.
Lillian, D.; Singh, J.B.; Appleby, A.; Lobban, L.; Aruts, R.; Gompert, R.;
Hague, R.; Tommey, J.; Kazazis, J.; Antell, M.; Hansen, D.; Scott, B.
(1975a) Atmospheric fates of halogenated hydrocarbons. Environ. Sci.
Techno!. 9:1042-1048.
Lillian, D.; Singh, H.B.; Appleby, A.; Lobban, L.; Arnts, R.; Gompert, R.;
Hague, R.; Toomey, J.; Kazazis, J.; Antell, M.; Hansen D.; Scott, B.
(1975b) Fates and levels of ambient halocarbons. Am. Chem. Soc. Symp.
Ser. 17:152-158.
Lowenbach and Schlesinger Associates (1979) Chloroform: a preliminary
materials balance. Prepared for the Office of Toxic Substances.
Washington, DC.: U.S. Environmental Protection Agency, (Cited in Rehm
et al., 1982).
Mabey, W.; Mill, T. (1978) Critical review of hydrolysis of organic
compounds in water under environmental conditions. J. Phys. Chem. Ref.
Data 7:383-415.
Merck index: an encyclopedia of chemicals and drugs. (1976) Windholz, M.,
ed. 9th ed. Rahway, NJ: Merck and Co., Inc.; p. 272.
Metcalf and Eddy, Inc. (1972) Waste water engineering, collection treatment
disposal. New York: McGraw-Hill Book Co. (Cited in Wagner et al.,
1980).
Murray, A.J.; Riley, J.P. (1973) Occurrence of some chlorinated aliphatic
hydrocarbons in the environment. Nature 242:37-38.
NAS (National Academy of Sciences) (1978) Chloroform, carbon tetrachloride,
and other halomethanes: An environmental assessment. Washington, D.C.:
National Research Council; 294 pp.
NCASI (National Council for Air and Stream Improvement) (1977) Analysis of
volatile halogenated organic compounds in bleached pulp mill effluent.
NCASI Technical Bulletin No. 298. New York, NY. (Cited in Rehm et al.,
1982).
NCI (National Cancer Institute). (1976) Report on carcinogenesis bioassay
of chloroform. Available from National Technical Information Service,
Springfield, VA. (NTIS PB-264-018).
3-47
-------
NIOSH (National Institute for Occupational Safety and Health) (1974)
Criteria for a recommended standard - occupation exposure to chloroform.
Publ. No. (NIOSH) 75: 114. Washington, D.C.: Dept. Health, Education,
and Welfare.
NIOSH (National Institute for Occupational Safety and Health) (1977)
Criteria for a recommended stnadard - occupational exposure to waste
anesthetic gases and vapors. Publ. No. (NIOSH) 77-140. Washington,
D.C.: Department of Health, Education, and Welfare.
Okuno, T.; Tsuji M.; Shintani, K. (1974) On the chlorination of
hydrocarbons in air. J. Japan. Soc. Air Pollut. 9:211.
Pearson, C.R.; McConnell, G. (1975) Chlorinate q and Co hydrocarbons in
the marine environment. Proc. R. Soc. Lond B. 189:305-332.
Pellizzari, E.D. (1977) The measurement of carcinogenic vapors in ambient
atmosphere. EPA-600/7-77-055. RTP, U.S. Environmental Protection
Agency.
Pellizzari, E.D.; Erickson, M.D.; Zweidinger, R.A. (1979) Analytical
protocols for making a preliminary assessment of halogenated organic
compounds in man and environmental media. EPA-560/13-79-010. U.S.
Environmental Protection Agency. Available from NTIS as PB80-109168.
Rasmussen, R.A.; Harsch, D.E.; Sweany, P.H.; Krasnec, J.P.; Crown, D.R.
(1977) Determination of atmospheric halocarbons by a temperature-
programmed gas chromatographic freezeout concentration method. J. Air
Pollut. Control Distr. 27:579-581.
Rem, R.M.; Anderson, M.E.; Ouletsky, S.A.; Misenheimer, D.C.; Rollius, H.F.
(1982) Chloroform Materials Balance, Draft Report. EPA
Contract 68-02-3168, Task 69.
Sax, N.I. (ed.) (1981) Dangerous properties of industrial materials report.
New York: Van Nostrand Reinhold Co. March/April, p.46. (Cited in Rehm
et al., 1982).
Singh, H.B. (1977) Atmospheric halocarbons: evidence in favor of reduced
average hydroxyl radical concentration in the troposphere. Geophys.
Res. Lett. 4:101-104.
Singh, H.B.; Salas, L.J.; Shigeishi, H.; Smith, A.J.; Scribner, E.;
Cavanagh, L.H. (1979) Atmospheric distributions, sources and sinks of
selected halocarbons, hydrocarbons, SF6, and N20. EPA-600/3-79-107.
U.S. Environmental Protection Agency.
Singh, H.B.; Salas, L.J.; Smith, A.; Shigeishi, H. (1980) Atmospheric
measurements of selected toxic organic chemicals-interim report - 1979.
EPA-600/3-80-072. U.S. Environmental Protection Agency. Available from
NTIS as PB80-198989.
Singh, H.B.; Salas, L.J.; Smith, A.J.; Shigeishi, H. (1981) Measurements of
some potentially hazardous organic chemicals in urban environments.
Atmos. Environ. 15:601-612.
3-48
-------
SRI International. (1983) 1983 Directory of chemical producers: United
States of America. Menlo Park, CA: SRI International; pp. 502, 503,
584.
Stevens, A.A.; Slocum, C.J.; Seeger, D.R.; Robeck, G.G. (1976) Chlorination
of organics drinking water. J. Am. Water Works Assoc. 68:615-620.
Symons, J.M.; Be liar, T.A.; Carswell, J.K.; DeMarco, J.; Kropp, K.L.;
Robeck, G.G.; Seeger, D.R.; Slocum, C.J.; Smith, B.L.; Stevens, A.A.
(1975) National Organics Reconnaissance Survey for Halogenated
Organics. J. Am. Water Works Assoc., Vol. 67 pp. 634-647.
Tabak, H.H.; Quave, S.A.; Mashni, C.I.; Barth, E.F. (1981) Biodegradability
studies with organic priority pollutant compounds. J. Water Pollut.
Control Fed. 53:1503-1518.
TAPPI (Technical.Association of the Pulp and Paper Industry). (1963) The
bleaching of pulp. TAPPI Monograph No. 27. New York, NY. (Cited in
Rehm et al., 1982).
U.S. EPA. (n.d.) National Organic Monitoring Survey. Washington, DC:
Technical Support Division, Office of Water Supply. (Cited in Rehm et
al., 1982).
U.S. EPA. (1976) Atmospheric freons and halogenated compounds.
EPA-600/3-78-108. Research Triangle Park, NC: Environmental Sciences
Research Laboratory. (Cited in Rehm et al., 1982).
U.S. EPA. (1978a) In-depth studies on health and environmental impactsof
selected water pollutants. U.S. Environmental Protection Agency.
Contract No. 68-01-4646. (Cited in U.S. EPA, 1980h.)
U.S. EPA. (1978b) Statement of basis and purpose for an amendment to the
national-interim primary drinking water regulations on trihalomethanes.
Washington, DC: Office of Water Supply. (Cited in U.S. EPA, 1980h.)
U.S. EPA. (1979a) Fate of priority pollutants in publicly owned treatment
works. EPA-400/1-70-301. Washington, DC: Office of Water Regulations
and Standards. (Cited in Rehm et al., 1982).
U.S. EPA. (1979b) Compilation of air pollution emissions factors. 3rd ed.
Supplement 12. AP-42. Transportation and marketing of petroleum
liquids. Research triangle Park, NC: Office of Air Quality Planning
and Standards; pp. 4.4-1 to 4.4-13. (Cited in Rehm et al., 1982).
U.S. EPA. (1980a) Organic chemical manufacturing volume 8: selected
processes. Report 5: chloromethanes by methane chlorination process.
EPA-450/3-80-028c. Research Triangle Park, NC: Office of Air Quality
Planning and Standards. (Cited in Rehm et al., 1982).
U.S. EPA. (1980b) Organic manufacturing volume 8: selected processes.
Report 6: chloromethanes by methanol hydrochlorination and methyl
chloride chlorination process. EPA-450/3-80-028c. Research Triangle
3-49
-------
Park, NC: Office of Air Quality Planning and Standards. (Cited in Rehm
et al., 1982).
U.S. EPA. (1980c) Synthetic organic chemical manufacturing industry -
background information for proposed standards, VOC fugitive emissions.
EPA-450/3-80-33a. Research Triangle Park, NC: Office of Air Quality
Planning and Standards. (Cited in Rehm et al., 1982).
U.S. EPA. (1980d) Organic chemical manufacturing volume 8: selected
processes. Report 1: ethylene dichloride. EPA-450/3-80-028c.
Research Triangle Park, NC: Office of Air Quality Planning and
Standards; pp. III-l to III-9. (Cited in Rehm et al., 1982).
U.S. EPA. (1980e) Development document for effluent limitations guidelines
and standards for the steam electric point source category.
ERA-440/l-80-029b. Washington, DC: Office of Water Regulations and
Standards. (Cited in Rehm et al., 1982).
U.S. EPA. (1980f) Development document for effluent limitations guidelines
and standards for the pulp, paper and paperboard and builders' paper and
board mills. EPA-440/l-80-025b. Washington, DC: Office of Water
Regulations and Standards. (Cited in Rehm et al., 1982).
U.S. EPA. (1980g) Organic chemical manufacturing volume 8: selected
processes. Report 3: fluorocarbon (Abbreviated Report).
EPA-450/3-80-028c. Research Triangle Park, NC: Office of Air Quality
Planning and Standards; p. III-l to III-6. (Cited in Rehm et al.,
1982).
U.S. EPA. (1980h) Ambient water quality criteria for chloroform. Available
as PB81-117442 from Springfield, VA: National Technical Information
Service.
U.S. EPA. (1981a) The determination of halogenated chemical indicators of
industrial contamination in water by the purge and trap method. Method
502.1. EPA-600/4-81-059. U.S. Environmental Protection Agency.
U.S. EPA. - (iŁ81tr) ~~~Air pollution emission factors. 3rd ed. Supplement 12.
AP-42. Storage of organic liquids. Research Triangle Park, NC: Office
of Air Quality Planning and Standards. (Cited in Rehm et al., 1982).
U.S. EPA. (1981c) The 1980 Needs Survey. EPA-430/9-81-008. Washington, DC:
Office of Water. (Cited in Rehm et al., 1982).
U.S. EPA. (1982a) Projections of total lead usage under alternative lead
phasedown programs. Washington, DC. (Cited in Rehm et al., 1982).
U.S. EPA. (1982b) Lead phasedown regulations. 47 FR 49322. November 1.
(Cited in Rehm et al., 1982).
U.S. EPA. (1982c) Chloroform Position Document 2. Washington, DC: Office
of Pesticides and Toxic Substances; pp. 1-3. (Cited in Rehm et al.,
1982).
U.S. EPA. (1985). Health assessment document for 1,2-Dichloroethane
3-50
-------
(ethylene dichloride). EPA 600/8-84 006F. U.S. Environmental
Protection Agency.
USITC. (1982) Synthetic organic chemicals. United States production and
sales, 1981. USITC Publication 1292. Washington, DC: U.S.
International Trade Commission; p. 245.
Utidjian, H.M.D. (1976) J. Occup. Med. 18:253. (Cited in ACGIH, 1980.)
Wagner, K.; Bryson, H.; Hunt, G.; Shochet, A. (1980) Draft report Level I
materials balance chloroform. U.S. Environmental Protection Agency
Contract No. 68-01-5793. Washington, D.C.
Wood, P.R.; Parsons, F.Z.; DeMarco, J.; Harween, H.J.; Lang, R.F.;
Payan, I.L.; Ruiz, M.C. (1981) Introductory study of the
biodegradation of the chlorinated methane, ethane and ethene compounds.
Am. Water Works Assoc. Ann. Conf. and Exposition, June 1981.
St. Louis, MO.
3-51
-------
4. DISPOSITION AND RELEVANT PHARMACOKINETICS
4.1. INTRODUCTION
Considering that chloroform was the major anesthetic agent in use during
the hundred years from its introduction by Simpson in 1847 (Waters, 1951;
Snow, 1858; Simpson, 1847) until after the Second World War, there is
relatively little detailed information about its pharmacokinetics and
metabolism in man. This undoubtedly is due to the fact that until recently,
specific and sensitive analytical methods were unavailable for the
measurement of CHC13 and its metabolites at the concentrations in which they
were likely to be present i_n vivo. Although chloroform as an anesthetic
agent has been replaced by drugs with less cardiac and hepatic toxicity, it
is still widely used in large bulk as an industrial solvent, as a chemical
intermediary, and as a grain fumigant. Chloroform is present in the water
supplies of many United States cities in concentrations reaching 311 i^g/L,
and also has been indentified as a contaminant of the air (U.S. Occupational
Safety and Health Administration (OSHA), 1978; National Institute for
Occupational Safety and Health (NIOSH), 1977b; Dowty et al., 1975; Symons
et al., 1975). Accordingly, ordinary exposure to chloroform occurs from the
workplace, food, drinking water, and ambient air (NIOSH, 1977b; Dowty et al.,
1975; McConnell et al., 1975). Such exposure can be chronic by both oral and
pulmonary routes, but at levels far below anesthetic concentrations (5000 to
10,000 ppm; 24.85 to 49.70 g/nr). Nonetheless, chloroform has been detected
in the breath of healthy people living in non-industrial environments (Conkle
et al., 1975) and in post-mortem human tissue samples (McConnell et al.,
1975).
4-1
-------
4.2. ABSORPTION
Chloroform is rapidly and extensively absorbed through the lungs and
from the gastrointestinal tract. Inhalation is considered the primary route
of entrance into man for occupational exposure and air pollution. Absorption
after oral ingestion is of particular interest since chloroform is a
contaminative component of drinking water and foodstuffs. Significant
absorption of chloroform through intact skin occurs only with liquid contact
and not with vapors.
4.2.1. Dermal Absorption
Absorption of chloroform through the skin from direct liquid contact
(immersion of -hands or arms) is a slow process. Early studies (Torkelson et
al., 1976; Schwenkenbecher, 1904; Witte, 1874) showed that chloroform does
penetrate the skin and can be absorbed into the body by this route. Tsurata
(1975, 1977) has studied the percutaneous absorption of a series of
chlorinated organic solvents applied to a standard area of shaved abdominal
mouse skin for 15 minute periods. Absorption was quantitated by presence of
the compound in the total mouse body plus expired air, as determined by GC.
For all solvents, percutaneous absorption linearly increased with time over
the short exposure period and was directly related to water solubility. For
chloroform the absorption rate was 329 pmoles/min/cm^ skin, third highest of
8 solvents measured. Tsurata extrapolated this absorption rate to a
calculation of the amount absorbed into the human body as the result of 1 min
immersion of both hands (800 cm2 area). The estimated amount absorbed,
19.7 mg/min, was equated to an inhalation exposure concentration of 2429 ppm
for 1 min. Tsurata concluded that skin absorption from liquid contact could
be a significant route of entry into the body for chloroform. More recently
Jakobson et al. (1983) carried out similar experiments with guinea pigs for
4-2
-------
10 chlorinated organic solvents (which however did not include chloroform).
Liquid contact (skin area, 3.1 cm^) was maintained for up to 12 hours and
solvent concentration in blood was monitored during, and for some solvents
after exposure. For these solvents, the blood elimination curves following
dermal exposure were nonlinear, corresponding to a kinetic model involving at
least two body compartments. Furthermore percutaneous absorption of these
solvents, as reflected by blood concentration profiles, showed three
different patterns that were related to water solubility. For solvents which
were relatively hydrophilic [300 to 900 mg/100 ml water] the blood
'concentration increased steadily during the entire dermal exposure,
indicating that absorption occurs fasten than elimination by metabolism or
•pulmonary excretion. Chlorof-orm with a water solubility of about
750 mg/100 ml water might be expected to be in this group.
4.2.2. Oral
The kinetics of gastrointestinal absorption of chloroform after oral
ingestion have not been specifically studied; however, transmucosal diffusive
passage occurs readily, as expected from its neutral and lipophilic proper-
ties (Tables 4-1 and 4-2), and as demonstrated by its biological effects
produced by peroral administration of a wide range of dosages and dosing
schedules for toxicity studies in rats, mice, guinea pigs, and dogs
(Fishbein, 1979; Hill et al., 1975; Brown et al., 1974; Kimura et al., 1971;
Klaassen and Plaa, 1967; Miklashevskii et al., 1966; Plaa et al., 1958;
Eschenbrenner and Miller, 1945), teratologic studies in rats and rabbits
(Thompson et al., 1974), and metabolism studies in mice, rats, rabbits,
monkeys, and man (Brown et al., 1974; Taylor et al., 1974; Fry et al., 1972;
Rubinstein and Kanics, 1964; Paul and Rubinstein, 1963). The rapid
appearance of clinical symptoms following accidental and intentional
4-3
-------
TABLE 4-1. PHYSICAL PROPERTIES OF CHLOROFORM AND OTHER CHLOROMETHANESa
Ostwald solubili
Dichloromethane
Chloroform
Carbon tetrachloride
Vapor pressure
at 25°C, torr
400
250
100
Water/
air
7.6
4.0
0.25
Blood/
air
9.7
10.3
2.4
ty, 37°C
01 ive oil/
air
152
401
361
^Conversion factors:
20°C; 750 mmHg
37°C; 760 mmHg
1 ppm in air = 4.97
1 ppm in air = 4.69
= 4.97 mg/m3
= 4.69 mg/m3
SOURCE: Sato and Nakajima, 1979.
TABLE 4-2. PARTITION COEFFICIENTS FOR HUMAN TISSUE AT 37°C
Tissue
Blood
Brain
Grey matter
White matter
Heart
Kidney
Liver
Lung
Muscle
Fat tissue
Coefficient
8.0
16
24
8
11
17
7
12
280
Relative to blood
2.0
3.0
1.0
1.4
2.1
0.9
1.5
35.0
SOURCE: Steward et al., 1973.
4-4
-------
ingestion of chloroform has also been reported in man (Storms, 1973;
Schroeder, 1965; Piersol et al., 1933).
Brown et al. (1974) and Taylor et al. (1974) found that 14C-chloroform
in olive oil given perorally to mice, rats, and monkeys (60 mg/kg) was
essentially completely absorbed by virtue of a 93 to 98% recovery of
radioactivity in exhaled air, urine, and carcass (Table 4-9). Absorption was
rapid, with peak blood levels at 1 hour in mice and monkeys. In man, Fry
et al. (1972) observed that 13C-chloroform (0.5 g) in olive oil swallowed in
a gelatin capsule resulted in rapid appearance of the stable isotope in
exhaled breath (Table 4-8), with peak blood levels at 1 hour.
Withey et al. (1982) have investigated the effect of dosing vehicle on
the intestinal absorption of-chloroform in fasting rats (400 g) following
intragastric intubation of equivalent doses (75 mg/kg) in about 4 ml of water
or corn oil. The postabsorptive peak blood concentration averaged 6.5 times
higher for water than corn oil (39 vs 6 ^g/ml), while the time to initial
peak blood concentration was essentially the same (5.6 vs. 6.0 min).
Although the absorption from water vehicle exhibited one blood concentration
peak, the absorption from corn oil showed two peaks in blood concentration at
6 and 40 minutes. The ratio of the areas under the blood concentration
curves for 5 hours after dosing (AUC, 5 hr) was 8.7; water, corn oil. These
results suggest that the absorption of chloroform with both vehicles is
rapid. However, the rate and extent of absorption may be diminished, and the
pattern of absorption altered, by intragastric intubation of high volumes
(for a rat) of corn oil vehicle. A slower partitioning of lipophilic
compounds dissolved in corn oil with mucosal lipids can be expected in
comparison with a water vehicle. Furthermore, in contrast to aqueous
absorption into the portal system and thence to the liver, corn oil and other
4-5
-------
liquids are extensively transported via mucosal lymphatic system which slowly
drains by way of the left lymphatic thoracic duct into the systemic
circulation via the superior vena cava. While these considerations are
unlikely to affect the pharmacokinetics of chloroform in man in any practical
way, they are of importance in relation to the modes of dosing employed in
long-term carcinogenicity tests of chloroform and other lipophilic compounds.
4.2.3. Pulmonary Absorption
Chloroform has a relatively high vapor pressure (250 torr at 25°C;
Table 4-1) and a high blood/air partition coefficient (8 to 10.3 at 37°C;
Table 4-2); hence its vapor in ambient air is a primary mode of exposure and
the lungs a principal route of entry into the body. The total amount
absorbed via the lungs (as for. all .vapors) is directly proportional to
(1) the concentration of the inspired air, (2) the duration in time of
exposure, (3) the blood/air Ostwald solubility coefficient, (4) the
solubility in the various body tissues, and (5) physical activity, which
increases pulmonary ventilation rate and cardiac output. Hence, the basic
kinetic parameters of the pulmonary absorption of chloroform and its
equilibration in the body are as valid for low concentrations expected in the
ambient environment as for the high vapor concentrations associated with its
use as an anesthetic (5000 to 10,000 ppm; 24.85 to 49.70 g/m3) (Smith et al.,
1973; Morris, 1951; Waters, 1951). These parameters have not been as well
studied as they have for modern anesthetics like halothane (Fiserova-
Bergerova and Holaday, 1979) or even for other common halogenated hydrocarbon
solvents like trichloroethylene, methylene chloride, or methylchloroform.
The earliest attempt at controlled studies of pulmonary absorption of
chloroform in man were conducted by Lehmann and Hasegawa (1910). These
investigators calculated retention values for chloroform (% inspired air
4-6
-------
concentration of chloroform retained in the body) from differences between
inspired and expired air concentrations (analyzed by alkali hydrolysis with
chloride titration). As expected, initial retention values were high, and
decreased with exposure duration as total body equilibrium with inspired air
concentration was approached (Table 4-3). Both the rate of uptake to
equilibration and the final retention value achieved are related to the
solubility of chloroform in blood (blood/air partition coefficient).
Figure 4-1 illustrates for chloroform and other anesthetic vapors that the
greater the Ostwald solubility coefficient for a vapor agent, the less
rapidly equilibrium occurs. From the data of Lehmann and Hasegawa
(Table 4-3) and recent data of Smith et al. (1973) concerning blood levels of
concern during anesthesia as shown in Figure 4-2, total body equilibrium with
inspired chloroform concentration requires at least >2 hours in normal man at
resting ventilation rate and cardiac output. The retention value at
equilibrium suggested by the Lehmann and Hasegawa (1910) data is -65%, and
is 67% as calculated from the data of Smith et al. (1973). The difference,
33 to 36%, represents body elimination of chloroform by routes other than
pulmonary (primarily by metabolism). The percent retention value is
independent of the inspired air concentration at equilibrium.
The magnitude of chloroform pulmonary uptake into the body (dose, body
burden) is directly related to the concentration of chloroform in the
inspired air and to the duration of exposure. The total amount retained in
the body during inhalation exposure can be estimated by multiplying percent
retention (R) by the volume of air inspired during the exposure period, or:
Amount uptake = (Cj - C^) • V • T
where V is ventilation rate (L/minute), T is exposure period (minute), and Cj
and CA are inspired air concentration and end alveolar air concentration,
4-7
-------
TABLE 4-3. RETENTION AND EXCRETION OF CHLOROFORM BY MAN DURING AND AFTER
INHALATION EXPOSURE TO ANESTHETIC CONCENTRATIONS
Subject
(Inspired air cone., ppm)
1
(4448)
2
(4920)
3
(4407)
Exposure period, min
0 to 5
5 to 10
10 to 15
15 to 20
20 to 25
25 to 30
Retention, %
74.5
72.4
6S.6
67.6
NR
NR
68.4
61.6
51.2
50.2
NR
NR
80.0
74.2
76.9
74.6
74.2
73.8
Excretion, mg/L expired air
Postexposure, min
0 to 10
10 to 20
20 to 30
NR
NR
NR
NR
NR
NR
1.70
0.97
0.85
NR = not reported; min = minutes.
SOURCE: Lehmann and Hasegawa, 1910.
4-8
-------
100
Figure U-1. Rate of rise of alveolar (arterial) concen-
tration toward inspired concentration for five anesthetic
agents of differing Ostwald solubilities (blood/air par-
tition coefficients): nitrous oxide, 0.^7; forane, 1.4;
halothane, 2.M; chloroform, 8; and methoxyflurane, 11.
Note rate of alveolar chloroform rise is less than that
of halothane with a smaller Ostwald coefficient and
greater than that of methoxyflurane with a larger coef-
ficient.
Source: Munson (1973)
4-9
-------
c
0)
o
ai
O.
o>
O
z
iu
(J
O
(J
5
ŁE
O
u.
O
cr
O
_i
x
u
12
10
CHLOROFORM
O VENOUS BLOOD
D ARTERIAL BLOOD
BASE
EXCESS
P»C02
PH
_J
+1.5
40.0
7.44
I
•2.0
36.0
7.39
-2.0
33.0
7.40
-3.0
33.0
7.39
-4.0
30.0
7.39
I
-5.0 -1.5
27.0 36.0
7.40 7.41
I
POST INDUCTION TIME, hours
Figure U-2. Arteriovenous blood concentrations of a
patient during anesthesia with chloroform. Note anes-
thetic blood concentration for chloroform, the decreasing
difference between arterial and venous concentrations at
2 to 3 hours, indicating whole-body equilibrium, and the
rapid fall of blood concentration with termination of
chloroform exposure.
Source: Smith et al. (1973).
4-10
-------
respectively. Physical activity increases uptake by increasing the
ventilation rate, V, and the cardiac output which influences rate of
distribution to the various tissues of the body.
During inhalation of chloroform (and in the post exposure elimination
phase), the arterial blood concentration of chloroform is directly
proportional to inspired air concentration (and end alveolar air
concentration). This fixed relationship is defined by the blood/air
partition coefficient in comparison to other solvents (Sato and Nakajima,
1979) (Table 4-1), and hence, for equivalent ambient air exposure
concentrations, the blood concentration of chloroform is proportionally
higher. For inspired air-concentration required for surgical anesthesia
(8,000 to 10,000 ppm; 39/76 to 49.70 g/m3), Smith et al. (1973) observed a
mean arterial blood chloroform concentration for 10 patients of 9.8 mg/dl
with a range of 7 to 16.5 mg/dl (Figure 4-2), and Morris (1951) found similar
values for his patients. For inspired air concentrations less than
anesthetic levels, for example low vapor concentrations of 10 to 100 ppm
(49.7 to 497 mg/m3), blood chloroform concentrations are lower in direct
proportion.
The amount of pulmonary absorption of chloroform is also influenced by
total body weight and by the total fat content of the body (average body fat
content is 8% of body weight) (Geigy Scientific Tables, 1973). The capacity
of adipose tissue to absorb chloroform in vivo is determined by the product
of adipose tissue weight and lipid solubility of chloroform. The lipid
solubility of chloroform is relatively high for this haloalkane (olive
oil/air, 401; Tables 4-1 and 4-2), and also, the adipose tissue/blood
partition coefficient is high (280 at 37°C); therefore, the uptake and
4-11
-------
storage of chloroform in adipose tissue can be substantial, and this uptake
and storage is increased with excess body weight and obesity.
4.3. TISSUE DISTRIBUTION
Chloroform, after pulmonary or peroral absorption, is distributed into
all body tissues. The compound crosses the placental barrier, as indicated
by embryotoxicity and teratogenicity in mice, rats, and rabbits after oral
and inhalation dosing (Murray et al., 1979; Oil ley et al., 1977; Schwetz
et al., 1974; Thompson et al., 1974). It has been found in fetal liver (von
Oettingen, 1964). Chloroform can be expected to also appear in human
colostrum and mature breast milk, since it has .been found in fresh cow's milk
and in high content in cheese and butter (Table 4-4).
As to be expected from the -lipophilie nature of chloroform and modest
water solubility (Table 4-1), highest concentrations are detected in tissues
with higher lipid content; relative tissue concentrations are reflected by
individual tissue/blood partition coefficients. Coefficients for human
tissues, given in Table 4-2, indicate that relative tissue concentrations are
expected in the order of adipose tissue > brain > liver > kidney > blood.
The absolute amounts of chloroform detected in these tissues at any given
time are proportional to the body dose (i.e., to the concentration in the
inspired air and duration of inhalation or to the oral dose, partition
coefficient, and to the tissue compartment size).
Gettler (1934) and Gettler and Blume (1931), using a modified Fujiwara
analytical method, determined the chloroform content of the brain, lungs, and
liver of nine patients who died during surgical anesthesia (presumably 5000
to 10,000 ppm; 24.85 to 49.50 g/m3 inspired air) as: brain, 120 to 182; lung,
92 to 145; and liver, 65 to 88 mg/kg tissue wet weight. Even higher values
(372 to 480 mg/kg in brain tissue) were detected in seven cases of death due
4-12
-------
TABLE 4-4. CHLOROFORM CONTENT IN UNITED KINGDOM FOODSTUFFS
AND IN HUMAN AUTOPSY TISSUE
Chloroform in U.K. foodstuffs
Ch
Foodstuff
Dairy produce
Fresh milk
Cheshire cheese
English butter
Hens' eggs
Meat
English beef (steak)
English beef (fat)
Pig's liver
Oils and fats
Margarine
Olive oil (Spanish)
Cod liver oil
Vegetable cooking oil
Beverages
Canned fruit drink
Light ale
Canned orange juice
Instant coffee
Tea (packet)
Fruit and vegetables
Potatoes (S. Wales)
Potatoes
(N.W. England)
Apples
Pears
Tomatoes
Fresh bread
loroform,
ng/kg
5
33
22
1.4
4
3
1
3
10
6
2
2
0.4
9
2
18
18
4
5
2
2
2
Chloroform in human autopsy
tissue
Chloroform,
Age of pg/kg
subject Sex Tissue (wet tissue)
76 F Body fat
Kidney
Liver
Brain
76 F Body fat
Kidney
Liver
Brain
82 F Body fat
Liver
48 M Body fat
Liver
65 M Body fat
Liver
75 M Body fat
Liver
66 M Body fat
74 F Body fat
19
2
5
4
5
5
1
2
67
8.7
67
9.5
64
8.8
65
10.0
68
52
SOURCE: McConnell et al., 1975.
4-13
-------
to excessive administration of chloroform (Gettler, 1934). The blood
concentration during surgical anesthesia has been recently determined (by GC)
to range from 70 to 165 mg/L in 10 patients (average, 98) by Smith et al.
(1973). These tissue concentrations are in general agreement with the
tissue/blood partition coefficients summarized from the literature by Steward
et al. (1973) and given in Table 4-2.
In contrast to the high tissue levels of chloroform detected in response
to inspired air concentrations required for anesthesia, McConnell et al.
(1975) recently analyzed post-mortem tissue from eight persons, four males
and four females, with an age range of 48 to 82, living in.the United Kingdom
in ordinary non-industrial circumstances, for chloroform and other
halogenated compounds (carbon tetrachloride, trichloroethylene,
perchloroethylene, hexachlorobutadiene). Significant tissue levels of three
chlorinated hydrocarbons were detected. Chloroform levels, pg/kg wet tissue
weight, were as follows: body fat, 5 to 68 (average of 51); liver, 1 to 10
(average of 7.2); kidney, 2 to 5; and brain, 2 to 4 (Table 4-4). Presumably.
these tissue levels of chloroform were derived from air, foodstuff
(Table 4-4), and drinking water contamination (OSHA, 1978; Dowty et al.,
1975; Symons et al., 1975).
There have been few controlled exposure studies in animals investigating
the distribution of chloroform in body tissues and determining dose-dependent
tissue concentrations. Chenoweth et al. (1962) determined blood and tissue
concentrations of chloroform in two normal fasted dogs after 2.5 hours of
surgical anesthesia. Concentration of chloroform in the inhaled stream
during anesthesia was not determined, but anesthesia was judged to be
satisfactory at an arterial level of 45 to 50 mg/dl. Blood and tissue
chloroform levels were determined by infrared spectroscopy after tissue
4-14
-------
extraction in cold carbon disulfide and distillation. Table 4-5 shows the
relative concentration of chloroform in body tissues. The highest
concentrations were detected in fat tissue, some 10-fold greater than blood,
and in adrenals (4-fold greater than blood); the concentrations in brain,
liver, and kidney were similar to blood.
Cohen and Hood (1969) used low-temperature whole-body autoradiography to
study the distribution of 14C-chloroform in mice. Individual mice were
administered 2.4 pi of 14C-labeled chloroform by inhalation over a 10-minute
period. The animals were sacrificed 0, 15, or 120 minutes after exposure.
Autoradiography of mice killed immediately after inhalation showed the
highest concentration of radioactivity in body fat and liver, while lesser
and relatively uniform amounts were seen in blood, brain, lung, kidney, and
muscle. By 120 minutes after exposure, a considerable decrease in total
radioactivity occurred, which was principally confined to liver, duodenum,
and fat. A mottled appearance in the liver suggested a segmental or
localized distribution. Biopsy specimens were taken from selected tissues in
each animal and radioactivity determined by scintillation counting.
Table 4-6 shows the distribution of radioactivity (chloroform and
metabolites) in these tissues, and tissue/blood concentration ratios.
Following sacrifice, after 10 minutes of exposure, most tissues approach a
unit concentration with blood. However, in both fat and liver, the
concentration exceeds unity. By 15 minutes, the ratio of radioactivity in
brown fat reaches its peak at 15 times that detected in blood. The relative
concentration of radioactivity in the liver continues to increase until the
termination of the experiment at 120 minutes, when it reaches a final value
6.7 times in excess of that in the blood. Kidney and lung tissues also
increased in relative concentration over the 2-hour period to a value of 1.53
4-15
-------
TABLE 4-5. CONCENTRATION OF CHLOROFORM IN VARIOUS TISSUES OF TWO DOGS
AFTER 2.5 HOURS OF ANESTHESIA
Arterial blood
Brain
Adrenal (total)
Fat, omentum
Right ventricle
Skeletal muscle
Lung
Liver
Spleen
Kidney
Bile
Thyroid
Pancreas
Urine
Dog A
^g/g wet
275
298
1185
2820
214
189
147
282
237
225
209
460
296
57
Dog B
tissue weight ± 5%a
397
392
1305
1450
314
155
336
290
255
226
205
760
350
73
^Chloroform concentration was determined by infrared spectrometry after
tissue extraction.
SOURCE: Chenoweth et al., 1962.
4-16
-------
TABLE 4-6. CONCENTRATIONS OF RADIOACTIVITY (CHLOROFORM PLUS METABOLITES)
IN VARIOUS TISSUES OF THE MOUSE (NMRI)a
I
I—>
~-J
Total radioactivity (counts/min/mg)
Tissue
Blood
Brain
Muscle
Lung
Kidney
Liver
Fat
Brown fat
0 min
260 ±
217 ±
288 ±
262 ±
284 ±
407 ±
1674 ±
3158 ±
22.0
16.4
44.4
24.1
35.0
36.9
201
384
15 min
103
112
110
149
145
208
953
1490
± 17
± 9
± 5
± 12
± 21
± 9
± 92
± 98
.5
.9
.1
.6
.6
.3
.7
.4
120
37 ±
23 ±
26 ±
53 ±
56 ±
250 ±
266 ±
211 ±
min
4.0
2.7
6.9
8.2
8.0
17.9
30.1
40.3
Tissue/blood ratio
0 min
1.
0.
0.
1.
1.
1.
6.
12.
00
84
87
01
08
56
42
12
15 min
1.00
1.12
1.07
1.44
1.41
2.10
9.25
14.70
120 min
1.00
0.63
0.70
1.43
1.53
6.76
7.18
5.70
^Animal sacrifices were at 0, 15, or 120 minutes following 10-minutes inhalation of chloroform.
represent duplicate determinations in each of two animals at each time sequence (± S.E).
Min = minutes.
SOURCE: Cohen and Hood, 1969.
Data
-------
and 1.43 of blood, respectively. The increasing ratios of liver and
kidney/blood radioactivity represent a continued accumulation of metabolites
within these organs. High body fat/blood concentrations shows that adipose
tissue represents an important storage site, prolonging retention of
chloroform in the body.
Whole-body autoradiography was also carried out by Brown et al. (1974)
on male and female Sprague-Dawley rats and squirrel monkeys given
14C-chloroform perorally (60 mg/kg). Male and female rats killed 3 hours
after dosage showed no apparent sex difference in distribution of
radioactivity. Radioactivity was greatest in body fat and liver, while
lesser amounts were seen -in blood, brain, lung, kidney, and muscle. Squirrel
monkeys showed a similar distribution, with the exception that high
concentrations of radioactivity were present in the bile and increased with
time. Examination of bile extract by gas-liquid chromatography showed the
bile radioactivity was unchanged chloroform, indicating an excretion of
chloroform by the biliary route in the monkey.
Brown and his colleagues (Taylor et al., 1974) also investigated the
tissue distribution of chloroform in 3 strains of mice (CF/LP, CBA, and C57)
by-whole-body autoradiography after oral dosing (60 mg/kg ^C-chloroform).
In the male mice of the three strains examined 3 hours after dosing, the
greatest amounts of radioactivity appeared in liver and kidneys, and lesser
amounts in renal cortex but not medulla. Female mice showed greatest
radioactivity in liver, intestine, and bladder, with much less radioactivity
in kidney and little differentiation between renal cortex and medulla. The
same general patterns were observed 5, 7, and 24 hours after dosing. Biopsy
samples of these tissues were taken, and radioactivity determined by
scintillation counting. Table 4-7 shows the distribution of l4C-chloroform
4-18
-------
TABLE 4-7. TISSUE DISTRIBUTION OF l^C-CHLOROFORM RADIOACTIVITY IN CF/LP MICE
AFTER ORAL ADMINISTRATION (60 mg/kg)a
Mean DPM/100 mg wet weight (SEM)
Tissue Male (6) Female (6)
Liver
Kidney
Brown fat
Blood
18,157
13,759
1,011
2,910
(1898)
(1047)
(80)
(423)
21,535 (2097)
3,920 (533)
1,074 (54)
2,906 (457)
aSimilar results were obtained for CBA and C57 strains.
SOURCE: Taylor et al., 1974.
radioactivity in male and female mice killed 5 hours after dosing. There is
a 3.5-fold difference between the activity present in the male and female
kidneys of each strain. Male mice had greater activity in the kidneys, but
female mice showed relatively greater activity in the liver. This sex
difference in distribution was abolished by castration or testosterone
administration to female mice. The sex difference in tissue distribution of
chloroform and its metabolites may relate to the nephrotoxic effect of
chloroform-that occurs in male mice but not in female mice (Bennet and
Whigham, 1964; Culliford and Hewitt, 1957; Hewitt, 1956; Shubik and Ritchie,
1953; Eschenbrenner and Miller, 1945a,b). The pattern of tissue distribution
of chloroform in mice also depends on mode of exposure. Cohen and Hood
(1969), after chloroform inhalation, found highest levels in body fat
(Table 4-6), while Brown et al. (1974) and Taylor et al. (1974) observed
lower levels in fat and highest levels in liver and kidney following oral
dosing (Table 4-7). The high liver levels of chloroform after oral
administration may be due, in part, to first passage and extraction by the
4-19
-------
liver after this route of administration, to differences of time after
exposure (2 versus 5 hours), and to metabolism and covalent binding of
metabolites to cellular macromolecules (see below).
The sex difference in tissue distribution and binding of chloroform (and
metabolites) in kidney and liver, noted by Brown and his colleagues (Taylor
et al., 1974), appeared to be peculiar to mice. These workers did not
observe such differences in male and female rats or squirrel monkeys (Brown
et al., 1974).
4.4. EXCRETION
Elimination of chloroform from the body is perforce the sum of
metabolism and excretion of unchanged chloroform via pulmonary and other
routes. Unmetabolized chloroform is excreted almost exclusively through the
lungs; however, metabolism of chloroform is extensive, with the proportion
excreted unchanged dependent on body dose. Surprisingly, considering its
historical importance, its longtime use as an industrial chemical and
anesthetic agent, few controlled experimental studies in man have been made
on the kinetics of excretion of chloroform.
4.4.1. Pulmonary Excretion
Figure 4-3 shows the time-course of pulmonary elimination of chloroform
after accidental inhalation exposure to a mixture of solvents, including
chloroform, carbon tetrachloride, trichloroethylene, and perchloroethylene.
Stewart et al. (1965) determined, post-exposure, the alveolar air
concentrations of these solvents by infrared and gas-liquid chromatography
analysis. The kinetics of pulmonary excretion of the solvents are
independent of one another. However, all, including chloroform, demonstrate
the typical kinetics of gaseous vapor pulmonary elimination, that has been
4-20
-------
i
o
-------
observed experimentally for relatively hydrophobia, volatile gaseous
anesthetics and industrial solvents (Eger, 1963; Fiserova-Bergerova et al . ,
1974, 1979, 1980; Droz et al., 1977). At termination of exposure with zero
concentration in inspired air, chloroform (and the other solvents)
immediately begins to be eliminated from the body into the lungs, with blood
and alveolar air concentrations describing parallel exponential decay curves
with three major components (Figure 4-3). These exponential components have
been related by many investigators (Eger, 1963; Fiserova-Bergerova et al.,
1978, 1979, 1980; Droz et al., 1977) to first-order kinetics of pulmonary
elimination associated with desaturation of physiological compartments in
accordance with a blood flow-limited-model in which the rate constants are
determined predominately by tissue perfusion^ volume of tissue distribution
and by partition coefficients:
Tissue uptake and Fj
desaturation of = F • \ . exp ( - - • t )
compartment, T T bl/air V
where F is blood flow through tissue compartment, V is volume of the
compartment, g is partition coefficient, and exp is base of natural
logarithm.
0.693 V \T/bl
t1/2 = - p -
Since three exponential components are typically observed
experimentally, three physiological tissue compartments are included in the
model described by a three-term exponential function of the form:
Uptake, Desaturation = Ae-at + Be-pt + Ce-vt
4-22
-------
where A, B, C are macrocoefficients and a, p, \ are hybrid constants (defined
above). These terms represent three flow-limited major body compartments:
(1) a vessel-rich group of tissues (VRG) with high blood flow and high
diffusion rate constant (VRG: brain, heart, kidneys, liver and endocrine and
digestive systems), (2) lean body mass (MG: muscle and skin), and
(3) adipose tissue (FG). More recently, Fiserova-Bergerova and coworkers
(1974, 1979, 1980) have mathematically reformulated this physiological,
first-order model to accommodate the effect of metabolism on uptake,
distribution and clearance of inhaled vapor compounds.
The half-times (t 1/2) of elimination from-the physiological
compartments (VRG < MG < FG) are-independent of the body dose, but are
dependent on tissue/blood partition coefficients and blood/air partition
coefficients. Since these solvent compounds have high solubility in body fat
(Table 4-1), they are eliminated slowly from fat depots with a long half-time
of elimination, as illustrated by Stewart's patient (Stewart et al., 1965) in
Figure 4-3. From Figure 4-3, it can be estimated graphically that chloroform
has a half-time of elimination from the fat compartment (FG) of =36 hours,
with similar long half-times for the highly fat soluble compounds,
perchloroethylene and carbon tetrachlorid-e.
There is limited information available for the half-times of pulmonary
elimination of chloroform from the VRG and MG. From the early data of
Lehmann and Hasegawa (1910) given in Table 4-3, the half-time of pulmonary
elimination from the VRG appears to be =30 minutes. A similar estimate can
be made from the data of Smith et al. (1973) and Morris (1951) at termination
of anesthesia in man; these workers determined that blood chloroform
concentration rapidly fell exponentially from 7 to 3.5 mg/dl within
30 minutes (Figure 4-2).
4-23
-------
Pulmonary elimination of chloroform was investigated by Fry et al.
(1972) in male and female volunteers given ^C-chloroform jn olive oil orally
(by gelatin capsule). Chloroform was determined in expired air by GLC.
Their data, summarized in Table 4-8, show that the amount of chloroform
excreted through the lungs within 8 hours (expressed as a percentage of the
dose, 0.1 to 1.0 g), increased (0 to 65%) in proportion to the dose.
Following a peak blood concentration (0.5 mg/dl for a 500 mg dose) 1 hour
after oral dosage, absorption, and distribution, the blood chloroform
concentration declined exponentially with three components: (1) a very rapid
disappearance, with half-time of 14 minutes possibly corresponding to VRG
compartment kinetics, (2) a slower disappearance, with half-time of
90 minutes corresponding to -MG kinetics, and (3) a very slow disappearance,
with very long half-time from adipose tissue. This half-time was
undetermined, but chloroform was detected in blood and breath 24 hours later.
Fry and coworkers (1972) noted a linear relationship for their subjects
between pulmonary excretion of chloroform and body weight deviation from
ideal, an index of excessive leanness or excessive body fat from normal.
Their data in Figure 4-4 show that for both male and female subjects given a
standard oral dose of chloroform, lean subjects eliminate via the lungs a
greater percentage of the dose, while overweight subjects eliminate less
chloroform. The different slopes of the linear relationship for men and
women presumably reflect the different proportion of adipose tissue in the
two sexes. The bodies of women tend to contain higher proportions of fat
than those of men (Geigy Scientific Tables, 1973). These observations
reinforce the role of adipose tissue as a storage site for chloroform.
Brown and his coworkers (Brown et al., 1974; Taylor et al., 1974) have
demonstrated an animal species difference in the amount of pulmonary
4-24
-------
TABLE 4-8. PULMONARY EXCRETION OF 13CHC13 FOLLOWING ORAL DOSE
(PERCENT OF DOSE)a
Subjects
8 M and F
1
1
1
Dose, g
0.5
1.0
0.25
0.10
Mean
for 8 hoursb
40.3
64.7
12.4
nil
Range
17.8 to 66.6
NA
NA
NA
Pulmonary excretion of 13C02 following 0.5 g oral dose of 13CHC13,
Cumulative percent of doseb
Time after dose, hr
Subjects
Male (1)
Female (62.7 kg) (1)
0.5
2.1
0.5
1.75
24.1
10.7
2.5
35.9
28.3
5.5
49.2
47.5
7.5
50.6
48.5
aRecalculated from the data of Fry et al., 1972.
bWithin 4% of value calculated for infinite time.
NA = not applicable.
excretion of ^C-chloroform from a standard oral dose (60 mg/kg body weight)
given in olive oil. Mice (three strains), rats, and squirrel monkeys excrete
chloroform via the lungs (6, 20, and 79%, respectively, of the standard
dose). This species difference is primarily related to the capacity to
metabolize chloroform rather than differences in pulmonary kinetics, since,
4-25
-------
80
§5
I I
>Z40
o:O
^f1
Z UJ
OCC
I>
a.
-6-4-2 0+2+4+6+8
BODY-WEIGHT DEVIATION FROM CALCULATED NORMAL, kg
Figure U-4. Relationship between total 8-hour pulmonary
excretion of chloroform following 0.5-g oral dose in man and
the deviation of body weight from ideal. The different slopes
of the linear relationship for men and women reflect the
different proportion of adipose tissue in the two sexes.
4-26
-------
as shown in Table 4-9, the percentage of the dose metabolized to -CC^ is
inversely proportional to that of pulmonary excretion. The mice, 48 hours
after dosing, retained only 2% of chloroform radioactivity (Table 4-9).
Withey and Collins (1980) determined the kinetics of distribution and
elimination of chloroform from blood of Wistar rats after intravenous
administration of 3, 6, 9, 12 or 15 m/kg of chloroform given in 1 ml water
intrajugularly. For all doses, the blood decay curves exhibited three
components of exponential disappearance of chloroform (a, p, A components) and
"best" fitted a first-order three compartment model. Table 4-10 summarized
the values obtained for the kinetic parameters. For volatile, lipophilic
compounds, for which a major route of elimination is pulmonary, experiments
utilizing dose administration via relatively large intravenous bolus
injections (relative to rat total blood volume), and which measures only
blood chloroform disappearance, provide a number of problems for data
interpretation of elimination (pulmonary and metabolism) and/or tissue
distribution. In these experiments, pulmonary elimination, which is rapid
for organic solvents, occurs simultaneously with distribution and metabolism;
in contrast, experiments in which the animal is preloaded by oral or
inhalation administration, distribution is more readily separable from
pulmonary elimination. After intravenous administration (Table 4-10), the
rate constant ke (for elimination of chloroform from the central compartment
blood out of the body (principally via pulmonary excretion and/or metabolism)
was dose-dependent and consistent with a half-time of elimination of only
3.6 min for the lowest dose and only 6.2 min for the highest dose. Since the
half-times for distribution into other tissue compartments from the central
compartment blood have longer half-times, it is likely that, of the dose
introduced into the blood, a major portion (depending on dose) was excreted
4-27
-------
TABLE 4-9. SPECIES DIFFERENCE IN THE METABOLISM OF 14C-CHLOROFORM
(ORAL DOSE OF 60 mg/kg)a
14c-radioactivity 48 hours after dose, mean values as
percent dose
Species
No.
Expired
or metabolites
Expired Urine +
14COo feces Carcass
Total
Mice
CF/LP,
CBA,C57
strains
Rats
S-D
Squirrel
monkeys
19 6.1
6 19.7
6 78.7
85.1 2.6 1.8 95.6
65.9 7.6 NR 93.2
17.6 2.0 NR 98.3
aRecalculated from the data of Brown et al., 1974.
NR = not recorded.
within a few minutes by the lungs. Further indication that the dose and/or
mode of administration influenced the distribution and elimination of
chloroform was shown by the proportional increase of the apparent volume of
distribution, Vd (45 ml; 3 mg/kg to 89 ml; 15 mg/kg) and the decrease with
dose in values for rate constants of transfer from blood to other tissue
compartments. The volume of distribution (Vd) of chloroform was 89 ml for
the highest dose or about 22% b.w., surprisingly low for a lipid soluble
compound that is known to diffuse into all the major organ systems
(Tables 4-5 and 4-6). Adipose tissue is known to be a major tissue
compartment for chloroform [Section 4.3]; the clearance of chloroform from
perirenal fat was found to be slow with a half-time of 106 min, and since the
4-28
-------
TABLE 4-10. KINETIC PARAMETERS FOR CHLOROFORM AFTER I.V. ADMINISTRATION TO RATS (min -1 ± S.E.)
r>o
Dose,
mg/kg
3.0
6.0
9.0
12.0
15.0
15.0;
a Vd,b
ml
45.07
±0.04
53.57
±12.61
64.46
±14.26
80.62
±20.20
89.13
±12.83
ti min
a
0.72
±0.11
0.64
±0.13
0.64
±0.084
0.32
±0.20
0.35
±0.048
2.1
P
0.135
±0.001
0.081
±0.01
0.095
±0.0001
0.060
±0.028
0.070
±0.006
9.9
Y
0.0287
±0.0064
0.0158
±0.0019
0.0189
±0.0009
0.0074
±0.0056
0.0134
±0.0005
51.7
ke
0.1907
±0.0239
0.1874
±0.0356
0.1284
±0.0217
0.1035
±0.0232
0.1124
±0.0110
6.2
k!2
0.2346
±0.0651
0.2681
±0.0631
0.2529
±0.0489
0.1071
±0.0774
0.1104
±0.0256
6.3
k21
0.2575
±0.0316
0.1862
±0.0188
0.2421
±0.0064
0.1192
±0.0676
0.1523
±0.0188
4.6
k!3
0.0921
±0.0008
0.0730
±0.0246
0.0594
±0.0034
0.0395
±0.0253
0.0396
±0.0104
17.5
k31
0.0421
±0.0098
0.0233
±0.0032
0.0281
±0.0042
0.0106
±0.0084
0.0193
±0.0018
35.9
a2 to 4 rats/dose.
"Vd = volume of distribution.
SOURCE: Withey et al., 1982.
-------
rate constant, k31 was given as 0.0193 min (for 15 mg/kg dose) indicating a
half-time of 36 min, the adipose tissue appears to be a deep compartment.
These investigators believe that their kinetic data show no evidence in the
rat of nonlinear or dose-dependent Michaelis-Menten kinetics and they suggest
that a dose of 15 mg/kg is below hepatic metabolism saturation.
However, Reynolds et al. (1984a,b) have observed dose-dependent
metabolism and first-pass effect in rats given 12 or 36 mg/kg 14c-chloroform
by gavage in mineral oil. For these two oral doses, 5% and 12%,
respectively, were excreted unchanged in exhaled air with peak rates of 0.71
and 1.81 mg hr~l kg~l at 15 and 30 min, respectively, after dosing. The
percentages of the two doses metabolized, (to 14fj02 collected in exhaled air)
was 67% and 68%, respectively, at peak" rate of 4.4 and 5.5 mg hr~l kg~l,
30-45 and 60-105 min after ministration. These investigators fitted the
pulmonary elimination data to a linear two-compartment model and determined
the apparent half-times of absorption, distribution, and pulmonary
elimination for the 12 and 36 mg/kg gavage doses, respectively; for
chloroform absorption, 0.08 and 0.13 hr, distribution 0.29 and 0.41 hr,
pulmonary elimination 3.83 and 2.27 hr; and for pulmonary elimination of CO?
metabolite, 2.1 and 5.6 hr. These dose-dependent half-times indicate
strongly a dose-dependent disposition of chloroform in the rat in accordance
with Michaelis-Menten kinetics for metabolism.
4.4.2. Other Routes of Excretion
Chloroform is not eliminated in significant amounts from the body by any
route other than pulmonary. Studies of chlorinated compounds in the urine
after chloroform inhalation exposure or peroral dosage to animals and humans
have failed to detect unchanged chloroform (Brown et al., 1974; Fry et al.,
1972). Brown et al. (1974) identified chloroform in high concentration in
4-30
-------
the bile of squirrel monkeys after oral ^C-chloroform dosage, and suggested
an active enterohepatic circulation in this species. For monkeys, they found
only 2% of dose radioactivity in combined urine and feces collected for
48 hours after dosing (Table 4-9), and only 8 and 3% for rats and mice,
respectively.
4.4.3. Adipose Tissue Storage
There is no definitive experimental evidence in the literature
concerning bioaccumulation after chronic or repeated daily exposure to
chloroform. However, there are practical reasons to believe that extended
residence in body fat occurs. In man, chloroform has a relatively high fat
tissue/blood partition coefficient of 35 (Table 4-2), a long half-time of
elimination from adipose tissue compartments of =36 hours (Figure 4-3), and
it has been detected in blood and breath 24 to 72 hours after a single
exposure (Stewart, 1974; Fry et al., 1972; Stewart et al., 1965). Figure 4-5
clearly shows the slow elimination of chloroform from the adipose tissue of
dogs following a 3-hour anesthesia (Chenoweth et al., 1962). Despite the
rapid exponential decline of blood levels of chloroform within 3 hours after
termination of anesthesia, significant levels of chloroform were still
present 20 hours later.
Fry et al. (1972) have provided indirect evidence in man of the storage
of chloroform in body fat (Figure 4-4), while analysis of body fat of animals
given a single inhalation exposure or oral dosage demonstrated marked
accumulation of chloroform in this tissue (Taylor et al., 1974; Cohen and
Hood, 1969) (Tables 4-5, 4-6, 4-7). Of direct relevance to man are the
observations of two research groups: McConnell et al. (1975), who
demonstrated the occurrence of significant amounts of chloroform (and other
chlorinated hydrocarbons) in autopsy tissues (highest concentrations in body
4-31
-------
PENTOBARBITAL
TIME, hr
Figure 4-5. Blood and adipose tissue concentrations of
chloroform during and after anesthesia in a dog. Note the high
prolonged levels of chloroform in adipose tissue (broken line)
for 20 hours even after rapid exponential fall in blood con-
centration (solid line) with termination of chloroform anes-
thesia after three hours.
Source: Chenoweth et al. (1962).
fat) of humans exposed only to ordinary ambient air (Table 4-4); and of
Conkle et al. (1975), who analyzed the GC-MS alveolar air of eight fasting
healthy men working in a nonindustrial environment and found in three men
significant rates of pulmonary excretion of chloroform (Table 4-11), as well
as other halocarbons (for example, methylene chloride, dichlorobenzene,
methylchloroform) in all eight men.
4.5. BIOTRANSFORMATION OF CHLOROFORM
4.5.1. Known Metabolites
The haloforms, and chloroform in particular, long have been known to
undergo extensive mammalian biotransformation. Zeller (1883) demonstrated an
increased daily urinary inorganic chloride excretion representing 25 to 60%
of the dose in dogs given oral doses of chloroform (7 to 10 g in gelatin
capsules). Eighty years later, Van Dyke et al. (1964), using
4-32
-------
TABLE 4-11. LEVELS OF CHLOROFORM IN BREATH OF FASTED NORMAL HEALTHY MEN
Subject
A
B
C
D
E
F
G
H
Age
34
28
33
38
47
28
38
23
Chloroform excretion, pg/hr
2.0
ND
ND
ND
11.0
ND
ND
0.22
ND = not detected.
SOURCE: Conkle et al., 1975.
Cl-chloroform, confirmed in rats that the extra urinary inorganic chloride
originated from the metabolism of chloroform. Zeller (1883) also found, in
the urine of his dogs given chloroform, a levo-rotatory oxidative metabolite
that he suggested to be the glucuronide of trichloromethanol, a compound only
recently postulated as an intermediate of P^Q oxidative metabolism
(Figure 4-6). Van Dyke and coworkers (1964) also found evidence for
•^C-metabolites (-2% dose) in the urine of their rats given chloroform.
However, other investigators (Brown et al., 1974; Fry et al., 1972; Paul and
Rubinstein, 1963; Butler, 1961) with newer methodologies have not been able
to identify lesser chloromethanes in the urine or breath of mouse, rat, or
man after chloroform exposure.
4-33
-------
MAJOR AEROBIC PATHWAY
H-C CI3
P450. O2
NADPH
MICROSOMES
[HOCCI3]
ACCEPTOR
PROTEIN
I
CO -«-
H2C - CH - C - OH
S NH
\ /
C
N
O
2-OXOTHIAZOLIDINE-
4-CARBOXYLIC ACID
-HCI
O=C CI2
PHOSGENE
>
CYSTEINE
CONDENSATION
H2O
• 2 HCI + CO2
GLUTATHIONE
CONJUGATES?
MINOR ANEROBIC PATHWAY
CHCI3
ANEROBIC
NADPH
-^- BA50 - Fe2+ : C CI2 + HCI
I +H2O
P450 - Fe2+ CO -^ CO + 2 H Cl
REDUCED
MICROSOMES
Figure 4-6. Metabolic pathways of chloroform biotransformation
(Identified CH Cl_ metabolites are underlined.)
4-34
-------
In addition to the chloride ion, it has been established from both
In vivo and i_n vitro studies that the major end-product metabolite of
chloroform is carbon dioxide (C02) (Brown et al., 1974; Fry et al., 1972;
Rubinstein and Kanics, 1964; Van Dyke et al., 1964; Paul and Rubinstein,
1963), with phosgene identified as the immediate precursor metabolite from
in vitro studies (Mansuy et al., 1977; Pohl et al., 1977; Ilett et al., 1973)
(Figure 4-6). C02 from chloroform metabolism is primarily excreted through
the lungs, but a small percentage (<7%) is incorporated into endogenous
metabolites and excreted into the urine as bicarbonate, urea, methionine, and
other amino acids (Brown et al., 1974). Carbon monoxide (CO) has also been
identified as a very minor metabolite of anaerobic chloroform metabolism
(Figure 4-6), both from i_n vitro studies (Ahmed et al., 1977; Wolf et al.,
1977) and i_n vivo animal studies (Anders et al., 1978; Bellar et al., 1974).
In addition to chloroform metabolites that are excreted, phosgene and
other "reactive intermediates" of chloroform metabolism interact with and
covalently bind to tissue acceptors such as protein and lipids (Docks and
Krisna, 1976; Uehleke and Werner, 1975; Brown et al., 1974; Ilett et al.,
1973; Cohen and Hood, 1969; Reynolds, 1967; Cessi et al., 1966).
The liver is the principal site of chloroform metabolism, although Paul
and Rubinstein (1963) and Butler (1961) found that rat kidney, adipose
tissue, and skeletal muscle also converted chloroform to C02 (25, 0.8, and 8%
of liver, respectively).
Recently it has been shown that chloroform is metabolized by the kidney
via metabolic pathways similar to those in the liver, although to a lesser
extent (per unit weight of renal cortical tissue) (Smith and Hook, 1983,
1984; Pohl et al., 1984; Branchflower et al., 1984).
4-35
-------
4.5.2. Magnitude of Chloroform Metabolism
Chloroform is metabolized to differing extents in man and other animal
species. Since chloroform and other halogenated hydrocarbons are thought to
produce pathological effects by metabolism in target tissues to reactive
intermediates that covalently bind to macromolecules (Chapter 5), the total
capacity to metabolize chloroform as well as individual tissue sites of
metabolism are important determinants of expected interspecies differences in
toxic susceptibility. This interdependence between intensity of toxic
response and metabolism, and interspecies differences in magnitude of
metabolism, are important considerations in extrapolation from experimental
animal to man (Reitz et al., 1978).
Few studies have been made of the capacity of man to metabolize
chloroform; virtually no studies have been made of the pharmacokinetic,
endocrine, genetic, and environmental factors modifying metabolism in man.
The early studies of Lehmann and Hasegawa (1910) on the retention of
chloroform from inspired air inhaled by three volunteers (4500 to 5000 ppm,
average of 64%) (Table 4-3) suggest that 36% of pulmonary uptake of
chloroform in man is metabolized. Similarly, a retention value of 67%,
calculated from the data of Smith et al. (1973) for patients inhaling
10,000 ppm chloroform during surgical anesthesia (Figure 4-2), indicates 33%
chloroform metabolism during inhalation exposure. A similar estimate of the
extent of chloroform metabolism during anesthesia has been made by Feingold
and Holaday (1977). These workers simulated, with a computer, chloroform
inhalation kinetics using a nonlinear whole-body compartmental model, and
found that the percent of chloroform uptake metabolized was 30%. This rate
of metabolism remained constant during 8 hours of anesthesia, and continued
4-36
-------
for several days following termination of anesthesia, presumably from
chloroform stored during anesthesia in adipose tissue.
Fry et al. (1972) have investigated the metabolism of chloroform in man
after single oral doses. Isotopically labeled ^C-chloroform dissolved in
1.0 ml olive oil/gelatin capsule was given to 12 healthy male and female
volunteers (58 to 60 kg body weight) at doses of 0.1 to 1.0 g. For two of
the volunteers, pulmonary excretion of ^QQ,, -jn expired air from the
metabolism of chloroform was serially connected over a 7.5 hour period and
analyzed by mass spectrometry. The results given in Table 4-8 show that 49
and 51% of a 0.5 g dose was metabolized to ^3C02. Pulmonary excretion of
unchanged ^C-chloroform during a comparable period (8 hours) in a separate
experiment with these two subjects were 67 and 40%, respectively. No
metabolites other than C0Ł (e.g., methylene dichloride, tetrachloroethane)
were found in expired air, and chloroform was not found in the urine. These
results indicate that (1) virtually all of an oral chloroform dose (0.5 g)
can be accounted for by pulmonary excretion of C02 and unchanged chloroform;
(2) metabolism of chloroform to C0Ł is =50% of this dose; and (3) absorption
and metabolism are rapid and virtually complete within 5 hours as shown in
Table 4-8, possibly because of first pass through the liver. In this
respect, the kinetics of metabolism of oral doses may differ substantially
from inhalation doses. Furthermore, the data of Table 4-8 suggest that the
fraction of the dose metabolized is dose-dependent. Thus, an oral dose of
0.1 g was completely metabolized (100%), with no chloroform excreted
unchanged through the lungs; but for a 1.0 g dose, 65% was excreted and only
35% metabolized. These results suggest that metabolism is rate-limited in
man, since a diminishing proportion of dose is metabolized with increasing
dose (Table 4-8).
4-37
-------
In man, Chiou (1975) has shown that up to 38% of an oral dose of
chloroform is metabolized in the liver, and up to about 17% is excreted
intact from the lungs before the chloroform reaches the systemic circulation,
an example of a first-pass effect.
Animal experiments demonstrate a marked species difference in the
metabolism of chloroform. Early experiments in the 1960s by Paul and
Rubinstein (1963), Van Dyke et al. (1964), and Cohen and Hood (1969) with
mice and rats given ^4C-chloroform indicated a minimal metabolism of
chloroform (=4%) occurred in these species. More recent studies by Brown
and his coworkers (Brown et al., 1974; Taylor et al., 1974) have shown th&t
mice and rats metabolize chloroform to CC^ extensively (65 to 85%), and to a
greater extent than-non-human primates, or man. These investigators gave
equivalent oral doses (60 mg/kg) of ^C-chloroform to mice (3 strains), rats,
and squirrel monkeys, and determined ^4C-labeled chloroform or volatile
metabolites and ^CQ^ in expired air, ^C-radioactivity in urine, and
^C-radioactivity remaining in animals at sacrifice 4 to 8 hours after
dosing. Total recovery of ^C-chloroform radioactivity was excellent and
accounted for 93 to 98% of the administered dose. Their results are
summarized in Table 4-9. Using ^C02 as a measure of the fraction of the
chloroform dose metabolized (intermediate chloromethane metabolites were not
found in breath or urine), mice metabolized 85% of the dose, rats, 66%, and
squirrel monkeys, 28%. A further 2 to 8% of 14C-radioactivity (14C02
incorporated into urea, bicarbonate, and amino acids) were found in urine.
They found no strain difference in mice, or sex difference in mice, rats, or
monkeys in capacity to metabolize chloroform, or in tissue distribution and
binding of metabolites, except for mice, where kidney radioactivity
concentration was greater in males than females and lesser in livers of males
4-38
-------
than females (Table 4-7). In rats, also, with oral doses of 12 and 36 mg/kg
chloroform, Reynolds et al. (1984a,b) found 67% and 69% of the dose,
respectively, was metabolized to CO?, indicating at least in the rat
extensive, rate-limited (above 60 mg/kg) metabolism. These findings of Brown
and coworkers of large interspecies differences for metabolism of chloroform
and marked sex differences in mice (but not other species) for tissue
distribution and covalent binding of intermediate metabolites to tissue
macromolecules in liver and kidney emphasize the difficulties and dangers of
extrapolating studies in lower animals to man (Reitz et al., 1978).
4.5.3. Enzymic Pathways of Biotransformation
The postulation has been made that a reactive metabolite of CHC13 is
responsible for its liver and renal toxicity- in man (von Oettingen, 1964;
Conlon, 1963) and experimental animals (Bhooshan et al., 1977; Pohl et al.,
1977; Ilett et al., 1973; Klaassen and Plaa, 1966), and possibly the
production of liver tumors in mice (Eschenbrenner and Miller, 1945a) (see
generally Docks and Krishna, 1976; Uehleke and Werner, 1975; Brown et al.,
1974; Ilett et al., 1973; Reynolds, 1967; Paul and Rubinstein, 1963). For
example, when rats or mice are treated with ^C-chloroform, the extent of
hepatic necrosis parallels the amount of ^C-label bound irreversibly to
liver protein (Docks and Krishna, 1976; Brown et al., 1974; Ilett et al.,
1973). Both necrosis and binding are potentiated by pretreatment of animals
with phenobarbital, a known inducer of liver microsomal metabolism, and
inhibited by pretreatment with the inhibitor piperonyl butoxide. Chloroform
administration also decreases the level of liver glutathione in rats
pretreated with phenobarbital, further suggesting that a reactive metabolite
is produced (Docks and Krishna, 1976; Brown et al., 1974). The results of
in vitro studies with rat and mouse liver and kidney microsomes support the
4-39
-------
in vivo observations by establishing that ^C-chloroform is metabolized to a
reactive metabolite which binds covalently to microsomal protein (Bhooshan
et al., 1977; Sipes et al., 1977; Uehleke and Werner, 1975; Ilett et al.,
1973; Pohl et al., 1984; Branchflower et al., 1984; Smith and Hook, 1984).
This metabolic process is oxygen dependent and appears to be mediated by
cytochrome P450 which is inducible by phenobarbitol (Sipes et al., 1977;
Uehleke and Werner, 1975; Ilett et al., 1973).
The demonstrations by Pohl et al. (1977) and Mansuy et al. (1977) of
carbonyl chloride (phosgene) formation from chloroform by rat microsomal
preparations suggest that phosgene may be the key causal agent for these
toxic effects. The finding of Weinhouse and collaborators (Shah et al.,
1979) that phosgene is also a reactive metabolic intermediate in the
metabolism of carbon tetrachloride emphasizes basic similarities in the
metabolism and toxicities of these two chloroalkanes. Figures 4-6 and 4-7
show for comparison the currently proposed pathways of metabolism of
chloroform in liver and kidney and of carbon tetrachloride.
Figure 4-6 indicates that the initial step in the metabolism of
chloroform involves the oxidation of the aliphatic carbon (H-C) to
trichloromethanol by phenobarbital inducible cytochrome P^Q (Sipes et al.,
1977; Uehleke and Werner, 1975; Ilett et al., 1973). This metabolic step has
been suggested by Mansuy et al. (1977) and Pohl et al. (1977) as the
precursor of phosgene formed by rat microsomes i_n vitro from chloroform.
Phosgene was confirmed as a metabolite by reaction with cysteine to give
2-oxothiazolidine-4-carboxylic acid which was identified by GC-CIMS.
Trichloromethanol is highly unstable and spontaneously dehydrochlorinates to
produce phosgene (Seppelt, 1977). The electrophilic phosgene reacts with
water to yield C02, a known metabolite of CHC13 in vitro (Rubinstein and
4-40
-------
C CI3C C CI3
CHCI3
ACCEPTOR
PROTEIN
I
CCL4
REDUCTIVE
DECHLORINATION
ANAEROBIC
MICROSOMES
NADPH
P450 - Fe2+C CI4
P450 - Fe24 • C CI3 + CI
LIPOPEROXIDATION
CONJUGATION
MALONALDEHYDE
(P450-Fe34-CI3COH]
•HCI
• O-CCL2
PHOSGENE
H2O
2 H Cl + CO,
H.C-CH-COOH
I I
S NH
V
CYSTEINE
CONDENSATION
2-OXOTHIAZOLIDINE
4-CARBOXYLIC ACID
Figure M-7. Metabolic pathways of carbon tetrachloride bio-
transformation. (C Clj. metabolites identified are underlined).
Source: Shah ct al. (1979).
4-41
-------
Kanics, 1964; Paul and Rubinstein, 1963) and i_n vivo (Brown et al., 1974; Fry
et al., 1972), with protein to form a covalently bound product (Pohl et al.,
1977; Sipes et al., 1977; Uehleke and Werner, 1975; Brown et al., 1974; Ilett
et al., 1973), or with cysteine (Pohl et al., 1977), and possibly with
glutathione (Docks and Krishna, 1976; Brown et al., 1974). The finding that
deuterated chloroform (CDC13) depletes glutathione in the livers of rats less
than CHC13 supports this notion (Docks and Krishna, 1976).
The postulated oxidation of the C-H bond of chloroform by P45o to
produce trichloromethanol which spontaneously yields phosgene is further
supported by the observations of Pohl and Krishna (1978). These workers
found that chloroform metabolism to phosgene by rat liver microsomes is
oxygen and NADPH dependent, and inhibited by CO and SKF 525-A. Moreover, in
the presence of cysteine and 1802 atmosphere, *802 is incorporated into the
2-oxo position of 2-oxothiazolidine-4-carboxylic acid. Oxidative cleavage of
the C-H bond appears to be the rate-determining step, since deuterium labeled
chloroform (CDC13) is biotransformed into phosgene slower than CHC13; CDC13
appears also to be less hepatotoxic than CHC13. Pohl (1980) has further
characterized the metabolism of chloroform in rat liver microsomes by
measuring the covalent binding of 14CHC13 and C3HC13 to microsomal protein.
Chloroform does not appear to be activated by reductive dechlorination to the
radical 'CHC12, because the 3H-label does not bind to microsome protein as
does the 14C-label.
Figure 4-7 summarizes current knowledge of the biotransformation of
carbon tetrachloride. The first step is a rapid reductive formation of the
trichloromethyl ('CC13) radical by complexing with one or more of the P450
cytochromes (Shah et al., 1979; Poyer et al., 1978; Recknagel and Glende,
1973). This radical undergoes several reactions in addition to binding to
4-42
-------
lipids (Villarruel and Castro, 1975; Uehleke and Werner, 1975; Villarruel
et al., 1975; Gordis, 1969; Reynolds, 1967) and protein (Uehleke et al.,
1977; Uehleke and Werner, 1973), although not to nucleic acids (Uehleke
et al., 1977; Uehleke and Werner, 1975; Reynolds, 1967). Anaerobically, the
addition of a proton and electron yields chloroform (Glende et al., 1976;
Uehleke et al., 1973; Fowler, 1969; Butler, 1961), dimerization to
hexachloroethane (Uehleke et al., 1973; Fowler, 1969), or further reductive
dechlorination to CO via the carbene CC12 (Wolf et al., 1977). Aerobically,
the *CCl3 radical is oxidized by the P^g system to trichloromethanol
(C^COH), which is the precursor of phosgene (C12CO) that Weinhouse and
colleagues (Shah et al., 1979) have shown to be an intermediate in carbon
tetrachloride metabolism by rat liver homogenates. Hydrolytic dechlorination
of phosgene yields C02 (Shah et al., 1979).
Under normal physiological conditions (i.e., aerobic conditions), a
minimal formation of chloroform might be expected to occur. Carbon
tetrachloride yields a chloroform most readily i_n vitro under anaerobic
conditions and its formation is inhibited by oxygen (Uehleke et al., 1977;
Glende et al., 1976). Shah et al. (1979) observed that chloroform does not
compete successfully with carbon tetrachloride for initial binding to P^^Q
cytochrome (Sipes et al., 1977; Recknagel and Glende, 1973). Wolf et al.
(1977) also found that binding of chloroform to reduced cytochrome P450 was
very slow compared to that of carbon tetrachloride.
Anders and coworkers (Anders et al., 1978; Ahmed et al., 1977) have
shown that dihalomethanes and trihalomethanes, including chloroform, also
yield CO as a metabolite. Intraperitoneal administration of haloforms (1 to
4 mmoles/kg) to rats led to dose-dependent elevations in blood CO levels.
Treatment of the rats with either phenobarbital (but not 3-methyl-
4-43
-------
cholanthrene) or SKF 525-A, respectively, increased or decreased metabolism
to CO. The order of yield of CO from the iodoforms was greatest for
iodoform > bromoform > chloroform for the same dose. Thus, chloroform was
minimally metabolized to CO (i.e., to less than one-tenth of that for
iodoform or bromoform). Similar findings were made by these workers with rat
liver microsomes (Ahmed et al., 1977); they found that metabolism of the
haloforms to CO by rat liver microsomes (1) required NADPH; (2) could proceed
anaerobically but was increased 2-fold by 02; (3) was increased by
pretreatment with phenobarbital and inhibited by SKF 525-A or COC12
pretreatment; and (4) was stimulated by glutathione or cysteine addition in
both anaerobic (3-fold) or aerobic (8-fold) conditions. These results
suggested that haloforms were metabolized to CO via a cytochrome P45o
dependent system. However, chloroform was a poor substrate compared to
iodoform or bromoform, yielding <2% of its quantity of CO as formed from
equimolar concentrations of these halomethanes. Wolf et al. (1977) also
found that chloroform, to a very limited extent, was metabolized to CO by
reduced rat P^Q preparations. These workers investigated the spectral and
biochemical interactions of a series of halogenated methanes with rat liver
microsomes under anaerobic reducing conditions. Tetrahalogens (e.g., CCl^
and trihalogens (e.g., CHC^) all formed complexes with reduced cytochrome
P450 with absorption peak at 460 to 465. A shift to 454 occurred with CO
formation and subsequent complexing of CO to P45Q- CO formation required
NADPH, was higher in microsomes from phenobarbital- and 3-methylcholanthrene-
treated rats, and was not found at high oxygen concentrations (>8%). Testai
and Vittozzi (1984) have recently reported similar observations. These
investigators compared the metabolism of chloroform to protein covalent bound
metabolites and complexing with cytochrome P by rat liver microsomes
4-44
-------
maintained at pH 7.4 under anaerobic (<1% pO?) and aerobic (>6% p02)
conditions with added NADPH and generator. Covalent binding to microsomal
protein in the absence of molecular oxygen was reduced to 30% of that
established with aerobic conditions, and a greater loss of functional
cytochrome P^gg was observed. Differential spectra obtained below a 6% p02
tension showed appearance of a maximum at about 450 nm. Since anaerobically
formed metabolites were very effective in producing PQ$Q loss, it was
suggested that the spectral change was possibly due to the formation of a
chloroform metabolite-adduct with cytochrome P45Q.
Figure 4-8 shows the relative rates of CO formation from carbon
tetrachloride and other polyha-lomethanes. Chloroform, in comparison to
carbon tetrachloride, was a very poor reaction substrate, and binding of
chloroform to reduced cytochrome PQ^Q was extremely slow compared to that of
carbon tetrachloride. Wolf et al. (1977) proposed the reduction sequence
shown in Figures 4-6 and 4-7 for the reductive dechlorination of chloroform
and of carbon tetrachloride to yield CO via a carbene (CC^) intermediate.
The physiological importance of this pathway of metabolism appears to be more
significant for carbon tetrachloride than for chloroform.
4.6. COVALENT BINDING TO CELLULAR MACROMOLECULES
4.6.1. Proteins and Lip ids
Reactive intermediates of the metabolism of chloroform (phosgene,
carbene, -Cl) and carbon tetrachloride (-CC^, phosgene, carbene, 'Cl) that
irreversibly bind to cellular macromolecules (covalent binding) are generally
believed to result in an alteration of cellular integrity, which leads to
centrolobular hepatic necrosis and renal proximal tubular epithelial damage.
Chloroform, mole for mole, is generally accepted to be less hepatotoxic than
4-45
-------
a
120
100
80
60
cc
g 40
O
O
20
CHCI3. CCI3Br
10
TIME, min
15
20
Figure 4-8. Rate of carbon monoxide formation after addition of
various halomethanes to sodium dithionite-reduced liver micro-
somal preparations from phenobarbitol-treated rats. Note the low
rate of metabolism of chloroform to CO compared to carbon tetra-
chloride.
Source: Wolf et al. (1977).
4-46
-------
carbon tetrachloride (Brown, 1972; Klaassen and Plaa, 1969; Plaa et al.,
1958).
Chloroform in non-lethal doses produces renal damage in mice, dogs, and
man (Bhoosan et al., 1977; Pohl et al., 1977; Ilett et al., 1973; Klaassen
and Plaa, 1966, 1967; Bennet and Whigham, 1964; von Oettingen, 1964; Conlon,
1963; Plaa et al., 1958; Culliford and Hewitt, 1957; Hewitt, 1956; Shubik and
Ritchie, 1953); whereas in experimental animals, carbon tetrachloride does
not do so (Storms, 1973; Klaassen and Plaa, 1966; Plaa and Larson, 1965;
Bennet and Whigham, 1964; Culliford and Hewitt, 1957; Hewitt, 1956; Shubik
and Ritchie, 1953), although it does in man (New et al., 1962; Guild et al.,
1958). To explain these species differences in toxicity as well as known
intraspecies (Hill et al., 1975; Deringer et al., 1953; Shubik and Ritchie,
1953) and sex differences (Taylor et al., 1974; Ilett et al., 1973; Bennet
and Whigham, 1964; Culliford and Hewitt, 1957; Hewitt, 1956; Deringer et al.,
1953; Shubik and Ritchie, 1953; Eschenbrenner and Miller, 1945b), prevailing
concepts implicate (1) differences in the rates of metabolism and organ
system capacities for metabolism, which in turn determine the amount of
irreversible macromolecular binding and (2) differences in the enzyme
pathways for metabolism of the two haloalkanes (Figures 4-6 and 4-7).
Carbon tetrachloride, by a reductive dechlorination via complexing with
reduced P45Q, yields the trichloromethyl free radical (-CC^) (Recknagel and
Glende, 1973; Slater, 1972) (Figure 4-7), which can covalently bind to lipid
and protein (Shah et al., 1979; Villarruel et al., 1975; Castro and Diaz
Gomez, 1972; Reynolds, 1967), and can also initiate peroxidation of polyenoic
fatty acids (Slater, 1972; Recknagel and Ghoshal, 1966). Chloroform does not
appear to be activated to free radicals ('CC^ or -CHC^), but does bind
covalently to liver lipid and protein (Sipes et al., 1977; Docks and Krishna,
4-47
-------
1976; Uehleke and Werner, 1975; Brown et al., 1974; Ilett et al., 1973), and
initiates lipid peroxidation in some circumstances (Koch et al., 1974; Ilett,
1973; Brown, 1972; Slater, 1972). Several investigators have shown that
diene conjugates (products of lipoperoxidation) are not increased in vivo in
normal rats when chloroform is inhaled or injected intraperitoneally (Brown
et al., 1974; Brown, 1972; Klaassen and Plaa, 1969), but only when rats are
pretreated with phenobarbital and the metabolism of chloroform is greatly
enhanced (Brown et al., 1974; Brown, 1972). Studies in vitro show diene
conjugation and malonaldehyde formation (an index of lipoperoxidation) by
microsomes of phenobarbital-pretreated rats were not increased but decreased
by the addition of chloroform (Brown, 1972; Klaassen and Plaa, 1969),
suggesting that with isolated microsomes, metabolism of chloroform is too
small for sufficient quantities of reactive intermediates to accumulate and
initiate lipoperoxidation. However, Rubinstein and Kanics (1964) found
chloroform to be more rapidly metabolized by rat microsomal fractions than
carbon tetrachloride (see also Table 4-12). These findings indicate that
differences in metabolic activation [carbon tetrachloride to produce free
radicals (Figure 4-7), but chloroform primarily to phosgene (Figure 4-6)]
explain the greater potential of carbon tetrachloride for initiating
lipoperoxidation. Table 4-12 shows the data of Uehleke et al. (Uehleke
et al., 1977; Uehleke and Werner, 1975) for the covalent binding of rabbit
microsomes following incubation with ^C-labeled chloroform and carbon
tetrachloride. Both protein and lipid binding of ^4C-radioactivity are
4-fold and 20-fold, respectively, more extensive for carbon tetrachloride
than chloroform; lipids are labeled preferentially by carbon tetrachloride
but are not by chloroform. Furthermore, covalent binding from chloroform
metabolism occurs mainly with anaerobic conditions (a minor metabolic
4-48
-------
TABLE 4-12. COVALENT BINDING OF RADIOACTIVITY FROM 14C-CHLOROFORM AND
14C-CARBON TETRACHLORIDE IN MICROSOMAL INCUBATION IN VITRO
(nmol/mg in 60 min)
Microsomal
14C-CC14
14C-CHC13
Incubation
condition
N2
N2
02
Protein
20.0
5.1
8.5
Lip id
76.0
4.1
7.0
To added
serum albumin
1.4
0.9
1.7
aMicrosomes from phenobarbital pretreated rabbits.
SOURCE: Uehleke et al., 1977.
pathway) and is not greatly increased with aerobic metabolism, the major
pathway for metabolism of chloroform, which is Ł>2 dependent (Figure 4-6).
Covalent binding occurs preferentially to lipids and proteins of the
endoplasmic reticulum proximate to P^Q system for metabolism. However,
considerable covalent binding from chloroform metabolites occurs in other
cell fractions of liver.and kidney, particularly to mitochondria (Uehleke and
Werner, 1975; Hill et al., 1975). Hill et al. (1975) found when C57BL male
mice were injected interperitoneally with 0.07 ml/kg ^4C-chloroform in oil
and sacrificed 12 hours later, that in the liver, 50% of the radioactivity
was irreversibly bound to microsome, 23% to mitochondria, 25% to cytosol, and
<2% to nuclei; for kidney, 38% of radioactivity was bound to microsome, 39%
to mitochondria, 22% to cytosol, and <2% to nuclei. A similar distribution
was found in male NMRI mice by Uehleke and Werner (1975), who observed
minimal binding to microsomal RNA but significant binding to nicotine-adenine
nucleotides. The data of Ilett et al. (1973), shown in Figure 4-9,
4-49
-------
**
o
a
O»
"5
c
o
z
ffi
H
O
u
23456
CHLOROFORM DOSE, mmol/kg
Figure 4-9. Effect of increasing dosage of i.p.-injected
14
C-chlorofonn on,extent of covalent binding of radioactivity in
vivo to liver and kidney proteins of male mice 6 hours after
administration.
Source: Ilett et al. (1973).
4-50
-------
demonstrate that in C57 BL/6 mice, the amount of covalent binding in liver
and kidney microsomal fractions increases proportionally with the chloroform
dose.
4.6.1.1. Genetic Strain Difference—Hill et al. (1975) described in mice two
genetic variations in chloroform toxicity paralleling genetic differences in
covalent binding in liver and kidney. In one inbred strain (DBA/2), the male
animals were 4 times more sensitive to the lethal effects of oral doses of
chloroform (LD50 of 0.08 ml/kg) than the second strain (C57 BL/6, LD50 of
0.33 ml/kg). Males of the F^ hybrid strain (B6D2F^/J) had an intermediate
LDgQ of 0.2 ml/kg, midway between those of the two parental strains. The
susceptibility of DBA mice was-related to a dose-dependent necrosis of the
proximal convoluted renal tubules. However, mice of all three genotypes that
received >0.17 ml/kg chloroform exhibited both renal tubular necrosis and
hepatic centrolobular necrosis. Males and females of the same strain
exhibited similar dose thresholds to hepatic damage, but females died of
chloroform-induced hepatic damage without developing renal lesions. This
sex-related absolute difference is dependent on androgen profile of the mice;
testosterone-treated females become sensitive to renal toxicity (Bennet and
Whigham, 1964; Culliford and Hewitt, 1957; Eschenbrenner and Miller, 1945b).
Table 4-13 shows the extent of covalent binding in liver and kidney of
these three strains after a single intraperitoneal injection of
14C-chloroform (0.07 ml/kg) to the males. Kidney homogenates from DBA/2J
male mice, more sensitive to renal necrosis, contained more than 2-fold as
much radioactivity as those from resistant C57BL/6J; covalent binding in the
FI hybrid was intermediate, as expected. A significant difference was also
noted in labeling of kidney subcellular fractions. While all subcellular
fractions of susceptible male DBA mice were labeled to a greater extent than
4-51
-------
TABLE 4-13 MOUSE STRAIN DIFFERENCE IN COVALENT BINDING OF RADIOACTIVITY
FROM HC-CHLOROFORMa
Tissue
Liver
Kidney
Liver
Nuclei
Mitochondria
Microsomes
Cell sap
Kidney
Nuclei
Mitochondria
Microsomes
Cell sap
Specific activity
DBA
Tissue homoqenates
0.82
2.41
Subcellular fractions
0.67
1.14
0.64
0.98
2.20
3.67
1.74
1.65
relative to C57BL
Fl
0.96
1.64
0.76
1.14
0.73
1.09
1.67
1.97
1.44
1.23
C57BL
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
aAdult male mice of each genotype given l^C-chloroform (0.07 ml/kg)
intraperitoneally and sacrificed 12 hr later. Genotype comparisons are
given as ratio of radioactivity to C57BL = 1.
SOURCE: Hill et al., 1975.
Fj_ or C57BL strains, the greatest increase was in labeling of the
mitochondria! fraction.
In the liver, the distribution of covalent binding was generally
opposite to that observed in the kidneys (Table 4-13), but neither liver
homogenates nor subcellular fractions showed significant strain differences.
4-52
-------
4.6.1.2. Sex Difference—Kidneys of male mice are known to covalently bind
more 14C-chloroform radioactivity than do those of females, but females bind
more in the liver than males (Taylor et al., 1974; Ilett et al., 1973) (see
Tables 4-7 and 4-14). Table 4-15 shows that pretreatment of male mice with
phenobarbital increases covalent binding in the liver but not in the kidney
(Ilett et al., 1973). A similar observation has been made by Kluwe et al.
(1978) in male mice. They found that phenobarbital increased liver but not
kidney microsomal activity; 3-methylcholanthrene, dioxin, and PCGs increased
both liver and kidney microsomal enzyme activities. From the renal and
hepatic toxicity profile to chloroform displayed by mice treated with these
various inducers, these investigators concluded that the chloroform
metabolite(s) responsible for hepatic damage is probably generated in the
liver, and the metabolite(s) responsible for renal damage is generated in the
kidney. However, it has now been firmly established that chloroform is
metabolized by the kidney to phosgene and to other reactive compounds by
enzyme pathways similar to those in the liver (Pohl et al., 1984;
Branchflower et al., 1984; Smith and Hook, 1984; for a discussion, see
Section 5.3.2., Nephrotoxicity).
4.6.1.3. Inter-species Difference—In addition to intra-species strain
(mice) differences in covalent binding noted above, Uehleke and Werner (1975)
have also observed an apparent inter-species difference. Figure 4-10 shows
the i_n vitro binding of radioactivity from ^C-chloroform by microsomal
preparations from rat, mouse, rabbit, and man. Human and rabbit microsomes
have the highest rate of covalent binding from chloroform, with the mouse
followed by the rat considerably lower. Inter-species differences in the
covalent binding rates for carbon tetrachloride were small. These species
differences in binding of chloroform metabolites to protein and lipid
4-53
-------
TABLE 4-14. IN VIVO COVALENT BINDING OF RADIOACTIVITY FROM 14CHC13 IN LIVER
AND KIDNEY OF MALE AND FEMALE C57BL/6 MICE (nmoles/mg protein ± S.E.)a
Male Female
Liver 2.92 ± 0.35 3.66 ± 0.39
Kidney 2.34 ± 0.16 0.39 ± 0.02
aMice were sacrificed 6 hr after intraperitoneal administration of
3.72 nmoles/kg of 14CHC13-
SOURCE: Ilett et al., 1973.
TABLE 4-15. IN VITRO COVALENT BINDING OF RADIOACTIVITY FROM 14CHC13 TO
MICROSOMAL PROTEIN FROM LIVER AND KIDNEY OF MALE AND FEMALE MICE
(pmoles/mg protein/5 min ± S.E.M.)
Pretreatment
Male NA
Male Phenobarbital
Female NA
Liver Kidney
572 ± 54 44.6 ± 4.1
1454 ± 143 41.0 ± 3.2
419 ± 20 14.6 ± 2.5
NA = not applicable.
SOURCE: Ilett et al. 1973.
4-54
-------
O HUMAN
D RABBIT
A MOUSE
O RAT
30 40 50 60 70
TIME, min
Figure 4-10. Comparison of irreversible binding of radioactivity
from C-CHCl^ to protein and lipid of microsomes from normal
rabbit, rat, mouse, and human liver incubated in vitro at 37°C in
°2*
Source: Uehleke and Werner (1975).
4-55
-------
in vitro do not, however, parallel the species differences in metabolism of
chloroform i_n vivo, as measured by the conversion of chloroform to C02
(Table 4-9); In vivo, the mouse has the greatest capacity to metabolize
chloroform (80%), followed by rats (65%), nonhuman primates (20%), and man
(30 to 50%).
4.6.1.4. Age Difference—Uehleke and Werner (1975) have shown that
irreversible protein binding of radioactivity from 14C-chloroform and
l^C-carbon tetrachloride to liver microsomes of newborn rats (18 hours old)
is low compared to that of microsomes from adult rats (32 days); however, the
binding was shown to be proportional to PQ$Q content of the microsomes, which
was proportionally low in microsomes from newborn rats.
4.6.2. Nucleic Acids
P45Q systems activate chloroform and carbon tetrachloride i_n vivo and
in vitro to reactive metabolites that extensively covalently bind to proteins
and lipids, but do so only minimally to nucleic acid (Uehleke et al., 1977;
Wolf et al., 1977; Uehleke and Werner, 1975; Fowler, 1969; Reynolds, 1967),
unlike many other carcinogens that bind DMA. Reitz et al. (1980) measured
DMA alkylation in liver and kidneys of mice after an oral dose of 240 mg/kg
^C-chloroform (specific activity not given) and found ^values of 3 x 10~4 and
1 x 10~4 mol % for liver and kidney DMA, respectively. These workers judged
that chloroform has very little direct interaction with DNA when compared to
known carcinogens, as reported in the literature for dimethylnitrosamine
(3.5 x 10'1 mol % alkylation, liver DNA), dimethylhydralazine (2.6 x 10~2,
colon DNA) and N-methyl-N-nitrosourea (1.5 x 10'1, brain DNA) but given by
parenteral routes (Pegg and Hui, 1978; Cooper et al., 1978; Kleihues and
Margison, 1974). The failure of chloroform or carbon tetrachloride reactive
species to significantly bind DNA has been ascribed to their short half-life
4-56
-------
compared to epoxides and to their lack of nuclear penetration. Recently,
however, Diaz Gomez and Castro (1980) have shown that highly purified rat
liver nuclear preparations are able to anaerobically activate carbon
tetrachloride, and to aerobically activate chloroform to reactive metabolites
that bind to nuclear lipids and proteins. Their data, given in Table 4-16,
show that activity in nuclear preparations is smaller than in microsomes, but
within the same order of magnitude. These results might be relevant to the
hepatocarcinogenic effects of chloroform and carbon tetrachloride in mice and
rats, since the nuclear targets (DMA, RNA, nuclear proteins) are in the
immediate neighborhood sites of activation, thus making unnecessary the
present_assumption that the highly reactive intermediates (-CC^, phosgene,
malonaldehyde, or carbene), produced at the endoplasmic reticulum, must
travel to the nucleus.
Pereira et al. (1984) have investigated the effects of chloroform on
hepatic and renal DNA synthesis in rats and mice and on ornithine
decarboxylase (OD) activity, which is a marker enzyme for cellular
proliferation, DNA synthesis, and tumorigenesis. Chloroform given i.p.
induced a dose-dependent increase of hepatic OD activity 18 hr after
tnjection. For nrke a 10-fold increase of 00 activity was found at a maximal
dose of 375 mg/kg chloroform, and in rats a 52-fold increase was found at
750 mg/kg. Hepatic and renal DNA synthesis rate (as measured by 3H-thymidine
incorporation) was increased in mice but decreased in rats. Thus, OD
activity increase in rat liver was not paralleled by an increase in DNA
synthesis. Furthermore, minimal microscopic changes were detected in the
liver and kidney under the conditions of the experiments. These
investigators interpreted their results to mean that the induction by
chloroform of liver and renal OD activity was not associated with a cellular
4-57
-------
TABLE 4-16. COVALENT BINDING OF RADIOACTIVITY FROM HC-CHLOROFORM AND
14C-CARBON TETRACHLORIDE IN RAT LIVER NUCLEAR AND MICROSOMAL INCUBATION
M VITRO (pmol/mg ± S.D.)
Incubation
condition
Protein
Lip id
14C-CC14
Nuclear
Microsomal
14C-CHC13
Nuclear
Microsomal
02
21.9 ± 2.5
50.3 ± 4
27.0 ± 3
68.0 ± 9
147 ± 12
190 ± 11
20 ± 3
57 ± 8
SOURCE: Diaz Gomez and Castro, 1980.
regenerative response and hence their results do not support a non-genotoxic
mechanism for chloroform tumorigenesis.
4.6.3. Role of Phosgene
Phosgene is a prominent intermediate of both chloroform and carbon
tetrachloride metabolisms (Figures 4-6, 4-7). It is known to be 'highly
reactive and toxic to cells and tissues (Pawlowsky and Frosolono, 1977), and
its two highly reactive chlorines suggest that it could act on cellular
macromolecules similar to bifunctional alkylating agents. Reynolds (1967)
showed that ^C-phosgene, given to intact rats, labeled liver protein (and
lipids to a smaller extent). The pattern of labeling was quite different
from that of ^C-carbon tetrachloride and more similar to ^C-chloroform.
Moreover, 36Cl-carbon tetrachloride radioactivity was also stably
incorporated into liver lipid and protein, pointing to the 'CC^ radical
rather than phosgene as the reactive form for carbon tetrachloride that
4-58
-------
labels lipid. Cessi et al. (1966) also reported that ^C-phosgene labeled
terminal amino group of polypeptides in a manner similar to i_n vivo protein
labeling produced by carbon tetrachloride.
4.6.4. Role of Glutathione
Ekstrom and Hogberg (1980) found that chloroform, in freshly isolated
rat liver cells, induced depletion of cellular glutathione. Brown et al.
(1974) demonstrated that exposure of rats to an atmosphere of 0.5% chloroform
for 2 hours markedly decreased glutathione (GSH) in the liver when the
animals were pretreated with phenobarbital to stimulate metabolism. GSH
liver content of untreated rats was not decreased. Phenobarbital
pretreatment has been shown to markedly potentiate toxicity of both
chloroform and carbon tetrachloride in rats (Docks and Krishna, 1976; Cornish
et al., 1973; McLean, 1970; Scholler, 1970). However, it has not been
possible to detect a decrease in the liver glutathione levels following
administration of carbon tetrachloride or trichlorobromomethane (Docks and
Krishna, 1976; Boyland and Chasseaud, 1970). Sipes et al. (1977) have shown
that the addition of GSH liver microsomes from phenobarbital pretreated rats
incubated in vitro with ^C-labeled halocarbons, chloroform, carbon
tetrachloride, and trichloromomethane, inhibited covalent binding =80% for
all three compounds. Their results, given in Table 4-17, also show the
effects of anaerobic and aerobic conditions on covalent binding. The
reduction in binding of chloroform by an atmosphere of N2 suggests that its
bioactivation is mediated by a cytochrome P450 oxidative pathway to phosgene
(Figure 4-6), while the enhanced binding of carbon tetrachloride in N2
reflects P45Q mediated reductive pathways (Figure 4-7) and formation of free
radical. These investigators suggest that in phenobarbital-treated animals,
chloroform depletes liver GSH by the formation of conjugate between the
4-59
-------
TABLE 4-17. EFFECT OF GLUTATHIONE, AIR, N2 OR C0:02 ATMOSPHERE ON
THE IN VITRO COVALENT BINDING OF CC14, CHCla and CBrCls TO RAT LIVER
MICROSOMAL PROTEIN (pmoles l4C-bound/mg microsomal protein/minute)
Substrate^
Incubation conditions
Air
N2
C0:02 (8:2)
SKF 525 A (0.5 mM) , Air
Glutathione, Air
NADPH omitted, Air
CC14
97 ±
310 ±
18 ±
109 ±
17 "±
6 ±
10
51
1
5
2
1
CHC13
59 ± 5
21 ± 1
20 ± 1
7 ± 7
15 ± 2
3 ± 0
CBrCl3
1456 ± 66
1370 ± 143
853 ± 62
2105 ± 159
218 ± 25
65 ± 13
a!4C-labeled substrate is a final concentration of 1 x 10~3 M incubated at
37°C. Microsomes were from phenobarbital pretreated rats.
SOURCE: Sipes et al., 1977.
reactive intermediate phosgene and GSH (Docks and Krishna, 1976; Brown
et al., 1974). In the case of carbon tetrachloride, they suggest that GSH
addition i_n vitro (Table 4-16) also conjugates with the phosgene metabolite
of carbon tetrachloride produced when incubated in air, but, in addition, GSH
decreases the levels of -CC13 by reducing the free radical to chloroform.
In vivo, it is postulated that the oxidized glutathione may be reduced back
to reduced GSH by glutatione reductase; this would explain the absence of
falling liver GSH content with carbon tetrachloride in vivo (Gillette, 1972).
Thus, the toxic effects of chloroform and carbon tetrachloride may be
mediated through different mechanisms of covalent binding, and GSH may play
4-60
-------
different roles for these chlorocarbons in preventing covalent binding of
reactive intermediates of metabolism.
Chloroform and carbon tetrachloride are known to cause greater liver
damage in fasted animals than in fed animals (Diaz Gomez et al., 1975; Jaeger
et al., 1975; Krishnan and Stenger, 1966; Goldschmidt et al., 1939; Davis and
Whipple, 1919). For chloroform, a decreased content of hepatic GSH from
fasting has been postulated to be responsible for the increased
susceptibility of fasted mice (Docks and Krishna, 1976; Brown et al., 1974).
Nakajima and Sato (1979) have recently offered an additional explanation.
These investigators studied the metabolism of the chlorocarbons j_n vitro with
microsomes from livers of fasted rats, and found that the Disappearance of
chloroform from incubation increased 3-fold for a 24 hour fast, although
fasting produced no significant increase in the microsomal protein and
cytochrome P^Q liver contents (Table 4-18). Similar results were obtained
for carbon tetrachloride. These observations suggest that the increased
toxicity of chloroform and carbon tetrachloride from food deprivation may be
due not only to decreased GSH, but also to a greater production of reactive
intermediates and covalent binding to cellular macromolecules.
4.7 SUMMARY
At ambient temperatures, chloroform is a volatile liquid with high lipid
solubility and appreciable solubility in water. Hence, chloroform is readily
absorbed into the body through the lungs and intestinal mucosa; the portals
of entry with exposure from air, water and food. Few data are available on
the pharmacokinetics of absorption and excretion of chloroform in man,
particularly at the low exposure concentrations expected in ambient air and
drinking water. However, studies show absorption from the gastrointestinal
tract in man, monkeys, rats and mice is rapid and complete, occurring by
4-61
-------
TABLE 4-18. EFFECTS OF 24-HOUR FOOD DEPRIVATION ON CHLOROFORM AND CARBON TETRACHLORIDE ]_N VITRO
MICROSOMAL METABOLISM, PROTEIN, AND P-450 LIVER CONTENTS OF RATS
i
en
Chloroform
Carbon
tetrachloride
Male
Female
Fed
Fasted
Ratio
Fed
Fasted
Ratio
Metabolism, nmole/g/min
19.7 ± 2.6 55.1 ± 7.5 2.8 15.3 ± 6.8 39.3 ± 2.5 2.6
1.9 ± 0.2 5.9 ± 0.8 3.1 1.1 ± 0.5 4.5 ± 0.3 4.1
Protein content, mg/kg liver
27.7 ± 3.7 23.0 ± 2.7 NR 22.5 ± 1.5 23.7 ± 1.7 NR
P-450, nmol/mg protein
0.842 ± 0.123 0.823 ± 0.03 NR 0.638 ± 0.051 0.673 ± 0.044 NR
NR = not reported.
SOURCE: Nakajima and Sato, 1979.
-------
first-order passive absorptive processes. A dose-dependent first-pass effect
with pulmonary elimination of unchanged chloroform occurs with oral ingestion
in man, thus decreasing the amount of chloroform reaching the systemic
circulation. In rats, the kinetics of peroral absorption are also influenced
by the dosing vehicle; the absorption rate is decreased for chloroform given
in corn oil vehicle as compared to an aqueous solution. Pulmonary uptake and
elimination occur also by first-order diffusion processes with three distinct
components with rate constants corresponding to tissue loading or
desaturation of at least three major body compartments. Half-times in man
have been found to be approximately 14-30 minutes, 90 minutes, and
24-36 hours, respectively. The longest half-time is associated with the
lipids and the adipose tissue compartment. During inhalation exposure, at
equilibrium with inspired air concentration, the blood/air partition
coefficient is about 8 at 37°C and the adipose tissue/blood partition
coefficient is 280 at 37°C. The quantity of chloroform absorbed is dependent
also on body weight and fat content of the body.
Tissue distribution of chloroform is consistent with its lipophilic
nature and modest water solubility. This chloroalkane readily crosses the
blood brain and placental barriers and distributes into breast milk.
Concentrations occurring in all major tissue organs are dose related to
inspired air concentrations or to oral dosage. Relative tissue
concentrations occur in the order of adipose tissue > brain > liver > kidney
> blood.
Elimination of chloroform from the body occurs by two major and parallel
occurring processes: (1) pulmonary elimination of unchanged chloroform by
first order kinetics and (2) metabolism of chloroform. Chloroform is
metabolized in the liver and to a lesser extent in the kidneys and other
4-63
-------
tissues. Metabolism is dose-dependent and saturable, with a greater
proportion of small doses being metabolized. There are striking differences
in the pharmacokinetics and quantitative metabolism of chloroform in man as
compared to other animals. For large steady-state body burdens, 30-40% is
metabolized by man, 20% by the nonhuman primate, >65% by the rat, and >85% by
the mouse. Metabolism produces phosgene and other putative reactive
metabolites that covalently bind extensively to cellular lipids and proteins,
although not significantly to DNA or other nucleic acids. The intensity of
metabolite binding and organ localization parallel the acute cellular
toxicity of chloroform in liver and kidney observed in experimental animals.
Both binding and toxicity are highly dependent on animal species and genetic
strain, as we 1-1 as on sw and age. An additional variable is the tissue
level of reduced glutathione, which plays an important role in protecting
against both binding and toxicity. Conversely, inducers of hepatic and renal
P450 metabolizing systems increase binding and toxicity.
4-64
-------
4.8 REFERENCES FOR CHAPTER 4
Ahmed, A.E.; Kubic, V.L.; Anders; M.W. (1977) Metabolism of haloforms to
carbon monoxide. I. I_n vitro studies. Drug Metab. Dispos. 5:198-204.
Anders, M.W.; Stevens, J.L.; Sprague, R.W.; Shaath Z.; Ahmed, A.E. (1978)
Metabolism of haloforms to carbon monoxide. II. In vivo studies. Drug
Metab. Dispos. 6:556-560.
Bellar, T.A.; Lichtenburg, J.J.; Kroner, R.C. (1974) Occurrence of
organohalides in chlorinated drinking waters. Amer. Water Works Assoc.
J. 66:703-706.
Bennet, R.A.; Whigham; A. (1964) Chloroform sensitivity of mice. Nature
204:1328.
Bhoosha, B.; Pohl; L.R.; Krishna, G. (1977) A biochemical mechanism for
chloroform activation. Fed. Proc. (Abstr.) 36:396.
Boyland, E.; Chasseaud, L.F. (1970) The effect of some carbonyl compounds
on rat liver glutathione levels. Biochem. Pharmacol. 19:1526-1528.
Branchflower, R.V.; Nunn, D.S.; Highet, R.J.; Smith, J.H.; Hook, J.B.;
Pohl, L.R. (1984) Nephrotoxicity of chloroform: metabolism to
phosgene by the mouse kidney. Toxicol. Appl. Pharmacol. 72:159-168.
Brown, D.M.; Langley, P.F.; Smith, D.; Taylor, D.C. (1974a) Metabolism of
chloroform. I. The metabolism of 14C-chloroform by different species.
Xenobiotica 4:151-163.
Brown, B.R.; Sipes K.G.; Sagalyn, A.M. (1974b) Mechanisms of acute hepatic
toxicity: chloroform, halothane and glutathione. Anesthesiology
41:554-561.
Brown, S.S. (1972) In: Mathway, D.C., ed. Foreign compound metabolism in
mammals. Vol. 2. Interactions of drugs and foreign compounds. London:
The Chemical Society; p. 456-474.
Butler, T.C. (1961) The reduction of carbon tetrachloride in vivo and
reduction of carbon tetrachloride to chloroform i_n vitro by tissues and
tissue constituents. J. Pharmacol. Exp. Ther. 134:311-319.
Castro, J.A.; Diaz Gomez, M.I. (1972) Studies on irreversible binding of
14C-CHC13 to microsomal lipids in rats under varying experimental
conditions. Toxicol. Appl. Pharmacol. 23:541-552.
Cessi, C.; Columbini, C.; Mameli, L. (1966) The reaction of liver proteins
with a metabolite of carbon tetrachloride. Biochem. J. 101:46c.
4-65
-------
Chenoweth, M.B.; Robertson, D.N.; Erly, D.S.; Goekhe, K. (1962) Blood and
tissue levels of ether, chloroform, halothane and methoxyflurane in
dogs. Anesthesiology 23:101-106.
Chiou, W.L. (1975) Quantitation of hepatic and pulmonary first-pass effects
and its implication in pharmacokinetic study. I. Pharmacokinetics of
chloroform in man. J. Pharmac. Biopharmaceut. 3:193-201.
Cohen, E.N.; Hood, W. (1969) Application of low-temperature
autoradiography studies of the uptake and metabolism of volatile
anesthetics in the mouse. Anesthesiology 30:306-314.
Conkle, J.P-; Camp, B.J.; Welch, B.E. (1975) Trace composition of human
respiratory gas. Arch. Environ. Health 30:290-295.
Conlon, M.F. (1963) Addiction to chlorodyne. Br. Med. J. 2:1177-1178.
Copper, H.K.; Buecheler, J.; Kleihues, P. (1978) DMA alkylation in mice
with genetically different susceptibility to 1,2-dimethylhydrazine-
induced colon carcinogenesis. Cancer Res. 38:3063-3065.
Cornish, H.H.; Ling, B.P.; Barth, M.L. (1973) Phenobarbital and organic
solvent toxicity. Am. Ind. Hyg. 34:487-492.
Cu111 ford, D.; Hewitt, H.B. (1957) The influence of sex hormone status on
the susceptibility of mice to chloroform-induced necrosis of renal
tubules. J. Endocrinol. 14:381-393.
Davis, N.; Whipple, C. (1919) The influence of fasting and various diets on
the liver injury effected by chloroform anesthesia. Arch. Int. Med.
23:612-633.
Deringer, M.K.; Dunn T.B.; Heston, W.E. (1953) Results of strain C3H mice
to chloroform. Proc. Soc. Exp. Biol. Med. 83:474-479.
Diaz Gomez, M.I.; Castro, J.A. (1980) Nuclear activation of carbon
tetrachloride on chloroform. Res. Comm. Chem. Path. Pharmacol.
27:191-194.
Diaz Gomez, M.I.; DeCastro, C.R.; DeFerreyra, E.C.; D'Acosta, N.;
DeFerros, O.M.; Castro, J.A. (1975) Mechanistic studies on carbon
tetrachloride hepatotoxicity in fasted and fed rats. Toxicol. Appl.
Pharmacol. 32:101-108.
Diem, K.; Lentner, C. (1973) Geigy Scientific Tables, eds. 7th ed.
Basle: Ciba-Geigy.
Dilley, J.V.; Chernoff, N.; Kay, D.; Winslow, N.; Newell, G.W. (1977)
Inhalation teratology studies of five chemicals in rats. Toxicol. Appl.
Pharmacol. 41:196.
Docks, E.L.; Krisna, G. (1976) The role of glutathione on chloroform-
induced hepatotoxicity. Exp. Mol. Pathol. 24:13-22.
4-66
-------
Dowty, B.J.; Carlisle O.R.; Laseter, L.L. (1975) New Orleans drinking water
sources tested by gas chromatography-mass spectrometry. Occurrence and
origin of aromatics and halogenated aliphatic hydrocarbons. Environ.
Sci. Technol. 9:762-765.
Droz, P.O.; Fernandez, J.G. (1977) Effect of physical workload on retention
and metabolism of inhaled organic solvents. A comparative theoretical
approach and its application with regards to exposure monitoring. Int.
Arch. Occup. Environ. Health 38:231-246.
Eger, E.I. (1963) Mathematical model of uptake and distribution. In:
Papper, E.M.; Kitz, R.J., eds. Uptake and distribution of anesthetic
agents. New York: McGraw-Hill; pp. 72-103.
Ekstrom, T.; Hogberg, J. (1980) Chloroform induced glutathione depletion
and toxicity in freshly isolated hepatocytes. Biochem. Pharmacol.
29:3059-3064.
Eschenbrenner, A.B.; Miller, E. (1945a) Induction of hepatomas in mice by
repeated oral administration of chloroform with observations on sex
differences. J. Natl. Cancer Inst. 5:251-255.
Echenbrenner, A.B. Miller, E. (1945b) Sex differences in kidney morphology
and chloroform necrosis. Science 102:302-303.
Feingold, A.; Holaday, D.A. (1977) The pharmacokinetics of metabolism of
inhalation anaesthetics. Br. J. Anaesth. 49:155-162.
Fiserova-Bergerova, V.; Vlach, J.; Cassady, J.C. (1980) Predictable
"individual differences" in uptake and excretion of gases and lipid
soluble vapours simulations study. Br. J. Ind. Med. 37:42-49.
Fiserova-Bergerova, V.; Holaday, D.A. (1979) Uptake and clearance of
inhalation anesthetics in man. Drug. Metab. Rev. 9:43-60.
Fiserova-Bergerova, V.; Vlach, J.; Singhal, K. (1974) Simulation and
prediction of uptake, distribution, and exhalation of organic solvents.
Br. J. Ind. Med. 31:45-52.
Fishbein, L. (1979) Potential halogenated industrial carcinogenic and
mutagenic chemicals. II. Halogenated saturated hydrocarbons. Sci.
Total Environ. 11:163.
Fowler, S.J.L. (1969) Carbon tetrachloride metabolism in the rabbit.
Br. J. Pharmacol. 37:733-737.
Fry, J.; Taylor, T.; Hathaway, D.F. (1972) Pulmonary elimination of
chloroform and its metabolite in man. Arch. Int. Pharmacodyn.
196:98-111.
Geigy Scientific Tables. (1973) Diem, K. and Lentner, C., eds. 7th ed.
Basel: Ciba-Geigy.
4-67
-------
Gettler, A.O. (1934) Medicolegal aspects of deaths associated with
chloroform or ether. Am. J. Surg. 26:168-174.
Gettler, A.O.; Blume, H. (1931) Chloroform in the brain, lungs and liver.
Quantitative recovery and determination. Arch. Pathol. 11:554-560.
Gillette, J.R. (1972) Pharmacology and the future of man. In: Proc.
5th Internat. Cong. Pharmacol, San Francisco. Basel: Karger.
Glende, E.A., Jr.; Hruszkewycz, A.M.; Recknagel, H.L. (1976) Critical role
of lipid peroxidation in carbon tetrachloride-induced loss of
aminopyridine demethylase, cytochrome P450 and glucose 6-phosphatase.
Biochem. Pharmacol. 25:2163-2170.
Goldschmidt, S.; Vars, N.M.; Ravdin, I. (1939) The influence of the
foodstuffs upon the susceptibility of the liver to injury by chloroform
and the probable mechanism of their action. J. Clin. Invest. 18:277-
289.
Gordis, E. (1969) Lipid metabolites of carbon terachloride. J. Clin.
Invest. 48:203-209.
Guild, W.R.; Young, J.V.; Merrill, J.P. (1958) Anuria due to carbon
tetrachloride intoxication. Ann. Int. Med. 48:1221-1227.
Hewitt, H.B. (1956) Renal necrosis in mice after accidental exposure
to chloroform. Br. J. Exp. Path. 37:32-29.
Hill, R.N.; Clemens, T.L.; Liu, O.K.; Vesell, E.S.; Johnson, W.D. (1975)
Genetic control of chloroform toxicity in mice. Science 190:159-160.
Ilett, K.F.; Reid, W.D.; Sipes, I.G.; Krishna, G. (1973) Chloroform
toxicity in mice: correlation of renal and hepatic necrosis with
covalent binding of metabolites to tissue macromolecules. Exp. Mol.
Pathol. 19:215-229.
Jaeger, R.J.; Connolly, R.B.; Murphy, S.D. (1975) Short-term inhalation
toxicity of halogenated hydrocarbons. Effects on fasting rats. Arch.
Environ. Health 30:26-30.
Jakobson, I.; Wahlberg, J.E.; Holmberg, B.; Johansson, G. (1982) Uptake via
the blood and elimination of 10 organic solvents following epicutaneous
exposure to anesthetized guinea pigs. Toxicol. Appl. Pharmacol. 63:181-
187.
Kimura, E.T.; Ebert, D.M.; Dodge, B.w. (1971) Acute toxicity and limits of
solvent residue for sixteen organic solvents. Toxicol. Appl. Pharmacol.
19:699-704.
Klaassen, C.D.; Plaa, G.L. (1969) Comparison of the biochemical alterations
elicited in livers from rats treated with carbon tetrachloride,
chloroform, and methylchloroforms. Biochem. Pharmacol. 18:2019-2027.
4-68
-------
Klaassen, C.D.; Plaa, G.L. (1967) Relative effects of various chlorinated
hydrocarbons on liver and kidney function in mice. Toxicol. Appl.
Pharmacol. 10:119-131.
Klaassen, C.D.; Plaa, C.L. (1966) Relative effects of various chlorinated
hydrocarbons on liver and kidney function in mice. Toxicol. Appl.
Pharmacol. 9:139-151.
Kleihues, P.; Margison, G.P. (1974) Carcinogenicity of N-methyl-N-
nitrosurea. Possible role of excision repair of 06-methylguanine from
DMA. J. Natl. Cancer Inst. 53:1839-1841.
Kluwe, W.M.; McCormack, K.M.; Hook, J.B. (1978) Selective modification of
the renal and hepatic toxicities of chloroform by induction of drug
metabolizing enzyme systems in kidney and liver. J. Pharmacol. Exp.
Therap. 207:566-573.
Koch, R.R.; Glende, E.A.; Recknagel, R.O. (1974) Hepatotoxicity of
bromotrichloromethane-bond dissociation energy and lipoperoxidation.
Biochem. Pharmacol. 23:2907-2915.
Krishnan, N.; Stenger, R. (1966) Effects of starvation on the
hepatotoxicity of carbon tetrachloride. Am. J. Pathol. 49:239-246.
Lehmann, K.B.; Hasegawa, 0. (1910) Studies of the absorption of chlorinated
hydrocarbons in animals and humans. Arch. Hyg. 72:327-342 (Ger.).
Mansuy, D.; Beaune, P.; Cresteil, T.; Lange, M.; Lerous, J.P. (1977)
Evidence for phosgene formation during liver microsomal oxidation of
chloroform. Biochem. Biophys. Res. Commun. 79:513-517.
McConnell, G.; Ferguson, D.M.; Pearson, C.R. (1975) Chlorinated
hydrocarbons and the environment. Endeavor 34:13-18.
McLean, A.E. (1970) The effects of protein deficiency and microsomal enzyme
induction by DDT and phenobarbitone on the active toxicity of
chloroform. Br. J. Exp. Pathol. 51:317-321.
Miklashevskii, V.E.; Tugarinova, V.N.; Rakhmanina, N.L.; Yakovleva, G.P.
(1966) Toxicity of chloroform administered perorally. Hyg. Sanit.
31:320-322.
Morris, L.E.V. (1951) Chloroform in blood and respired atmosphere.
In: Waters, R.M., ed. Chloroform, a study after 100 years. Madison,
WI: University of Wisconsin Press; pp. 95-119.
Munson, E.S. (1973) Discussion: chloroform, halothane and regional
anesthesia. Anesth. Analg. 52:8-9.
Murray, F.J.; Schwetz, B.A.; McBride, J.G.; Staples, R.E. (1979) Toxicity
of inhaled chloroform in pregnant mice and their offspring. Toxicol.
Appl. Pharmacol. 50:515-522.
4-69
-------
Nakajima T.; Sato, A. (1979) Enhanced activity of liver drug-metabolizing
enzymes for aromatic and chlorinated hydrocarbons following food
deprivation. Toxicol. Appl. Pharmacol. 50:549-556.
National Institute for Occupational Safety and Health (NIOSH). (1977)
National Occupational Hazard Survey. Vol. III. Survey analyses and
supplemental tables. Cincinnati: USDHEW; pp. 2864-2866.
New, P.S.; Lubash, G.D.; Scherr, L.; Rubin, A.L. (1962) Acute renal failure
associated with carbon tetrachloride intoxication. J. Am. Med. Assoc.
181:903-906.
Paul, 6.G.; Rubinstein, D. (1963) Metabolism of carbon tetrachloride and
chloroform by the rat. J. Pharmacol. Exp. Ther. 141:141-148.
Pawlowsky, R.; Frosolono, M.F. (1977) Effect of phosgene on rat lungs after
single high exposure. Arch. Environ. Health 32:278-283.
Pegg, A.E.; Hui, G. (1978) Formation and subsequent removal of
06-methylguanine from deoxyribonucleic acid in rat liver and kidney
after small doses of dimethylnitrosamine. Biochem. J. 173:739-748.
Pereira, M.A.; Savage, R.E.; Guion, C.W.; Wernsing, P.A. (1984) Effect of
chloroform on hepatic and renal ONA synthesis and ornithine
decarboxylase activity in mice and rats. Toxicol. Lett. 21:357-364.
Piersol, G.M.; Tiemen, H.J.; Kau, L.S. (1933) Fatal poisoning following the
ingestion of chloroform. Med. Clin. North Am. 17:587-601.
Plaa, G.L.; Larson, R.E. (1965) Relative nephrotoxic properties of
chlorinated methane, ethane, and D24 ethylene derivatives in mice.
Toxicol. Appl. Pharmacol. 7:37-44.
Plaa, G.L.; Evans, E.A.; Hine, C.H. (1958) Relative hepatotoxicity of seven
halogenated hydrocarbons. J. Pharmacol. Exp. Ther. 123:224-229.
Pohl, L.R. (1980) Investigations of the mechanism of activation of
chloroform. Toxicol. Res. Proj. Dir. 5(No. 8.0463).
Pohl, L.R.; Bhooshan, B.; Whitaker, N.F.; Krishna, G. (1977) Phosgene:
a metabolite of chloroform. Biochem. Biophys. Res. Commun. 79:684-691.
Pohl, L.R.; Krishna, G. .(1978) Deuterium isotope effect in bioactivation
and hepatotoxicity of chloroform. Life Sci. 23:1067.
Pohl, L.R.; George, J.W.; Satch, H. (1984) Strain and sex differences in
chloroform-induced nephrotoxicity. Different rate of metabolism of
chloroform to phosgene by the mouse kidney. Drug Metab. Dispos.
12:304-308.
Poyer, J.R.; Floyd, R.A.; McCoy, P.B.; Janzen, E.G.; Davis, E.R. (1978)
Spin-trapping of the trichloromethyl radical produced during enzymic
NADPH oxidation in the presence of carbon tetrachloride or
bromotrichloroethane. Biochem. Biophys. Acta 539:402-409.
4-70
-------
Recknagel, R.O.; Ghoshal, A.K. (1966) Quantitative estimation of
peroxidative degeneration of rat liver microsomal and mitochondria!
lipids after carbon tetrachloride poisoning. Exp. Mol. Pathol. 5:413-
426.
Recknagel, R.O.; Glende, Jr., E.A. (1973) Carbon tetrachloride
hepatotoxicity: an example of lethal cleavage. CRC Crit. Rev. Toxicol.
2:263-297.
Reitz, R.H.; Quast, J.F.; Scott, W.T.; Watanabe, P.G.; Behring, O.J. (1980)
Pharmacokinetics and macromolecular effects of chloroform in rats and
mice: implications for carcinogenic risk estimation. Water Chlorinat.
Environ. Impact Health Eff. 3:983-993.
Reitz, R.H.; Gehring, P.J.; Park, C.N. (1978) Carcinogenic risk estimation
for chloroform: an alternative to EPA's procecures. Food Cosmet.
Toxicol. 16:511-514.
Reynolds, E.S. (1967) Liver parenchymal cell injury. IV. Pattern of
incorporation of carbon and chlorine from carbon tetrachloride into
chemical constituents of liver i_n vivo. J. Pharmacol. Exp. Ther.
155:177-126.
Reynolds, E.S.; Treinen, R.J.; Farrish, H.H.; Moslen, M.T. (1984a)
Relationships between the pharmacokinetics of carbon tetrachloride
conversion to carbon dioxide and chloroform and liver injury. Arch.
Toxicol. Suppl. 7:393-306.
Reynolds, E.S.; Treinen, R.J.; Farrish, H.H.; Moslen, M.T. (1984b)
Metabolism of 14C-carbon tetrachloride to exhaled, excreted and bound
metabolites. Dose-response, time-course, and pharmacokinetics.
Biochem. Pharmacol. 33:3363-3374.
Rubinstein, D.; Kanics, L. (1964) The conversion of carbon tetrachloride
and chloroform to carbon dioxide by rat liver homogenates. Can. J.
Biochem. 42:1577-1585.
Sato, A.; Nakajima, T. (1979) A structure-activity relationship of some
chlorinated hydrocarbons. Arch. Environ. Health 34:69-75.
Scholler, K.L. (1970) Modification of the effects of chloroform on the rat
liver. Br. J. Anaesth. 42:603-605.
Schroeder, H.G. (1965) Acute and delayed chloroform poisoning. A case
report. Br. J. Anaesth. 37:972-975.
Schwenkenbecher, F.A. (1904) The absorption capability of the skin. Archiv
fuer Anatomie und Physiologie, Physiologische Abteilung. 28:121-165.
Schwetz, B.A.; Leong, B.K.J.; Gehring, P.J. (1974) Embryo- and fetotoxicity
of inhaled chloroform in rats. Toxicol. Appl. Pharmacol. 28:442-451.
Seppelt, K. (1977) Trifluromethanol, CFaOH. Angew. Chem. Int. Ed. Engl.
16:322-323.
4-71
-------
Shah, H.; Hartman, S.P.; Weinhouse, S. (1979) Formation of carbonyl
chloride In carbon tetrachloride metabolism by rat liver in vitro.
Cancer Res. 39:3942-3947.
Shubik, P.; Ritchie, A.L. (1953) Sensitivity of male DBA mice to the
toxicity of chloroform as a laboratory hazard. Science 117:285.
Simpson, J.Y. (1847) On a new anaesthetic agent, more efficient than
sulfuric ether. Lancet 2:549-550.
Sipes, I.G.; Krishna, G.; Gillette, J.R. (1977) Bioactivation of carbon
tetrachloride, chloroform and bromotrichloromethane: role of cytochrome
P-450. Life Sci. 20:1541-1548.
Slater, T.F. (1972) Free radical mechanisms in tissue injury. London:
Pion. Ltd.
Smith, A.A.; Volpetto, P.O.; Gremling, Z.W.; DeVore, M.B.; Glassman, A.B.
(1973) Chloroform, halothane and regional anesthesia. A comparative
study. Anesth. Analg. 52:1-11.
Smith, J.H.; Hook, J.B. (1983) Mechanism of chloroform nephrotoxicity. II.
In vivo evidence for renal metabolism of chloroform in mice. Toxicol.
Appl. Pharmacol. 70:480-485.
Smith, J.H.; Hook, J.B. (1984) Mechanism of chloroform nephrotoxicity.
III. Renal and hepatic microsomal metabolism of chloroform in mice.
Toxicol. Appl. Pharmacol. 73:511-524.
Snow, J. (1858) On chloroform and other anesthetics. London:
John Churchill.
Steward, A.; Allot, P.R.; Cowles, A.L.; Mapleson, W.W. (1973) Solubility
coefficients for inhaled anaesthetics for water, oil, and biological
media. Br. J. Anaesth. 45:282-293.
Stewart, R.D. (1974) The use of breath analysis in clinical toxicology.
In: Essays in toxicology. F.R. Blood ed. New York: Academic Press;
Chapter 5.
Stewart, R.D.; Dodd, H.C.; Erly, D.S.; Holder, B.B. (1965) Diagnosis of
solvent poisoning. J. Am. Med. Assoc. 193:1097-1100.
Storms, W.W. (1973) Chloroform parties. J. Am. Med. Assoc. 225:160.
Symons, J.M.; Bellar, T.A.; Carswell, J.K.; Demarco, J.; Kropp, K.L.;
Robeck, G.G.; Seeger, D.R.; Slocum, C.J.; Smith, B.L.; Stevens, A.A.
(1975) National Organics Reconnaissance Survey for halogenated
organics. Am. Water Works Assoc. J. 67:634-647.
Taylor, D.C.; Brown, D.M.; Kuble, R.; Langley, P.F. (1974) Metabolism of
chloroform. II. A sex difference in the metabolism of I4c-chloroform
in mice. Xenobiotica 4:165-174.
4-72
-------
Thompson, D.J.; Warnet, S.D.; Robinson, V.V. (1974) Teratology studies on
orally administered chloroform in the rat and rabbit. Toxicol. Appl.
Pharmacol. 29:348-357.
Torkelson, T.R.; Oyen, F.; Rowe, V.K. (1976) The toxicity of chloroform as
determined by single and repeated exposure of laboratory animals. Am.
Ind. Hyg. Assoc. J. 37:697-704.
Tsurata, H. (1975) Percutaneous absorption of organic solvent.
1. Comparative study of the i_n vivo percutaneous absorption of
chlorinated solvents in mice. Ind. Health 13:227-236.
Tsurata, H. (1977) Percutaneous absorption of organic solvents.
2. A method for measuring the penetration rate of chlorinated solvents
through excised rat skin. Ind. Health 15:131-140.
Uehleke, H.; Werner, T. (1975) A comparative study on the irreversible
binding of labeled halothane, trichlorofluoromethane, chloroform, and
carbon terachloride to hepatic protein and lipids in vitro and i_n vivo.
Arch. Toxicol. 34:289-303.
Uehleke, H.; Hellmer, K.H.; Tarabelli, S. (1973) Binding of 14C-carbon
tetrachloride to microsomal proteins i_n vitro and formation of CHC13 by
reduced liver microsomes. Xenobiotica 3:1-11.
Uehleke, H.; Werner, T.; Greim, H.; Kramer, M. (1977) Metabolic activation
of halothane and tests i_n vitro for mutagenicity. Xenobiotica
7:393-400.
U.S. Occupational Safety and Health Administration (OSHA). (1978) Air
contaminants. U.S. Code Fed. Reg., Title 29, part 1910:1000;
pp. 612-614.
Van Dyke, R.A.; Chenoweth, M.B.; van Poznak, A. (1964) Metabolism of
volatile anesthetics. I. Conversion i_n vivo of several anesthetics to
14C02 and chloride. Biochem. Pharmacol. 13:1239-1247.
Villarruel, M.C.; Castro, J.A. (1975) Irreversible binding of carbon
tetrachloride to microsomal phospholipids. Free radical nature of the
reactive species and alterations in the physico-chemical properties of
the target fatty acids. Res. Commun. Chem. Pathol. Pharmacol.
10:105-115.
Villarruel, M.C.; Diaz-Gomez, M.I.; Castro, J.A. (1975) The nature of the
irreversible binding of carbon-tetrachloride to microsomal lipids.
Toxicol. Appl. Pharmacol. 33:106-116.
von Oettingen, W.F. (1964) The halogenated hydrocarbons of industrial and
toxicological importance. Amsterdam: Elsevier Publ. Co.; pp. 77-108.
Waters, R.M. (1951) Chloroform. A study after 100 years. Madison, WI:
University of Wisconsin Press.
4-73
-------
Withey, J.R.; Collins, B.T.; Collins, P.G. (1982) Effect of vehicle on the
pharmacokinetics and uptake of four halogenated hydrocarbons from the
gastro-intestinal tract of the rat. Pre-print paper submitted to
J. Appl. Toxicol., December 1982.
Withey, J.R.; Collins, B.T. (1980) Chlorinated aliphatic hydrocarbons used
in the foods industry: the comparative pharmacokinetics of methylene
chloride, 1,2-dichloroethane, chloroform and trichloroethylene after
i.v. administration in the rat. J. Environ. Pathol. Toxicol. 3:313-332.
Witte, C. (1874) XXXIV. Experiments concerning the effects of chloroform
on blood circulation. Dtsch. Z. Chir. 4:548-566.
Wolf, C.R.; Mansuy, D.; Nastainczyk, W.; Deutschmann, G.; Ullrich, V. (1977)
The reduction of polyhalogenated methanes by liver microsomal cytochrome
P-450. Mol. Pharmacol. 13:698-705.
Zeller, A. (1883) On the fate of iodoforms and chloroform in the organism.
Hoppe-Seyler's Z. Physiol. Chem. 8:78-79.
4-74
-------
5. TOXICITY
5.1. EFFECTS OF ACUTE EXPOSURE TO CHLOROFORM
In both humans and experimental animals, characteristic effects of acute
exposure to chloroform are depression of the central nervous system and
hepatic damage. Renal and cardiac effects also occur. The systemic toxic
effects of chloroform appear to be similar regardless of whether exposure or
administration occurred by inhalation, oral, or parenteral routes. The only
systemic effect documented for dermal administration, however, is renal
damage.
5.1.1. Humans
5.1.1.1. Acute Inhalation Exposure in Humans—Information on the effects of
acute inhalation exposure of chloroform on humans has been obtained primarily
during its use as an inhalation anesthetic. The relationship of the
concentration of chloroform in inspired air and blood to anesthesia is
described in Table 5-1 (Goodman and Gilman, 1980). Concentrations of
chloroform used for the induction of anesthesia were in the range of 2-3
volumes % (20,000-40,000 ppm), followed by lower maintenance levels (NIOSH,
1974; Adriani, 1970).
Chloroform inhalation has a depressive effect on the central nervous
system. Excitement due to release of inhibitions is followed by progressive
depression of the cortex, higher centers, medulla and spinal cord (Wood-Smith
and Stewart, 1964). Centers controlling temperature regulation, respiration,
vomiting, vasomotor, and vagal activity are all depressed (Adriani, 1970).
The cardiovascular system is also affected by anesthetic use of
chloroform. The myocardium is directly depressed in deeper planes of
anesthesia. A blood level sufficient to cause respiratory failure may also
5-1
-------
TABLE 5-1. RELATIONSHIP OF CHLOROFORM CONCENTRATION
IN INSPIRED AIR AND BLOOD TO ANESTHESIA
In inhaled air, In blood,
volumes % mg%
Not sufficient for anesthesia O.15 <2
Light anesthesia 0.15 to 0.20 2 to 10
(after induction)
Deep anesthesia 0.20 to 1.50 10 to 20
Respiratory failure 2.0 20 to 25
*SOURCE: Goodman and Gilman, 1980.
cause cardiac arrest. In addition, chloroform sensitizes the autononiic
tissues of the heart to epinephrine, causing arrhythmTas. It has been found
that under chloroform anesthesia regarded as normal ,Hhe heart is subject to
arrhythmias and extrasystoles (Kurtz et al., 1936; Orth et al., 1951). Orth
et al. (1951) found a high incidence of ventricular arrhythmias, 20 of 52
cases investigated, and four cases of temporary cardiac arrest. Blood
pressure is lowered by chloroform as a result of a 3-fold action: cardiac
slowing due to vagal stimulation, depression of the vasomotor center, and
dilation of splanchnic blood vessels (Krantz and Carr, 1965).
Respiratory effects of chloroform inhalation include increased rate and
depth of respiration during induction and in light anesthesia, and decreased
minute volume exchange in deeper planes of anesthesia. The Hering-Breuer
reflex remains active. Bronchial smooth muscle is relaxed and secretions are
increased. Laryngeal spasms are caused by high concentrations (Adriani,
1970).
In the gastrointestinal tract, chloroform markedly stimulates the flow
of saliva during induction and recovery, but salivation is inhibited in
5-2
-------
deeper planes of anesthesia (Goodman and Gilman, 1980). The pharyngeal or
gag reflex is depressed. Under anoxic conditions, pharyngeal muscle spasms
result in stertorous respiration and thick mucus is excreted (Adriani, 1970).
Stomach movements are decreased or abolished as tone is reduced. Gastric
secretory activity is inhibited or abolished. Post-anesthetic dilation of
the stomach occurs in nearly all cases. Nausea and vomiting often occur
during recovery from anesthesia. The mechanism is central rather than local,
but may be due in part to irritation of the stomach by swallowed vapor
(Goodman and Gilman, 1980). Intestinal tone, motility, and secretory
activity are inhibited or abolished (Adriani, 1970).
In the urinary tract, chloroform anesthesia results in a decrease in
urine flow, possibly due to the release of antidiuretic hormone and renal
vasoconstriction, leading to a decrease in renal blood flow and glomerular
filtration. Polyuria occurs after recovery (Goodman and Gilman, 1980).
Chloroform anesthesia may be followed by albuminuria and glycosuria.
Post-operative urine retention occurs frequently. Renal tubular necrosis has
been found in cases of severe poisoning (Wood-Smith and Stewart, 1964).
During obstetric use of chloroform, uterine contractions are only
slightly decreased by light anesthesia, but are markedly inhibited in deeper
planes. Chloroform rapidly crosses the placental barrier, and respiratory
depression in the infant is likely to occur (Wood-Smith and Stewart, 1964).
Chloroform anesthesia also has metabolic effects in humans. A rise in
blood glucose accompanies anesthesia. Levels may rise >2-fold and remain
elevated for several hours. The liver glycogen falls coincident with the
rise in blood sugar. This, in turn, is a result of the release of
epinephrine from the adrenal medulla during the period of excitation. There
is also a decrease in glucose utilization in the periphery (Goodman and
5-3
-------
Gilman, 1980; Krantz and Carr, 1965). Acidosis occurs, characterized by a
fall in plasma biocarbonate and phosphate.
Chloroform is acutely toxic to the liver, although in so-called delayed
chloroform poisoning, the full effects of damage done during and shortly
after administration are not seen for 24-48 hours. The glycogen content of
the liver is rapidly depleted; three-fourths in the first half-hour and less
rapidly thereafter. There is centrilobular and, in severe cases, mid-zonal
and massive necrosis. Cells which survive show fatty degeneration. Symptoms
include progressive weakness, prolonged vomiting, delirium, coma, and death.
They develop from the first to the third day after exposure. Jaundice,
increased serum bilirubin, bile in the urine, reduction in liver function,
increased nitrogen excretion, lowered blood prothrombin and fibrinogen, and
the appearance of leucine, tryosine, acetone, and diacetic acid in the urine
are some of the more prominent findings. The hemorrhagic tendency is due to
reduced prothrombin formation by the injured liver. Death usually occurs on
the fourth or fifth day. and autopsy reveals degeneration and necrosis of
liver tissue, most marked around the central veins (Goodman and Gilman, 1980;
Wood-Smith and Stewart, 1964).
Hematologic effects due to acute chloroform inhalation are seen during
anesthesia. Erythrocytes are increased in number as the spleen is
constricted and red blood cells are extruded into the circulation.
Leukocytes are increased in number during the post-anesthetic period,
reaching a maximum within 24 hours and returning to normal in 48 hours.
There is an increase in polymorphonuclear cells. Platelets remain unchanged.
One half-hour after exposure, there is a decrease in clotting time.
Prothrombin time is increased. Prothrombin synthesis is impaired by liver
toxicity as previously noted (Adriani, 1970).
5-4
-------
The effects of chloroform on the eye include dilation of the pupils,
with reduced reaction to light as well as reduced intraocular pressure (Sax,
1979; Winslow and Gerstner, 1978).
Signs of chloroform poisoning include a characteristic sweetish odor on
the breath, cold and clammy skin, and dilated pupils (Winslow and Gerstner,
1978). Nausea and vomiting commonly occur. Ketosis, due to incomplete
oxidation of fats, as well as a rise in blood sugar, accompanies chloroform
intoxication. Initial excitation alternating with apathy is followed by
prostration, unconsciousness, and possible death due to cardiac and central
nervous system depression (Winslow and Gerstner, 1978).
The-above discuss ion presents observations made on the effects of
chloroform inhalation during general anesthesia. Information on the effects
of experimental acute inhalation exposure of chloroform in humans is limited
to the work of Lehman and Hasegawa (1910) and Lehman and Schmidt-Kehl (1936)
as reviewed by NIOSH (1974). The duration of exposure was <30 minutes and
only the subjective responses of the subjects were measured. The
dose-response relationships as tabulated by NIOSH (1974) are presented in
Table 5-2.
5.1.1.2. Acute Oral Exposure in Humans—Case reports of suicides (Piersol et
al., 1933; Schroeder, 1965) and of recreational abuse (Storms, 1973) of
chloroform present some information on the effects of acute imbibition. A
fatal dose of ingested chloroform may be as little as one-third of an ounce
(10 ml) (Schroeder, 1965). The initial effect is usually unconsciousness and
possibly death (within 12 hours without treatment) due to respiratory or
cardiac arrest. If the patient survives, delayed effects are observed within
48 hours after regained consciousness. These symptoms include vomiting,
anorexia, jaundice, liver enlargement, albinuria, ketosis, ketonuria and
5-5
-------
TABLE 5-2. DOSE-RESPONSE RELATIONSHIPS
160 ppm (0.8 mg/L) for unspecified time - no odor
205 ppm (1.0 mg/L) for unspecified time - light transient odor
390 ppm (1.9 mg/L) for 30 minutes -light transient odor
920 ppm (4.5 mg/L) for 7 minutes - stronger, lasting odor; dizziness, vertigo
after 3 minutes
680 ppm (3.3 mg/L) to 1000 ppm (5.0 mg/L) for 30 minutes - moderately strong
odor; taste
1100 ppm (5.4 mg/L) for 5 minutes - still stronger, permanent odor;
dizziness, vertigo after 2 minutes
1400 ppm (6.6 mg/L) to 1800 ppm (8.57 mg/L) for 30 minutes - stronger odor,
tiredness, salivation, giddiness, vertigo, headache, taste
3000 ppm (14.46 mg/L) for 30 minutes - all above plus pounding heart, gagging
4300 ppm (20.8 mg/L) to 5000 ppm (25 mg/L) for 20 minutes - dizziness and
light intoxication
5100 ppm (25 mg/L) for 20 minutes - dizziness and light intoxication
7200 ppm (35.3 mg/L) for 15 minutes - dizziness and light intoxication as
above but more pronounced
SOURCE: Lehman and Hasegawa (1910) and Lehman and Schmidt-Kehl (1936).
glucosuria, hemorrhage due to lowered blood fibrinogen and prothrombin,
reduced serum bicarbonate, increased blood sugar, coma and possible death.
Upon autopsy, extensive hepatic centrilobular necrosis is evident.
5.1.1.3. Acute Dermal and Ocular Exposure in Humans—Chloroform is absorbed
through the intact skin (von Oettingen, 1964). Application of chloroform to
the skin is followed after 3 minutes by a pungent and burning pain reaching
its maximum after 5 minutes, associated with erythema, hyperemia, and finally
vesication (Oettel, 1936). Exposure of the eye to concentrated chloroform
vapors causes a stinging sensation. Splashing the substance into the eyes
5-6
-------
evokes burning, pain, and redness of the conjunctiva! tissue. The corneal
epithelium is sometimes impaired; however, regeneration starts rapidly and
leads to full recovery within 1-3 days (Winslow and Gerstner, 1978).
5.1.2 Experimental Animals
5.1.2.1. Acute Inhalation Exposure in Animals—Tolerance of animals to
chloroform has been summarized by Lehmann and Flury (1943) and by Sax (1979).
Similar central nervous system effects are seen in animals at approximately
the same magnitude of exposure that produced these effects in humans. In
mice, exposure to 2500 ppm for 2 hours produced no obvious effects, 3100 ppm
for 1 hour produced slight narcosis, while 4000 ppm induced deep narcosis
within one-half hour. Only slight symptoms are seen at 2000-6000 ppm for
longer exposures. Fatal exposures were 4100-8200 ppm for mice, 12,300 ppm
for rabbits, and 16,300-20,500 ppm for guinea pigs (duration of fatal
exposures not specified). In cats, exposure to 7200 ppm resulted in
disturbance of the equilibrium after 5 minutes, light narcosis after
60 minutes, and deep narcosis after 93 minutes of exposure. Exposure to
21,500 ppm produced disturbances in equilibrium after 5 minutes, light
narcosis after 10 minutes, and deep narcosis in cats after 13 minutes of
exposure.
Kyi in et al. (1963) described the effects of a single exposure of mice
to 100, 200, 400, or 800 ppm of chloroform for 4 hours. The mice exposed to
100 ppm did not develop demonstrable liver necroses, although moderate fatty
infiltration of the liver was noted. In mice exposed to 200 ppm, some
necrotic areas appeared in the liver and there was an increase in serum
ornithine-carbamyl transferase. Exposure to chloroform at 400 and 800 ppm
resulted in increased hepatic necrosis and serum enzyme activity.
5-7
-------
More recent data regarding toxic effects of acute inhalation exposure to
chloroform were presented by Wood et al. (1982), although the study was
designed primarily to investigate the role of hydrogen bonding in the
anesthetic mechanism. Groups of mice in a rotating cage were given a single
exposure of up to 3 hours of varying concentrations of chloroform or
deuterated chloroform, each concentration being held constant for about
20 minutes and then being raised until the mice had lost their righting
reflex. The concentration was then lowered to 1/2 the ED5Q (-1500 ppm)
where it remained until the mice had regained their righting reflex. The
duration of these manipulated exposures never exceeded 3 hours. Only 4 of
47 mice given chloroform gained the righting reflex; indeed, some mice died
or were comatose. Upon histological examination of*animals sacrificed
3-6 hours after exposure, mild hepatic centrilobular necrosis and very mild
renal tubular necrosis was observed. The an-imals receiving deuterated
chloroform survived for 24 hours, after which they were sacrificed. The
liver and kidney lesions in these mice were more severe, perhaps owing to the
longer survival time, which may have allowed these lesions to develop.
5.1.2.2. Acute Oral Exposure in Animals—Kimura et al. (1971) performed
acute oral toxicity studies in newborn (5-8 g), 14-day-old (16-50 g), young
adult (80-160 g), and older adult (360-470 g) rats. The chloroform was given
in undiluted form to unfasted rats. LD50 values in ml/kg (95% confidence
limits) were reported as follows: 14-day-old, 0.3 (0.2-0.5); young adult,
0.9 (0.8-1.1); and older adult, 0.8 (0.7-0.9). The young and older adult
rats were males; the other two groups contained rats of both sexes. When
compared with 15 other solvents included in this study, the LD50 values for
chloroform were the lowest in the two adult groups and next to lowest in the
14-day-old rats. Only a rough approximation of the LD5Q could be obtained
5-8
-------
for the newborn rats; volumes of 1 ml/kg body weight were generally fatal.
Lower volumes could not be measured with any degree of accuracy and were not
attempted.
Torkelson et al. (1976) reported an oral LD50 of 2.0 g/kg (1.05-3.80) in
male rats. Animals receiving as little as 0.25 g/kg showed adverse effects.
Other recent studies of acute oral toxicity have reported LD^Q values (with
95% confidence limits) of 1120 mg/kg (789-1590) in ICR male mice and
1400 mg/kg (1120-1680) in females (Bowman et al., 1978), and 908 mg/kg
(750-1082) and 1117 mg/kg (843-1514), respectively, in male and female rats
(Chu et al., 1980).
In a study that- compare4-the toxioities of halogenated hydrocarbons, a
single oral dose of 60 mg/kg chloroform to mice had no toxic effect (Hjelle
et al., 1982).
Hill (1978) performed experiments in mice designed to study variability
in susceptibility to chloroform toxicity from single oral doses based upon
genetic sex differences. For three strains of mice, the LDgQ values (ml/kg)
were as follows: DBA/2J, 0.08; B6D2F1/J, 0.20; and C57BL/6J, 0.33. The
animals more sensitive to chloroform-induced death were also found to be more
susceptible to renal toxicity. Males were found to be more sensitive to
renal damage and death than were females. This difference was related to
testosterone and it was further noted that C57BL/10J males are relatively
testosterone deficient in comparison to DBA males. The C57BL/6J strain used
in the Hill (1978) study is closely related to the C57BL/10J strain and may,
therefore, also be testosterone deficient.
In male B6C3F1 mice, severe diffuse renal necrosis occurred after a
single oral dose of 240 mg/kg and focal tubular regeneration occurred after a
single dose of 60 or 240 mg/kg. These effects were not seen after 15 mg/kg
5-9
-------
(Reitz et al., 1980). Liver damage (hepatocellular necrosis and swelling
with inflammatory cell infiltration) occurred only at the highest doses.
Chu et al. (1982a) studied the effects of acute oral exposure of
chloroform on Sprague-Dawley rats. Groups consisted of 10 male and 10 female
animals given a single oral dose of 0, 546, 765, 1071, 1500, or 2100 nig/kg of
chloroform in a volume of 5 ml/kg corn oil. Clinical signs of toxicity
included depression and coma, but the authors did not specify whether these
signs occurred at all dose levels. Treated rats surviving for 14 days
consumed less food and had depressed growth rates. Gross examination
revealed increased liver and kidney weights at 1071 mg/kg. Upon
comprehensive histological examination, only mild to moderate lesions, even
at high doses, were observed in these organs. No changes were noted in other
organs, including brain and heart. The hepatic and renal lesions were
characterized by hepatocyte variations and occasional vesiculation of biliary
epithelial nuclei in the liver, and by bilateral focal interstitial nephritis
and fibrosis in the kidney. Changes in hematological and biochemical
parameters were also observed in the 1071 and 1500 mg/kg treated groups.
Cholesterol levels increased while lactate dehydrogenase activity and liver
protein levels decreased. In female rats, the activity of microsomal aniline
hydroxylase was induced by chloroform exposure. The numbers of lymphocytes
were reduced in both males and females, as were hemoglobin and hematocrit
values. Upon longer exposures of 5, 50, or 500 ppm chloroform in drinking
water for 28 days, the only toxic effect observed was a decreased number of
neutrophils in the highest dose-exposed rats. Examinations were performed as
in the single-dose experiment.
Yannai (1983) has reported that chloroform has profound effects on the
adrenal-pituitary axis, and that these effects may be a parameter in
5-10
-------
chloroform toxicities. After administration of single oral doses of
chloroform (1.5 g/kg) to rats, adrenal hypertrophy developed and persisted
for 12 days. Adrenal cholesterol content decreased within 3 hr (an
indication of increased corticosteroid synthesis) and was accompanied by an
increase of plasma corticosterone within 15 min, remaining elevated for 4 hr.
The blood clearance rate of corticosterone was not changed. Since
cortisteroids play an important role in the stress response, these result
suggest that toxic agents such a chloroform may activate the adrenal-
pituitary axis which may in turn modify the toxic response.
5.1.2.3 Acute Dermal and Ocular Exposure in Anima1s--Torke1son et al. (1976)
found that chloroform, when applied to the skin of rabbits, produced slight
to moderate irritation and delayed healing of abraded skin. When applied to
the uncovered ear of rabbits, slight hyperemia and exfoliation occurred after
one to four treatments. No greater injury was noted after 10 applications.
One to two 24-hour applications, on a cotton pad bandaged on the shaven belly
of the same rabbits, produced a slight hyperemia with moderate necrosis and a
resulting eschar formation. Healing appeared to be delayed on the site as
well as on abraded areas which were also covered for 24 hours with a cotton
pad soaked in chloroform.
Single application of either 1.0, 2.0, or 3.98 g/kg for 24 hours under
an impermeable plastic cuff held tightly around the clipped bellies of each
of two rabbits did not result in any deaths. However, extensive necrosis of
the skin and considerable weight loss occurred at all levels. All animals
were sacrificed for study 2 weeks after exposure. All treated rabbits
exhibited degenerative changes in the kidney tubules graded in intensity with
dosage levels. The livers were not grossly affected.
5-11
-------
In the same study (Torkelson et al., 1976), liquid chloroform, dropped
into the eyes of three rabbits, caused slight irritation of the conjunctiva
which was barely detectable 1 week after treatment. In addition, slight but
definite corneal injury occurred, as evidenced by staining with fluorescein.
A purulent exudate occurred for >2 days after treatment. Although started
30 seconds after instilling the chloroform, thorough washing of one eye of
each rabbit with a stream of running water for 2 minutes did not
significantly alter the response in the washed eyes from that of the unwashed
eyes.
5.1.2.4 . Intraperitoneal and Subcutaneous Administration in Animals—The
toxicity of chloroform—in mice after subcutaneous .administration (Kutob and
Plaa, 1962b) and intraperitoneal administration (Klaassen and Plaa, 1966) has
been compared with that of other halogenated hydrocarbons (Pohl, 1979). In
these studies, the LD^Q values for carbon tetrachloride, chloroform, and
dichloromethane were, respectively, 200, 27.5, and 76 mmol/kg after
subcutaneous administration and 20, 14, and 23 mmol/kg when given
intraperitoneally. When the relative hepatotoxicity of these compounds was
compared, a subcutaneous dose of 0.5 mmol/kg of carbon tetrachloride produced
approximately the same degree of liver damage as "6.2 mmol/kg of chloroform.
After intraperitoneal administration, these values dropped to 0.01 mmol/kg
for carbon tetrachloride and 2.3 mmol/kg for chloroform. Dichloromethane did
not cause significant hist'ological changes in the liver by either route of
administration. At doses that produced liver toxicity, chloroform caused
kidney lesions which ranged from the presence of hyaline droplets, nuclear
pycnosis, hydropic degeneration, and increased eosinophilia, to necrosis with
karyolysis and loss of epithelium of the convoluted tubules.
5-12
-------
Ilett et al. (1973) found that intraperitoneal administration of
chloroform caused centrilobular hepatic necrosis in mice of both sexes,
whereas renal necrosis was observed only in male mice.
5.2. EFFECTS OF CHRONIC EXPOSURE TO CHLOROFORM
A characteristic effect of chronic exposure to chloroform is hepatic
damage; this effect has been documented primarily in studies with
experimental animals. As was the case for acute exposure to chloroform,
hepatic damage in chronic studies results from either inhalation or oral
administration of this chemical. Effects on the kidneys and thyroids have
also been observed in some experiments. This section will discuss both
subchronic (=90 days) and chronic exposure studies, because many of the
subchronic studies were preliminary range-finding tests for the chronic
studies.
5.2.1. Humans_
5.2.1.1. Chronic Inhalation Exposure in Humans--0n1y two chronic inhalation
studies that reported measurements of exposure concentrations, as well as
effects on human health, were found. Neither study (Challen et al., 1958;
Bomski et al., 1967) was performed adequately.
Challen et al. (1958) investigated complaints of workers (mainly women)
in a plant manufacturing lozenges that contained chloroform as a principle
ingredient. Before exhaust ventilation was installed, 9 of the 10 exposed
workers had complained of symptoms of tiredness, dull-wittedness, depression,
gastrointestinal distress, and frequent and scalding urination.
Breathing-zone monitoring during simulation of "pre-ventilation" working
conditions suggested that the employees had been exposed to =77-237 ppm.
Discussions with management revealed that some of these workers had
occasionally been observed to behave in a silly manner or to stagger about
5-13
-------
during the workday. Another group of workers (N = 10) had been exposed
primarily to concentrations of 22-71 ppm. Eight of these workers complained
of less severe symptoms. Apparently, both groups of workers had been exposed
to occasional peak concentrations of =1163 ppm lasting 1.5-2 minutes. At
least four workers in each group worked half-time. None of the five controls
reported symptoms similar to those reported by the exposed workers. Eight of
the higher exposure employees (77-237 ppm for 3-10 years, followed by
-2 years without exposure), nine of the lower exposure employees (22-77 ppm
for 10-24 months), and five unexposed employees submitted to physical
examinations, including liver function tests (thymol turbidity, serum
biliruhin, and urine urobilinogen tests). These examinations and tests
revealed no evidence of any organic lesion, including liver damage,
attributable to exposure to chloroform.
In humans, hepatic damage is the most common toxic effect of acute
exposure to chloroform, as noted previously. According to Pohl (1979), only
one report of liver abnormalities in humans after chronic exposure to
chloroform has been found in the literature and no additional reports were
found in the more recent literature. In this study (Bomski et al., 1967), 17
cases of hepatomegaly were found in a group of 68 industrial workers who were
exposed to chloroform in concentrations ranging from 2-205 ppm for 1-4 years.
These apparently were unspecified area, as opposed to breathing zone,
concentrations. Three of the 17 workers with hepatomegaly were judged to
have toxic hepatitis on the basis of elevated serum enzymes. The frequency
of viral hepatitis among the 68 chloroform-exposed workers was higher (4.4%
versus 0.38%) than the frequency among a group of inhabitants of the city,
>18 years of age. This phenomenon also occurred in the 2 previous years.
Ten cases of splenomegaly were also diagnosed among the 68 workers. There
5-14
-------
appears to be no comparison with incidences of these conditions in nonexposed
workers.
Recently Phoon et al. (1985) have reported on the occurrence of toxic
jaundice from chemical exposure of 31 factory workers to chronic, relatively
low levels of chloroform. The clinical symptomatology was similar to viral
hepatitis, without the high temperature or fever associated with that
disease. These workers had originally been diagnosed as having viral
hepatitis, but further investigation showed that each worker had been exposed
to chloroform. Two outbreaks were described. Between 1973 and 1974, over a
period of 9 months, 13 workers were afflicted in a large factory in Singapore
manufacturing small household appliances for which a degreaser (chloroform)
was used. The workers (mostly young females) complained of anorexia, nausea,
vomiting and jaundice. The level of chloroform in the air was estimated at
higher than 400 ppm, and blood samples showed blood chloroform levels of 0.10
to 0.20 mg%. A second outbreak of 11 cases occurred within a four-week
period in 1980 and a further 7 between 1980 and 1981, all from the same
section of a second factory making radios and clock radios. Preliminarily
diagnosed as acute infectious hepatitis, they were all negative for
hepatitis B surface antigen. These affected workers (again mostly young
females) worked in the casing department where components were assembled,
encased and packed for export, and where chloroform was used as a plastic
adhesive. Air analysis showed levels of 14.4 to 50.4 ppm on different days.
Blood samples revealed chloroform in some but not all workers. The duration
of exposure to chloroform, before onset of jaundice, averaged 1 to 4 months.
5.2.1.2. Chronic Oral Exposure in Humans--There are relatively few reports
of toxic effects following chronic ingestion of chloroform (Pohl, 1979), and
more recent reports were not located in the literature. In one case, it was
5-15
-------
estimated that a male patient ingested 1.6-2.6 g of chloroform in a cough
medicine daily for =10 years (Wallace, 1950). Blood and urine analyses, as
well as liver function tests, indicated the individual suffered from
hepatitis and nephrosis. Another report described three patients addicted to
chlorodyne, a tincture containing chloroform and morphine (Conlon, 1963).
Liver biopsy showed severe cellular damage in one of these individuals who
had ingested 21 ml of chloroform daily for an undetermined period of time.
All three displayed evidence of serious mental and physical deterioration,
including peripheral neuropathy. It is not possible to determine if the
adverse effects were due to chloroform, morphine, or ethanol.
More recently, the safety of a dentifrice containing 3.4% chloroform and
a mouthwash containing 0.43% was assessed in studies lasting >1 year (DeSalva
et al., 1975). The subjects using the dentifrice were exposed to =70 mg
(0.047 ml) of chloroform each day, whereas the groups using the mouthwash
were exposed to =178 mg (0.12 ml). The results of liver function tests and
blood urea nitrogen determinations showed no statistical differences between
control and experimental subjects.
Epidemiologic studies of humans exposed to chloroform in their drinking
water have focused on carcinogenic end points, and, hence, are discussed in
the chapter on carcinogenicity.
5.2.2. Experimental Animals
5.2.2.1. Chronic Inhalation Exposure in Animals—Experiments with several
species of animals (Torkelson et al., 1976) give some information regarding
potential effects of long-term inhalation exposure to chloroform. The
animals were exposed to chloroform 5 days/week for 6 months. Exposure to
25 ppm of chloroform for up to 4 hours/day had no adverse effects in male
rats as judged by organ and body weights, and by gross and histological
5-16
-------
examination of livers and kidneys. Exposure to 25 ppm for 7 hours/day,
however, produced histopathological changes in the livers and kidneys of male
but not female rats. These changes were characterized as lobular granular
degeneration and focal necrosis throughout the liver and cloudy swelling of
the kidneys. The hepatic and renal effects appeared to be reversible because
rats exposed according to the same protocol, but given a 6 week recovery
period following exposure, appeared normal by the criteria tested.
Increasingly pronounced changes were observed in the livers and kidneys of
both sexes of rats exposed to 50 or 85 ppm for 7 hours/day. Hematologic
indices, clinical chemistry, and urinalysis values, tested at higher levels
of exposure, were within normal limits. Each exposure group and control
group had 10-12 rats/sex except for the 25 ppm, 4 hour/day group, which had
10 male rats and no females.
Similar experiments with guinea pigs (N = 8-12/sex/group) and rabbits
(N = 2-3/sex/group) gave somewhat inconsistent results. Histopathological
changes were observed in livers and kidneys of both species at 25 ppm but not
at 50 ppm in either species, nor even at 85 ppm in guinea pigs. The results
of these studies are summarized in Table 5-3.
Other reports of effects of chronic inhalation exposure to chloroform in
experimental animals were not found in the more recent literature.
5.2.2.2. Chronic Oral Exposure in Animals—The data from several studies on
the effects of chronic and subchronic oral exposure to chloroform are
summarized in Table 5-4. Low levels of exposure (15-64 mg/kg/day,
6 days/week) have been reported to increase survival in mice and rats (Roe
et al., 1979; Palmer et al., 1979). and to be associated with (1) possible
transient CNS depression and mild hepatic changes in mice and rats (Jorgenson
and Rushbrook, 1980; Palmer et al., 1979); (2) hepatic damage in dogs
5-17
-------
TABLE 5-3. EFFECTS OF INHALATION EXPOSURE OF ANIMALS TO CHLOROFORM, 5 DAYS/WEEK FOR 6 MONTHSa
Exposure
Species Sex ppm
hours/day
Number 1n Group
Started Survived
Effects
rats M 85
10
en
i
CD
F 85
H 50
F 50
M 25
F 25
10 10
10 9
10 10
12 9
12 12
Excess mortality attributed to pneumonia on basis of gross and microscopic appearance
of lungs; slight depression of final body weight; Increase (p<0.05) 1n relative but
not absolute weights of liver, kidneys, and testes; no effect on spleen weight;
hlstologlcal findings Included marked centrllobular granular degeneration of the
livers and cloudy swelling of the kidneys but no hlstopathologlcal changes 1n testes;
hematologlc values (Including differential count), urlnalysls values and SGPT, SUN,
and SAP values all "within normal limits"
No evidence of pneumonia; final body weights and weights of liver and spleen
unaffected, relative and absolute kidney weights Increased; hlstologlcal findings
Included marked centrllobular granular degeneration of the livers and cloudy
swelling of the kidneys; hematologlc, urlnalysls, and SGPT, SUN, and SAT values all
"within normal limits"
Depression of final body weights (p<0.05); Increases 1n relative (p<0.05) but not
absolute weights of kidneys, spleen, and testes; hlstopathologlcal changes 1n livers
and kidneys similar to those seen at 85 ppm; hematologlc values, urlnalysls values,
and SGPT, SUN, and SAP values all "within normal limits"
Final body weights and weights of liver and spleen unaffected; Increase 1n relative
kidney weight (p<0.05); hlstopathologlcal changes 1n livers and kidneys similar to
those seen at 85 ppm but somewhat less marked; hematologlc values, urlnalysls values,
and SGPT, SUN, and SAP values all "within normal limits"
No effect on final body weights or weight of liver, spleen, or testes; Increased
relative kidney weight (p<0.05); lobular granular degeneration with focal areas of
necrosis throughout the liver; cloudy swelling of renal tubular epithelium
No statistically significant effect on body weight; Increased relative but not
absolute kidney and spleen weights; no hlstopathologlc changes 1n kidneys and spleen;
microscopic appearance of livers not specified
-------
TABLE 5-3 (continued)
en
i
Species
rats
Exposure
Sex ppm hours/day
M.F 25 7
plus
0 ppm
for 6 weeks
(recovery
period)
Number In Group
Started Survived
12/sex 8 M,
10 F
Effects
"Normal" by the criteria tested at this dosage level
(see 25 ppm above)
25
1.2.
or 4
guinea M,F 85,50, 7
pigs or 25
rabbits M.F 85, 50, 7
or 25
dogs M.F 25
10, 10, 7, 8, 4, No evidence «f adverse effects by the criteria tested (I.e., final body weight;
10 respec- weights of livers, kidneys, spleen, testes; and probably gross and microscopic
lively appearance of at least the liver and kidneys)
8 to 12 50 to 92% No adverse effects at 50 or 85 ppm other than marked pneutnonltis 1n F at 85 ppm; some
/sex/ (mortality hlstopathologlcal changes 1n livers of both sexes and kidneys of H at 25 ppm
exposure not related (criteria tested were body weights, organ weights, and gross and microscopic
level to exposure) appearance of organs)
2 to 3 0 to 1 No adverse effects at 50 ppm; some hlstopathologlc changes 1n kidneys and liver
/sex/ death/group, and pneumonltls 1n lungs at 25 and 85 ppm. Hematologlc and clinical chemistry
exposure not related values within normal limits at 85 ppm (criteria tested were same as for rats)
level to exposure
I/sex I/sex No adverse effects In M; marked cloudy swelling of renal tubular epithelium and
Increase 1n capsular space 1n glomerull of kidneys 1n F (criteria tested were same as
for rats and Included clinical chemistry and hematologlcal studies)
^Controls for each species and sex Included at least one unexposed and one air-exposed group, each comparable 1n number of animals to the exposed
groups. In statistical comparisons of organ and body weights, values for the control group (unexposed or air-exposed) closer 1n body weight to the
test group were used. Mortality In control groups was similar to mortality In treated groups, with the exception of excess mortality 1n male rats
exposed to 85 ppm for 7 hours/day or to 25 ppm for 4 hours/day. No explanation was given by Torkelson et al. (1976) for the high mortality In the
4 hours/day group. Strains of animals and age or weight at the start of the experiment were not specified. Purity of the chloroform used was 99.3%
(0.4% ethyl alcohol and <0.3% of an unknown).
M = male; F = female; SGPT = serum glutamlc pyruvlc transamlnase; SUN = serum urea nitrogen; SAP = serum alkaline phosphatase.
SOURCE: Torkelson et al., 1976.
-------
TABLE 5-4. EFFECTS OF SUBCHRONIC OR CHRONIC ORAL ADMINISTRATION OF CHLOROFORM TO ANIMALS
Species, strain
age/weight at
start
Sex
No. at
start
Vehicle
Dosage
Duration
Response
Reference
Rats, Sprague-Dawley
weanling, 101 g M,
94 g F
M.F
20/sex/dose
level
drinking
water
en
i
PO
o
0, 5, 50, 500,
or 2500 ppm 1n
drinking water
ad lib. corre-
sponding to
Intakesa of 0,
0.11-0.71, 1.2-
1.5. 8.9-14, or
29-55 mg/rat/day
The highest dosea
corresponds to
= 291 mg/kg/day F,
310 mg/kg/day H
90 days, after
which 10 rats/
group were killed
and 10 rats/
group observed
for additional
90 days
Increased mortality, Chu et al., 1982b
decreased growth rate,
and decreased food
Intake at highest dose;
Increased frequency of
mild to moderate
liver and thyroid
lesions at highest
dose. Including
Increases 1n cytoplasmlc
homogeneity, hepato-
cyte density, and
cytoplasmlc volume,
vacuolizatlon due to
fatty Infiltration, some
veslculatlon of biliary
epithelial nuclei, and
hyperplasla 1n livers;
and reduced follicle
and colloid density,
Increased epithelial
height, some focal
collapse of follicles 1n
thyroid; no hlstopatho-
loglcal effects 1n
kidney, brain, and heart;
after the 90-day recovery,
the lesions were very
mild and similar to
those seen 1n controls
-------
TABLE 5-4. (continued)
en
i
ro
Species, strain
age/weight at
start Sex
Rats, Osborne-Mendel M
6 weeks/190 g
No. at
start Vehicle Dosage Duration
30/group drinking 0, 200, 400, 600 10 rats/group
except water 900, or 1800 ppm killed at 30,
40 ad lib. 1n drinking water 60, and 90 days
controls ad lib. cor- of exposure
responding to
Intakes^ of 0,
20, 38, 57, 81, or
160 mg/kg/day,
plus 0 ppm group
matched with
1800 ppm group
for water
consumption
Response Reference
Dose-related signs Jorgenson and
of depression during Rushbrook, 1980
1st week only; dose-
related reduction In
water consumption;
decreased weight gain
1n 160 mg/kg group;
increased Incidence of
"hepatosls" 1n livers
of treated rats at 30
and 60 days but not
90 days (not dose-
related) ; no other
treatment-related
effects on serum
clinical chemistry
values or urlnalysls
values, or gross and
microscopic appearance
of tissues, Including
kidney.
-------
TABLE 5-4. (continued)
Species, strain
age/weight at
start
Sex
No. at
start
Vehicle
Dosage
Duration
Response
Reference
rats, Osborne-Mendel
52 days/240 g H
175 g F
M.F
50/sex/dose
level; 20/
sex-matched
controls;
= 99/sex
colony
controls
corn oil, M: 0, 90, or 180 78 weeks treat-
gavage mg/kg/day; ment plus
F: 0, 100, or 200 33 weeks
mg/kg/day (TWA); observation
5 days/week
in
i
ro
IN)
Treated animals had NCI, 1976
dose-related decrease
In survival and weight
gain, slight decrease
In food consumption,
Increased severity and
Incidence of pulmonary
lesions characteristic
of pneumonia; necrosis of
hepatic parenchyma NS 1n
controls, 3/50 low dose M,
4/50 high dose M, 3/49 low
dose F, 11/48 high dose F;
hyperplasla of urinary
bladder epithelium 1/18
control M, 7/45 low dose
M, 1/45 high dose M, NS
for control F, 6/43 low
dose F, 2/40 high dose
F; Increased splenic
hematopolesls 1n 1/18
control M, 3/45 low dose
M, 6/45 high dose M; both
control and treated
animals had chronic
nephritis; low but
statistically signi-
ficant Increased
Incidence of renal
epithelial tumors
1n treated M (see
Carc1nogen1c1ty
section)
-------
TABLE 5-4. (continued)
Species, strain
age/weight at
start
Sex
No. at
start
Vehicle
Dosage
Duration
Response
Reference
rats, Sprague-Dawley
NS
M.F
10/sex/dose
level
toothpaste,
gavage
0, 15, 30, 150,
or 410 mg/kg/day
6 days/week
13 weeks
en
i
ro
rats, Sprague-Oawley
SPF. 180 to 240 g M
130 to 175 g F
M.F
50/sex/group
toothpaste,
gavage
0 or 60 mg/kg/day
6 days/week
80 weeks
exposure
plus 15 weeks
observation
At 410 mg/kg/day,
Increased liver weight
with fatty change
and necrosis,
gonadal atrophy,
increased eellular
proliferation 1n
bone marrow; at
150 mg/kg/day,
changes less pro-
nounced but effect
(NS) on relative
liver and kidney
weights; pre-
sumably no effects at
lower dosage levels
Survival of treated
animals slightly
better than that of
controls (32* treated M,
22% control M, 26%
treated F, 14% control
females survived to 95
weeks); body weights of
treated rats slightly and
progressively depressed;
Intercurrent respiratory
and renal disease 1n all
groups; minor hlstologl-
cal changes 1n livers but
no evidence of "treatment-
related toxic effect" 1n
Palmer et al.,
1979
Palmer et al.,
1979
-------
TABLE 5-4. (continued)
Species, strain
age/weight at
start
Sex
No. at
start
Vehicle
Dosage
Duration
Response
Reference
en
i
ro
mice, B6C3F1,
6 weeks/19 g
30/group
except 40
ad lib.
controls
drinking
water
0, 200, 400, 600,
900, 1800, or
2700 ppm 1n
drinking water
ad lib. cor-
responding to
Intakesc
of =0, 32, 64,
97, 145, or 290
mg/kg/day; addi-
tional 0 ppm
group matched
with 2700 ppm
group for water
consumption
10 mice/group
killed at 30, 60,
and 90 days of
exposure
livers; decrease
(p<0.01) In relative
liver weight 1n
treated females; no
gross or hlstologlc
treatment-related
changes 1n brain;
possible effect (NS) on
Incidence of severe glomerulo-
nephrltls; decrease
1n plasma chollnes-
terase 1n treated females
Dose-related signs
of depression during
first week only; marked
reduction In
water consumption
in higher-dose
groups during first
2 weeks; body
weight losses (p<0.05)
In 97, 145, and 290 mg/kg
groups and 1n matched
controls during first week;
mild hepatic centrllobular
fatty change In 64, 97,
145, and 290 mg/kg groups
at 30 days, but only in 2
highest dosage groups at
60 and 90 days; Increase 1n
liver fat/11verwe1ght
Jorgenson and
Rushbrook, 1980
-------
TABLE 5-4. (continued)
Species, strain
age/weight at
start
Sex
No. at
start
Vehicle
Dosage
Duration
Response
Reference
mice, B6C3F1
35 days/18g M,
17 g F
M.F
en
i
ro
en
50/sex/dose
level; ZO/sex
matched
controls
corn oil,
gavage
mice, Schofleld
M.F
10/sex
per dosage
toothpaste,
gavage
M: 0. 138, or 227
mg/kg/day;
F: 0, 238, or 447
mg/kg/day;
5 days/week
78 weeks treat-
ment plus 14 to
15 weeks
observation
0. 60, 150,
or 425 mg/kg/
day; 6 days/
week
6 weeks
(p<0.05) for
290 mg/kg group at
all 3 periods; no other
treatment-related changes
In serum enzyme levels,
urlnalysls values, or gross
or microscopic appearance
of tissues Including kidney
Survival decreased NCI, 1976
1n high dose F,
unaffected 1n
other treated groups;
high Incidences of
hepatocellular car-
cinoma 1n treated mice
(see Cardnogenlclty sec-
tion); renal Inflamma-
tion 1n 10/18 control M,
2/50 low dose M,
1/50 high dose M
At 425 mg/kg, 100% Roe et al., 1979
mortality; at 150 mg/kg,
8/10 M died and
weight gain of F
markedly retarded; at
60 mg/kg, weight gain
of both sexes moder-
ately retarded; no
other observations
mentioned
-------
TABLE 5-4. (continued)
in
i
ro
Species, strain
age/weight at
start
mice, ICI (expt. 1)
ICI (SPF) (expt. 2)
ICI, CBA, C57BL,
CF/1 (expt. 3)
slO weeks old
No. at
Sex start Vehicle Dosage Duration
expt. 1: treated: toothpaste 1n 0, 17 (expt.l), 80 weeks treat-
M,F 52/sex/dose all 3 expt. or 60 mg/kg/day, ment; 16 to 24
expt. 2 level; con- for all 6 days/week weeks observa-
and 3: trol: 52 to strains and tlon according
M 206/sex/ sexes plus to numbers of
treated strain/ aracMs oil survivors
and con- vehicle plus In expt. 3
trol, untreated for ICI,
plus gavage
some F
control
Response Reference
Survival generally Roe et al., 1979
better 1n 60 mg/kg
groups than 1n con-
trols except for
CF/1 animals or when
chloroform given 1n
arachls oil; slight
retardation of weight
gain 1n 60 mg/kg
groups; no effect
on hematologlc values
(tested 1n expt. 2
only); no treatment-
related adverse effect
on liver or other
tissues except 1n
kidneys as follows:
60 mn/kg 1n tooth-
paste - Increased Inci-
dence of moderate
to severe renal
changes (p<0.001) 1n
CBA and CF/1 M,
60 mg/kg 1n oil-
Increased Incidence of
moderate to severe kid-
ney disease (p<0.05) In
ICI M; Increased Incidence
of benign and malignant
kidney tumors 1n ICI M
treated with 60 mg/kg
1n toothpaste or oil (see
Cardnogenldty section)
-------
TABLE 5-4. (continued)
Species, strain
age/weight at
start
Sex
No. at
start
Vehicle
Dosage
Duration
Response
Reference
dogs, beagle
18 to 24 weeks
7 to 8 g
M.F I/sex/dose
level for 90
and 120 mg/
kg; 2/sex/
dose level
for lower
dosages
toothpaste
1n gelatin
capsule.
oral ly
30, 45, 60, 90,
or 120 mg/kg/day,
7 days/week
13 weeks for
30 and 45 mg/kg,
18 weeks for
60 mg/kg.
12 weeks
for 90 and
120 mg/kg
No deaths; occasional Heywood et al..
vomiting; marked
weight loss 1n all
dogs and poor
general condition In
some at 60 mg/kg or
higher; apparent
01
I
ro
—i
suppression of appe-
tite Initially at all
dosages and through-
out at 60 mg/kg and
higher; jaundice and
Increased SAP, SCOT,
SGPT, blUrubln, and
ICD values 1n male at
120 mg/kg; Increased
SGPT values 1n 4/4 and
Increased SAP and SCOT
values 1n 2/4 at
60 mg/kg; hepatocyte
enlargement and vacuo-
latlon with fat depo-
sition at 60 mg/kg and
higher; discoloration of
liver. Increased liver
weight, and slight fat
deposition 1n hepato-
cytes at 45 mg/kg; no
effect on any of these
clinical chemistry or
hlstologlcal parameters
at 30 mg/kg
-------
TABLE 5-4. (continued)
Species, strain
age/weight at
start
Sex
No. at
start
Vehicle
Dosage
Duration
Response
Reference
dogs, beagle,
8 to 24 weeks,
7 to 8 kg
M.F
8/sex/dose
level; 16/
sex vehicle
controls;
plus other
controls
toothpaste
1n gelatin
capsule,
orally
0, 15, or 30
mg/kg/day;
6 days/week
7.5 years
treatment plus
20 to 24 weeks
observation
tn
l
ro
CD
No effect on survival,
growth, organ
weights, hematologlc
or urlnalysls values
(checked at Intervals
throughout); moderate
dose-related elevation
of SGPT reaching peak
1n sixth year of study,
reverting to normal
levels after treatment
discontinued; other
serum enzyme Indicators
of hepatic damage
(checked during the
latter portion of the
study) followed
pattern similar to SGPT,
but BSP retention
and ICD values were
unaffected; aggregation
of vacuolated hlstlo-
cytes ("fatty cysts")
1n livers of all groups
but cysts were larger
and more numerous 1n
treated dogs and
persisted after treat-
ment ended; fat depo-
sition affected more
renal glomerull In
30 mg/kg group than
In other groups
Heywood et al., 1979
-------
i
ro
10
TABLE 5-4. (continued)
Species, strain
age/weight at
start
No. at
Sex start Vehicle Dosage Duration Response Reference
aCalculated by Chu et at. (198?b) by multiplying the fluid Intake volume by the concentration of chloroform.
t>Calculated by Jorgenson and Rushbrook (1980) from measured average body weights and water consumption.
cCalculated from Jorgenson and Rushbrook s statement that the mice had actual Intake levels of from 148 to 175% of the
Intended levels of 20, 40, 60, 90, 180, and 270 mg/kg/day.
dGrowth rate data were given only for the highest dose and mg/kg/day were calculated from this Information by taking average weights over the 90-day
period of exposure.
SPF = specific pathogen-free; SGPT = serum glutamlc-pyruvlc transamlnase; SAP = serum alkaline phosphatase; BSP = bromsulphthaleln;
ICO = 1soc1tr1c dehydrogenase (serum)
Purity of the chloroform samples used 1n all these studies was generally high and 1s discussed 1n the section on Carc1nogen1c1ty.
M = male; F = female; TWA = time-weighted average dose for days on which chemical was administered; NS = Not specified.
-------
(Heywood et al., 1979), and (3) renal damage in male mice of some sensitive
strains (Roe et al., 1979), and in dogs (Heywood et al., 1979). In addition
to hepatic damage, ingestion of =300 mg/kg/day of chloroform produced
thyroid lesions in rats (Chu et al., 1982b). In one study, a decreased
incidence of renal inflammation occurred in male mice treated with chloroform
at 138 or 227 mg/kg/day, 5 days/week (NCI, 1976). A similar effect may have
occurred in rats (Palmer et al., 1979), but the authors did not specify
whether the effect of chloroform was to increase or decrease the incidence of
intercurrent renal disease. The NCI (1976) report stated that hepatic
necrosis, hyperplasia of the urinary bladder epithelium, and increased
splenic hematopoiesis in rats may have been related to chloroform treatment,
but incidences in controls for some of these effects were not reported and
the data, shown in Table 5-4, are difficult to interpret. Many of the
studies summarized in Table 5-4 were at least partially designed as
investigations of carcinogenicity and, hence, are also discussed in the
carcinogenicity section of this document; some experimental details are
discussed more fully in that section.
From the data presented in Table 5-4, it appears that rats and mice can
tolerate higher daily intakes of chloroform when it is given in their
drinking water or in a toothpaste base (by gavage) than they can when the
chemical is administered in corn or arachis oil (by gavage). In the
subchronic study of Jorgenson and Rushbrook (1980), rats and mice appeared to
adapt to low levels of chloroform intake (up to =100 mg/kg/day); signs of
depression and mild hepatic damage that occurred initially had disappeared by
90 days of treatment. Elevated indices of liver damage (e.g., SGPT levels)
in dogs chronically exposed to chloroform reverted to "normal" after
treatment was discontinued, although histological changes persisted.
5-30
-------
Similarly, the mild liver and thyroid lesions seen in rats exposed to high
doses of chloroform via their drinking water for 90 days were no longer
apparent in rats allowed to recover for an additional 90 days (Chu et al.,
1982b).
5.3. INVESTIGATION OF TARGET ORGAN TOXICITY IN EXPERIMENTAL ANIMALS
Chloroform is a hepatic and renal toxicant in a variety of animal
species including man. In mice, however, nephrotoxicity occurs only in males
and not in females, although the hepatotoxicity is similar in both sexes.
This sex difference in mice appears to be androgen-related. Condie et al.
(1983) have compared the renal and hepatoxicity of chloroform with other
halomethanes in male mice (CD-I) .after daily oral dosing in corn oil for
14 consecutive days at three dose levels. Toxicity was evaluated by
measuring change in total body weight, uptake of p-aminohippurate (PAH) into
renal cortical slices, blood urea nitrogen, serum creatinine, serum GPT, and
by performing histopathologic examination of liver and kidney tissues.
Chloroform was the most potent of the series followed by
dibromochloromethane, bromodichloromethane, bromoform, and methylene chloride
with the least organ toxicity.
5.3.1. Hepatotoxicity
An extensive review of the early literature dealing with
chloroform-induced liver damage by von Oettingen (1964) notes studies
beginning in 1891. More recently, Groger and Grey (1979) summarized reports
describing chloroform-induced liver hepatotoxicity as follows: typical
effects of chloroform on liver cells are extensive vacuolization,
disappearance of glycogen, fatty degeneration, swelling, and necrosis, all
starting in the centrilobular areas. There is also often hemorrhaging into
the parenchyma and infiltration of polymorphonuclear cells and monocytes.
5-31
-------
Electron-microscopic observations of liver parenchyma! cells from
chloroform-intoxicated rats as carried out by Scholler (1966, 1967) revealed
deposition of lipid droplets in the cytoplasm, partial destruction of the
mitochondrial matrix, proliferation of smooth endoplasmic reticulum, and
swelling of the rough endoplasmic reticulum with detachment of ribosomes.
Kylin et al. (1963) conducted a study of the hepatotoxic effects of
inhaled trichloroethylene, tetrachloroethylene, and chloroform in mice with
the objective of finding the lowest (individual) concentration of the
substances producing signs of liver damage after a single 4-hour exposure
period. Histological examination showed that concentrations of these three
agents at 100 ppm caused moderate fatty infiltration in mice killed 1 day
after exposure. At >200 ppm, the extent of the alteration increased with
concentration and was more pronounced after 1 day than 3. Thus, judging from
the histological picture, the smallest concentrations (ppm) of the different
agents to produce more severe alterations in the exposed group than in the
controls were as follows:
Trichloro- Tetrachloro-
ethylene ethylene
(ppm) (ppm)
1 day after 1600-3200 <200
exposure
Chloroform
(ppm)
<100
3 days after
exposure
>3200
200 to 400
100 to 200
On this basis, the hepatotoxic effects of trichloroethylene, tetrachloro-
ethylene, and chloroform are in the approximate ratios 1:10:20. The amount
of liver fat was raised at 400 ppm of chloroform. A third indicator of liver
toxicity was an increase in serum ornithine carbamyl transferase (S-OCT)
5-32
-------
activity at 24 hours in animals exposed to 200, 400, and 800 ppm of
chloroform.
A study of the effect of oral doses of chloroform on the extent of liver
damage in white mice ("of a Swiss strain") was conducted by Jones et al.
(1958). Minimal changes characterized by midzonal fatty infiltration were
observed 72 hours after the administration of 30 mg (0.02 ml)/kg. When the
dose was increased to 133 mg (0.09 ml)/kg, a massive fatty infiltration of
the total liver lobule was found. At a level of 355 mg (0.24 ml)/kg, massive
fatty infiltraton occurred along with severe central lobular necrosis.
Information on the hepatotoxicity of long-term chloroform administration has
been presented in the sections.on Effects of Chronic .Exposure to Chloroform.
Inhalation exposure of rats to 25, 50, or 85 ppm chloroform for 7 hours/day.
5 days/week for 6 months produced centrilobular granular degeneration and
focal necrosis in their livers. In subchronic studies, ingestion of up to
=100 mg/kg/day of chloroform produced mild, transient histological changes
in the livers of rats and mice (Jorgenson and Rushbrook, 1980), ingestion of
145 or 190 mg/kg/day produced fatty change in the livers of mice (Jorgenson
and Rushbrook, 1980), and administration of 410 mg/kg/day by gavage produced
fatty change and necrosis in the livers of rats. Dogs treated subchronically
with 45 mg/kg/day by the oral route had histological evidence of slight
hepatic fatty change, with increasingly severe changes noted at dosages of 60
and 120 mg/kg/day.
In chronic oral studies, rats had minor histological change in their
livers and a decrease in relative liver weights when given 60 mg/kg/day of
chloroform, 6 days/week, while the livers of mice were unaffected at this
dosage. Dogs had some evidence of liver damage (clinical chemistry
parameters) and an increase in the number and size of fatty cysts in their
5-33
-------
lifetime when administered 15 or 30 mg/kg/day orally for 6 days/week. The
mechanism by which chloroform exerts its hepatotoxic effects has been widely
investigated and efforts have been made to identify the responsible
metabolite(s).
As long ago as 1928, it was suspected that the liver damage induced by
chloroform may be due not only to the chemical itself, but might be caused by
a degradation product (Lucas, 1928). Chloroform has been determined to be
metabolized before being excreted (Butler, 1961; Paul and Rubenstein, 1963;
Van Dyke et al., 1964; Reid and Krishna, 1973). These studies indicate that
the tissue necrosis induced by chloroform is associated with the covalent
binding of toxic metabolites and alkylation of tissue proteins.
Autoradiograms have revealed that this binding occurs predominantly in the
necrotic areas (IIlet et al., 1973). McLean (1970) has shown that
pretreatment of rats with phenobarbital (a microsomal enzyme-inducing agent)
greatly enhances the lethality of chloroform.
Brown et al. (1974) proposed a mechanism of chloroform hepatotoxicity
implicating a free radical metabolite which can react with glutathione (GSH)
(a tripeptide which protects against hepatotoxicity), diminishing GSH levels
in the liver. According to this hypothesis, once GSH levels are depleted,
further metabolism would result in the reaction of the metabolite with
microsomal protein, and hence, necrosis. This proposal was based on
observations of phenobarbital pretreated rats anesthetized with chloroform,
in which hepatotoxicity was enhanced and GSH levels were decreased by the
induction of microsomal enzymes. Covalent binding of chloroform metabolites
to microsomal proteins in vitro was also enhanced by enzyme induction, an
effect prevented by GSH. Similar findings were reported in mice by Ilett
et al. (1973), who found severe chloroform-induced centrilobular necrosis in
5-34
-------
phenobarbital pretreated mice, but only slight centrilobular damage in mice
exposed only to chloroform.
Thus, the hepatotoxicity of chloroform appears to depend on (1) the rate
of its biotransformation to produce reactive metabolite(s), and (2) the
amount of GSH available to conjugate with and thus inactivate the
metabolite(s).
The role of GSH in chloroform-induced hepatotoxicity was further studied
by Docks and Krishna (1976), who found that only those doses of chloroform
that decreased liver GSH caused liver necrosis when administered to
phenobarbital pretreated rats.
More recently, Ekstrom et al. (1982) studied the. mechanism of GSH
depletion by chloroform in rats pretreated with phenobarbital. The synthesis
of GSH proceeds via two enzymatic steps, the first of which is rate limiting:
Y-glutamyl-cysteine
glutamate + cysteine _ dipeptide
synthetase
In the presence of glycine, the reaction continues via GSH synthetase to
produce GSH. When the soluble fraction from livers of rats sacrificed at
various times after chloroform exposure was incubated in the presence of
these amino acids, it was found that GSH synthesis was inhibited within
4-6 hours, while liver necrosis was evident only after 6 hours. When glycine
was eliminated from the initial part of the incubation, the dipeptide
accumulated, but at a lower rate in the presence of chloroform than in its
absence. Later addition of glycine resulted in GSH synthesis at a rate
similar to control values. Thus, it appears that chloroform, or rather a
reactive metabolite, inhibited GSH synthesis at the rate limiting step (i.e.,
the formation of dipeptide by y-glutamyl-cysteine synthetase).
5-35
-------
Savage et al. (1982) found that rats treated with chloroform for 1 to
7 days showed a dose-dependent increase in hepatic ornithine decarboxylase
(ODC), with a threshold at 100 mg/kg body weight. Females were 2 to 4 times
more susceptible than males, and rats dosed from 7 days showed a decline in
ODC susceptibility. Chloroform, rather than increasing the activity of renal
ODC, resulted in a 35% reduction. The induction by chloroform of hepatic ODC
might be associated with regenerative hyperplasia, while the renal
carcinogenicity of chloroform could not be associated with ODC induction.
The same researchers have reported more recently that chloroform's role in
inducing liver and renal ODC activity does not appear to be associated with a
regenerative response (Pereira et al., 1984).
The biotransformation of chloroform (as discussed in Chapter 4) depends
on the activity of the microsomal drug metabolizing enzymes. Substances that
induce these enzymes were shown to, indeed, enhance the hepatotoxicity of
chloroform, as evidenced by increased serum glutamic-pyruvic transaminase
(SGPT) levels and decreased hepatic glucose-6-phosphatase activity (Lavigne
and Marchand, 1974). An inhibitor of the drug metabolizing enzymes SKF-525A,
however, while increasing the excretion of ^4C-carbon monoxide in rats
administered 14C-labelled chloroform, failed to diminish the hepatotoxicity
of chloroform, leading these authors to conclude that factors other than
metabolism may be involved.
McMartin et al. (1981) demonstrated that altering the cytochrome P-450
concentrations in the livers of chloroform-exposed rats also altered the
hepatotoxicity, as measured by the incidence of hepatic lesions and by serum
alanine aminotransferase activities. Both fasting and phenobarbital
pretreatment increased the cytochrome P-450 content and liver damage, while
cadmium produced the opposite effect.
5-36
-------
Theories of chloroform hepatotoxicity Involve the formation of reactive
intermediates by liver enzymes. How these intermediates exert their
hepatotoxic effect has been the subject of several studies. It has been
suggested by Masuda et al. (1980) that, based on the chloroform-induced
indices of hepatotoxicity of decreased microsomal glucose-6-phosphatase
activity and cytochrome P-450 content with increased hepatic malondialdehyde
levels, the lipid peroxidation hypothesis proposed for carbon tetrachloride
may also apply to the case of chloroform. Qualitative and mechanistic
differences of hepatotoxicity between the two chemicals were noted, however.
The interactive hepatotoxicity of chloroform and carbon tetrachloride
was studied by Harris et al. (1982), who found that neither chemical alone
given at subthreshold dose altered SGPT activity, hepatic triglyceride
content, or hepatic calcium content. However, when given together,
chloroform and carbon tetrachloride increased the toxic response in rats.
Administration of either or both chemicals had no effect on GSH levels or
conjugated diene formation, but ethane expiration was increased in rats given
both chemicals. Diene conjugation and ethane expiration are indices of lipid
peroxidation. Histopathological changes were more severe from the
combination than from etther chemical alone. Although the mechanism of the
hepatotoxic interaction between chloroform and carbon tetrachloride is
unclear, the authors suggested that there might be a combined effect of
phosgene formation and lipid peroxidation initiation.
It should be noted that the prevailing theories implicate phosgene as
the major metabolite responsible for chloroform hepatotoxicity (Reynolds and
Yee, 1967; Sipes et al., 1977; Mansuy et al., 1977; Pohl et al., 1977).
Other potential toxic metabolites discussed by Pohl (1979) in a review of
5-37
-------
this subject are a trichloromethyl radical and dichlorocarbene; however, they
are considered less important than phosgene in this regard.
A study by Stevens and Anders (1981) supports the phosgene-mediated
mechanism. The time course of changes in SGPT levels and covalent binding of
^C to proteins was examined in microsomal and soluble fractions from
phenobarbital-pretreated rats sacrificed at various times after chloroform or
14C-chloroform administration. It was found that ^4C binding was maximal at
6 hours while indices of liver damage peaked at 18 hours after chloroform
exposure. Further experiments were performed in which diethyl maleate (which
depletes GSH depletor) treatment caused increased ^C-binding to soluble and
microsomal fractions and increased SGPT levels, perhaps by inhibiting the
metabolism of phosgene to carbon monoxide or stable conjugates. Cysteine,
which reacts with phosgene to produce 2-oxothiazolidine-4-carboxylic acid,
had a protective effect. Diethyl maleate also diminished, but did not
eliminate the deuterium isotope effect on the GSH dependent chloroform
metabolism to carbon monoxide, which would be expected if carbon monoxide
formation occurred subsequent to phosgene production. Thus, the
hepatotoxicity of chloroform can be altered by altering the various reaction
pathways of phosgene, strongly indicating that phosgene is the toxic inter-
mediate.
From the above discussion, it appears that to be hepatotoxic, chloroform
must first be metabolized by microsomal drug metabolizing enzymes to an
active intermediate, probably phosgene, which in turn can react by various
pathways, depending on GSH levels. One pathway is the covalent binding to
liver proteins, resulting in necrotic lesions. Lavigne et al. (1983) have
pointed out that circadian rhythms may also influence chloroform-induced
hepatotoxicity. These investigators determined the diurnal variation of
5-38
-------
chloroform-induced hepatotoxicity in Sprague-Dawley rats. Animals were dosed
at 9:00, 13:00, 17:00, 21:00 or 03.00 hr (i.p., 0.5 ml/kg) and sacrificed
4 hr after each dosing. Measures of toxicity included serum GPT, GOT, LDH
and liver glucose-6-phosphatase. Toxicity was minimum after 9:00 hr dosing
and maximal 12 hr later at 21:00 hr dosing. Prior 16 hr food deprivation
increased the level of toxicity. Lavigne et al. suggest that a correlation
exists between the diurnatity of chloroform-induced hepatotoxicity and drug
metabolizing enzyme activity, which in rats is higher in the evening than in
the morning.
5.3.2. Nephrotoxicit.y
As noted by Watrous and Plaa (1972), the extensive body of research on
the hepatotoxicity of halogenated hydrocarbons has tended to overshadow the
fact that some of these agents are also nephrotoxic. Earlier reports of
chloroform nephrotoxicity include those of Heller and Smirk (1932), who found
that rats anesthetized with chloroform showed a diminished ability to excrete
a water load given prior to anesthesia, and Knocher and Mandelstam (1944),
who noted that chloroform injection produced a fatty infiltration of the
kidney.
Renal necrosis produced by the oral administration of chloroform was
described by Eschenbrenner and Miller (1945a). The necrosis, observed only
in male mice, involved portions of both proximal and distal convoluted
tubules. The nuclei of the epithelial cells were often absent or fragmented
and the cytoplasm was coarsely granular and deeply eosinophi1ic. The
glomeruli and collecting tubules appeared normal.
Sex and strain differences in the sensitivity of mice to chloroform
nephrotoxicity were further studied by Deringer et al. (1953). Exposure of
strain C3H mice to air containing =5 mg/L of chloroform for 1, 2, or 3 hours
5-39
-------
resulted in lesions of the kidneys of all of the males but in none of the
females. In animals dying within 1 day after exposure, epithelium of the
proximal tubules and portions of the distal tubules were generally necrotic.
The lumens of those segments of tubules were dilated. The glomeruli were
relatively unaffected. The mice dying or sacrificed at later time intervals
exhibited calcification in the necrotic area.
Similar lesions were found in males of strains C3H, C3Hf, A, and HR.
However, strains C57BL, C57BR/cd, C57L, and ST were resistant to chloroform-
induced nephrotoxicity.
Comparable results were reported by Krus and Zaleska-Rutczynski (1970).
Subcutaneous administration of chloroform to C3H/He male mice resulted in
renal tubular necrosis, with death ensuing 4-9 days later. The lesions were
calcified with no evidence of regeneration. Female mice of this strain,
males and females of the C57BL/6JN and BN strains, and F^ generation males of
the cross of female C3H/He with male C57BL/6JN mice survived the
administration of chloroform (0.1 ml of 0.05 g chloroform in 1 ml ethyl
laureate). Additional studies were performed with males and females of this
FI generation and the resistant BN strain, in which animals were sacrificed
at various times after chloroform administration. All mice survived, while
all female mice were resistant, showing no kidney lesions at any time in the
experiment. Renal damage was morphologically apparent in all male mice by
12 hours, but regeneration developed by day 4 and continued until the end of
the experiment. It was concluded that although all male mice had tubular
lesions, the ones surviving had tubules that did not calcify and a large
degree of renal regeneration.
Several investigators have studied the influence of testosterone on
chloroform-induced renal damage. Eschenbrenner and Miller (1945b) performed
5-40
-------
experiments in which they saw extensive necrosis of portions of the proximal
and distal renal tubules in normal male and in testosterone-treated castrated
male mice following the acute oral administration of chloroform. However, no
necrosis was found after chloroform was administered to female mice or
castrated male mice. Continuing this line of investigation, Culliford and
Hewitt (1957) reported the following results:
1) Adult male mice of two strains (CBA and WH) developed extensive
necrosis of the renal tubules after exposure to low concentrations of
chloroform vapor (7-10 mg/L for 2 hours). Adult females showed no
renal damage after equivalent exposure.
2) Adult females became fully susceptible to necrosis after treatment
with androgens. The susceptibility of males was greatly reduced by
treatment with estrogens.
3) Castration removed the susceptibility of the males of one strain, but
did not completely remove it in another. The residual susceptibility
of castrates was abolished by adrenalectomy.
4) Male mice under 11 days old were not susceptible to necrosis even
after massive doses of androgen. Between 11 and 30 days, they were
susceptible if given androgen. Thereafter, they became spontaneously
susceptible.
5) Liver damage occurred in nearly all exposed mice and was not
correlated with sex hormone status.
6) Susceptibility could be induced in gonadectomized mice by methyl
testosterone, testosterone propionate, dehydroepiandrosterone,
progesterone, and large doses of cortisone acetate.
Hill (1978) also performed experiments demonstrating similar strain and
sex differences in chloroform-induced renal toxicity. The renal toxicity of
a fixed oral dose of chloroform to castrated male mice was increased with
increasing doses of administered testosterone. Plasma levels of testosterone
in resistant strains tended to be lower than levels in susceptible strains.
Hill (1978) conjectured that a testosterone may act by sensitizing the renal
proximal convoluted tubules to chloroform through a testosterone receptor
mechanism. Eschenbrenner and Miller (1945b), however, linked susceptibility
5-41
-------
to the nephrotoxic action of chloroform to differences in kidney morphology
and physiology induced by testosterone.
Information on the nephrotoxicity of long-term chloroform administration
has been presented in the section on Effects of Chronic Exposure to
Chloroform. Inhalation of 25, 50, or 85 ppm of chloroform 7 hours/day,
5 days/week for 6 months produced cloudy swelling of the renal tubular
epithelium in rats. Male mice of certain sensitive strains had increased
incidences of moderate to severe renal disease when treated orally with
chloroform at a dosage of 60 mg/kg/day, 6 days/week in a chronic study (Roe
et al., 1979). Chronic oral administration of 30 mg/kg/day of chloroform,
6 days/week, to dogs produced an increase in the numbers of renal glomeruli
affected by fat deposition (Heywood et a I., 1979).
The mechanism of the nephrotoxicity of chloroform has been extensively
studied in the last few years. Earlier, the possibility that hepatotoxic
metabolites produced in the liver might be transported via the circulation to
the kidney to produce nephrotoxic effects had been considered (Ilett et al.,
1973). Recent studies have shown that chloroform is metabolized within the
kidney and the nephrotoxic effects of chloroform are the resultant of that
metabolism. The evidence can be summarized as follows:
1) It has been known that kidney slices from rats (Paul and Rubinstein,
1963) and more recently from mice (Smith and Hook, 1983) can
metabolize chloroform to C02, which is a known degradation producer
of phosgene (Pohl et al., 1980). Smith and Hook (1983, 1984) have
further shown that l^C-labelled chloroform is metabolized by renal
cortical microsomes from male ICR mice (but to a markedly lesser
extent by renal microsomes from female mice) to 14C02 and to reactive
metabolites that covalently bind to microsomal protein and lipids.
5-42
-------
renal metabolism, although 3- to 5-fold less active than hepatic
microsomal metabolism, also required NADPH and 02. P450 binding of
chloroform, as in the liver, was Type 1 and was inhibited by CO,
while addition of GSH decreased ^CQ? production and covalent
binding. Pohl et al. (1984) found that renal cortical homogenates
from chloroform sensitive ICR male mice (10-fold more active than
female mice) fortified with NADPH, 02 and cysteine metabolized
chloroform to phosgene, which was detected by HPLC as the cysteine
derivative, 2-oxothiazolidine-4-carboxylic acid (OTZ). P45Q in both
microsomal and mitochondrial fractions (microsomes 20x more active)
catalyzed chloroform to phosgene. Cleavage of the C-H bond of
"chloroform was the rate-limiting step since chloroform metabolism was
3x more rapid than with the deuterated analog CDCls. Branchflower
et al. (1984) have reported entirely similar findings; after
incubation of kidney homogenates with GSH they identified OTZ and
OTZG (N-(2-oxothiazolidine-4-carboxyl)-glycine) formed after initial
reaction of phosgene with glutathione. The intermediate
diglutathionyl dithiocarbonate (GSCOSG) was not identifiable, but it
was found to be rapidly converted to OTCG and OTZ by kidney y-
glutamyl transpeptidase and cystinyl glycinase. Figure 5-1 shows the
probable pathways of metabolisim of chloroform in the kidney. OTZ
and GSCOSG are not nephrotoxic either i_n vivo or j_n vitro (as
assessed by effects on PAH accumulation by renal cortical slices).
The pathway therefore represents a detoxification mechanism for
phosgene, the reactive species formed in the kidney by P45Q mediated
oxidative metabolism of chloroform.
5-43
-------
Cl —
H
1
C-CI
1
Cl
P450
system
62 NADPH
OH
1
Cl — C — Cl
1
Cl
Cl.
Cl
c=o
H2O
Phosgene
+ 2 GSH
CO2 + HCI
GS- C- SCH2CH - C- NHCH2CO2H
I
NH- C- CH2CH2CH—CO2H
O NH2
Diglutathionyl dithiocarbonate (GSCSG)
Y glutamyl
transpeptidase
O O
GS - C- SCH2CH- C- NHCH2C - OH
NH2
Glutathionylcysteinyl glycinyl dithiocarbonate (GSCOSCG)
r-r
C- NHCH2COOH
C
II
O
Cysteinyl glycinase
COOH
c
II
O
N-(2-oxothiazolidine-4-carbonyl)glycine 2-oxo-thiazolidine-4-carboxylicacid
(OTZG) (OTZ)
Figure 5-1. Probable pathways of metabolism of chloroform in the kidney
(from Branchflower et a!., 1984).
5-44
-------
2) The administration of chloroform to mice or rats produces a depletion
of kidney GSH, which indicates that phosgene is formed as a
metabolite in vivo (Kluwe and Hook, 1981), while deuterated
chloroform (CDCls) depletes renal GSH to a lesser extent and is less
toxic (Branchflower et al., 1984). Similarly GSH added to renal
homogenates or microsomes decreases covalent binding from chloroform
metabolism (Smith and Hook, 1984).
3) The capacity of the kidney to metabolize chloroform to phosgene
correlates with strain and sex differences in susceptibility to
chloroform-induced nephrotoxicity. Ahmadizadeh et al. (1984a)
investigated the possibility that morphoTogtcal differences in cell
type between sexes and strains of mice may account for differences in
renal toxicity to chloroform. However, these workers found that
renal lesions induced by chloroform in male strains of mice were
predominantly located in proximal convoluted tubule epithelial cells;
no histopathological changes were observed in cuboidal cells of
Bowman's capsule, although differences in cell type occurs between
sexes as well as strain in capsular epithelium. They concluded that
these morphological differences between sexes and strains probably
are not responsible for the pathophysiological difference seen with
chloroform, but rather a common factor such as testosterone effect on
metabolism and on morphology. Smith et al. (1983) explored the time
course of chloroform toxicity in male and female ICR mice. They
observed a rapid nephrotoxic effect in male mice as early as 2 hr
after s.c. injections of chloroform (250 ^l/kg). Renal cortical
slices from treated male mice exhibited decreased uptake and
accumulation of PAH and TEA, paralleled by a decrease of glutathione
5-45
-------
tissue content. Cytopathologic changes appeared 5 hr after
chloroform treatment. Female mice showed no evidence of renal
toxicity, even with the administration of diethylmaleate to deplete
tissue glutathione. On the other hand, hepatotoxicity was similar in
male and female rats and occurred with a depletion of hepatic
glutathione. These workers concluded that the temporal events in
liver and kidney toxicity suggested that chloroform was metabolized
by both liver and kidney in male mice but only in the liver of female
mice. Smith et al. (1983) provided further evidence from i_n vitro
studies. Renal cortical slices from ICR mice incubated with
chloroform resulted in a decrase of PAH or TEA uptake and
accumulation when the slices were from male mice but not female mice.
Deuterated chloroform (CDC13) had a lesser effect than chloroform.
Furthermore, the chloroform toxicity was inhibited by CO or by
incubation of the slices at 0°C. Conversely, renal slice toxicity
(as measured by PAH or TEA uptake) was increased by pretreatment of
the male mouse with diethylmaleate, a depletor of tissue GSH. Smith
et al. (1984) also investigated the effect of sex hormonal status on
chloroform nephrotoxicity and renal mixed function oxidases in ICR
mice. They found that renal microsomal content of cytochrome P45Q
was greater (5-fold) in kidneys from male mice versus female mice.
Similarly cytotochrome b5 content was greater (2-fold) and P45Q mixed
function oxidase activities (ethoxycoumarin 0-deethylase) was greater
(3- to 5-fold). Treatment of female mice with testosterone increased
renal P45Q content and activity to the male mouse level, while
castration of male mice reduced P450 content and activity to that of
female mice. Smith and Hook (1984) showed further that renal
5-46
-------
cortical microsomes from ICR male mice, but not female mice,
metabolized 14C-chloroform to reactive metabolites that covalently
bind to protein and lipid. These studies are consistent with the
concept that chloroform is metabolized by cytochrome P45Q mechanisms
in the kidney and the resulting sex difference of renal metabolic
lesions is correlated with a sex difference in renal metabolic
capacity to metabolize chloroform. Pohn et al. (1984) have also
provided strong experimental evidence for this concept. They found
that kidney homogenates of sensitive male DBA/2J mice metabolized
chloroform to phosgene 2x more rapidly than less sensitive C57BL/6J
strain, and kidney homogenates from male IRC sensitive strain
metabolized chloroform to phosgene lOx more rapidly than homogenates
from female mice of this strain. Furthermore treatment of IRC female
mice with testosterone induced in their kidney homogenates the
ability to metabolize chloroform to phosgene to the same level of
male mice.
4) Uptake hepatic metabolism and hepatotoxicity of chloroform,
potentiation of chloroform toxicity by P45Q inducers, example
phenobarbital, cannot be demonstrated in rat or mouse kidneys, which
are not inducible by phenobarbital (McMartin et al., 1981;
Ahmadizadeh et al., 1984b). Similarly, inhibitors of P45Q
metabolizing system, SKF-525A, piperony butoxide, etc., do not reduce
the nephrotoxicity of chloroform in sensitive mice (Kluwe and Hook,
1981; Lavigne and Marchand, 1974). Ahmadizadeh et al. (1984b)
pretreated subchronically the chloroform renal-sensitive DBA/2J
strain of male mice and the relatively resistant C57BL/6J male strain
with P450 inducers phenobarbital, (3-naphthoflavone, and
5-47
-------
polybrominated biphenyl (PBBs). None of these inducers increased
renal toxicity in DBA/2J mice, although all increased hepatic
toxicity. However, PBBs increased both hepatic and renal toxicity
and induced P450 systems in the resistant C57BL/6J strain
(Ahmadizadeh et al., 1984b; Kluwe and Hook, 1978). McMartin et al.
(1981) have also shown that fasting increased P45Q concentrations in
both liver and kidney of rats, and chloroform-induced damage was
enhanced in both organs of fasted animals. Bailie et al. (1984) have
recently shown potentiation of chloroform metabolism and toxicity
occurs in rabbit kidneys after pretreatment with phenobarbital.
Phenobarbital is a known inducer of the ?450 metabolizing system in
rabbit kidneys. Renal cortical slices from phenobarbital-treated
rabbits incubated with chloroform exhibited greater chloroform dose-
related decreases in uptake and accumulation of PAH and TEA than non-
phenobarbital treated controls. Covalent binding from l4C-chloroform
metabolism was enhanced 5-fold in renal slices and microsomes
prepared from phenobarbital-treated rabbits. Addition of cysteine to
the microsomal incubations resulted in decreased binding, but an
increase of 14C-OTZ, a measure of phosgene production. The lack of
correlation in mice and rats, but a good correlation in rabbits,
between hepatotoxicity and nephrotoxicity of chloroform following
pretreatment with' enzyme inducers supports the concept that the
nephrotoxic metabolite(s) of chloroform are generated in the kidney
and do not derive from liver metabolism.
The mechanism underlying the genetically determined strain and sex
differences for chloroform metabolism and nephrotoxicity in the rodent remain
incompletely resolved. Pohl et al. (1984), noting that testosterone is known
5-48
-------
to elicit a variety of inductive effects in the mouse kidney, which include
hypertrophy and augmentation of many enzyme activities, including the
metabolism of chloroform to phosgene, suggest plasma testosterone levels
(3,5-fold higher in sensitive DBA/2J male mice than in resistant C57BL/6J
male mice) may be responsible, at least in part, for both sex and strain
differences in sensitivity to chloroform-induced nephrotoxicity.
5.4. FACTORS MODIFYING THE TOXICITY OF CHLOROFORM
From the preceding discussion, the severity of toxic effects induced by
a given amount of chloroform is influenced by alterations of microsomal
enzyme activity or hepatic GSH levels. Thus, many factors including exposure
to other chemicals alter chloroform toxicity by affecting these parameters or
acting through" other mechanisms. These factors are of interest because they
fall into categories of nutritional status and accidental or intentional
chemical exposure by humans. As an example of action and interactions of
these categories, Sato and Nakajima (1984) have recently reviewed the
experimental evidence for dietary carbohydrate and ethanol-induced alteration
of the metabolism and toxicity of chemical substances with chloroform and
carbon tetrachloride as illustrative compounds. Sato and Nakajima put forth
three main conclusions:
1) Food deprivation causes a 2- to 3-fold increase in the metabolism of
various chemicals in rat liver, such as chloroform and carbon
tetrachloride. The lack of carbohydrate rather than the lack of
protein or fat accompanying food deprivation is primarily responsible
for the increase.
2) In contrast to general belief, dietary carbohydrate plays an
important role in regulating drug-metabolizing enzyme activity in the
liver; a low-carbohydrate diet enhances, whereas a high-carbohydrate
5-49
-------
diet represses, the hepatic metabolism of a variety of volatile
hydrocarbons, irrespective of protein and/or fat content(s).
3) Ethanol and carbohydrate are antagonistic in connection with hepatic
metabolism: the former increases and the latter decreases it. A
decrease (increase) in carbohydrate intake augments (suppresses) the
action of ethanol in a dose-related manner: ethanol consumed with
lowered carbohydrate intake results in a more remarkable increase in
hepatic metabolism than does ethanol consumed with moderate
carbohydrate intake.
Alcohol, dietary components, pesticides, and steroids are some of the
substances which are discussed further below.
5.4.1. Factors that Increase the ToxJcity
The effect of ethanol pretreatment on chloroform-induced heptatoxicity
in mice was studied by Kutob and Plaa (1962a). An intoxicating dose (5 g/kg)
of ethanol was administered orally to mice daily for 15 days initially, with
a systematic shortening of the duration to a single exposure. A challenging
dose of chloroform (0.08 ml/kg) was administered subcutaneously either 12,
15, or 24 hours after ethanol treatment. Liver dysfunction was measured by
prolongation of phenobarbital sleeping time, bromsulphalein (HSP) retention,
liver succinic dehydrogenase activity, and histological examination.
Regardless of the ethanol treatment period, phenobarbital sleeping time was
significantly increased in mice receiving ethanol followed by chloroform when
compared with mice receiving either substance alone. Similar findings were
found for BSP retention. The in vitro succinic dehydrogenase activity was
significantly reduced by ethanol pretreatment followed by chloroform
administration 12 or 24 hours, but not 48 hours, later, when compared with
activities from mice receiving only chloroform. Histological changes were
5-50
-------
seen in the livers of mice given ethanol 15 hours to 4 days prior to
chloroform challenge, while mice receiving either chemical alone had
morphologically normal livers. It was also determined that the ethanol
treatment increased liver triglyceride content, with a maximum at 15 hours,
and that ethanol pretreatment significantly increased the concentration of
chloroform in the livers with a maximum at 12 hours after chloroform
challenge. From these results, it was noted that a single dose of ethanol
was just as effective as multiple doses. A mechanism was proposed for the
ethanol enhanced chloroform-induced hepatotoxicity in which ethanol increases
liver lipid content (as evidenced by increased triglycerides) resulting in
increased concentrations-of chloroform to be metabolized in the liver.
In support of this-mechanism is the observation that oral isopropanol
pretreatment for 5 days (0.3 ml/100 g for 2 days and 0.15 ml/100 g for
3 days) followed 12 hours later by 5 daily inhalations of chloroform (5000
ppm first day, 2500 ppm on the next four days), 2 hours/day led to severe
fatty infiltration of the liver. Chloroform alone increased the pool of
triglycerides (Danni et al., 1981).
In contrast, Sato et al. (1980) studied the mechanism by which ethanol
enhances hydrocarbon metabolism, including that of chloroform. Rats
ingesting ethanol in their drinking water for 3 weeks were sacrificed
10 hours after the final exposure. Control rats were given isocaloric
glucose solutions. Liver microsomal enzyme systems were prepared and liver
protein and cytochrome P-450 contents analyzed, the increased contents being
indicative of microsomal enzyme synthesis in response to alcohol. When
chloroform was added as a substrate, its metabolism was enhanced by 6 times,
much more than could be accounted for by enzyme induction alone. Microsomes
prepared from rats that were withdrawn from ethanol 24 hours prior to
5-51
-------
sacrifice did not show enhanced activity. In a subsequent study (Sato et
al., 1981), rats receiving a single gavage dose of 0, 2, 3, 4, or 5 g/kg
ethanol were sacrificed 18 hours later. The in vitro metabolism of
chloroform by microsomes prepared from these rats was enhanced very little at
2 g/kg, slightly more at 3 g/kg, and dramatically at 4 g/kg. At 5 g/kg,
however, enhanced enzyme activity was no greater than at 3 g/kg ethanol.
When ethanol was added directly to the incubation system, the metabolism of
chloroform was inhibited. Rats receiving 5 g/kg ethanol retained relatively
large amounts in the blood and liver, while those receiving 4 g/kg retained
almost none. If the ethanol remaining in the rats exerted an inhibitory
effect on enzyme activity, then microsomal enzymes prepared from 5 g/kg
ethanol-treated rats, mixed with the soluble fraction from control rats,
should show increased activity when compared with both microsomal and soluble
fractions from ethanol-treated rats. This was found to be the case.
Based on these results and studies on metabolism of other hydrocarbons
In vitro and in vivo, Sato et al. (1981) suggested that ethanol is both a
stimulator and an inhibitor of drug metabolizing enzymes, depending on how
much ethanol remains in the body and thus how much time has elapsed since
ethanol ingestion. Thus, when ethanol is first ingested, it acts as a
competitive inhibitor of microsomal enzyme activity, but as it disappears
from the body, an optimum for stimulation may be reached and metabolism
enhanced. It was postulated that since the metabolism of chloroform was
enhanced to a much greater extent than can be explained from enzyme induction
alone, perhaps ethanol modifies the enzyme activities by other mechanisms
such as modification of membrane properties, allosteric effects, or by
displacement of substrate already bound.
5-52
-------
Polybrominated biphenyls (PBBs) have also been found to potentiate the
toxicity of chloroform (Kluwe and Hook, 1978). Mice were fed diets
containing 0, 1, 25, or 100 ppm PBB for 14 days. One day before sacrifice,
the mice were given a single intraperitoneal injection of 0, 0.5, 2.5, 5.0,
or 50 pi/kg chloroform. PBB enhanced the toxicity of chloroform in both the
liver and the kidney as evidenced by results of blood urea nitrogen (BUN) and
serum glutamic oxaloacetic transaminase (SGOT) determinations and by
inhibition of p-aminohippuric acid (PAH) uptake by renal slices. PBB also
reduced the LD^Q of chloroform in these mice, and the deaths were attributed
to hepatic necrosis. Since PBBs were known to induce the drug metabolizing
enzymes, their effects on chloroform were assumed to be due to enhanced
chloroform metabolism. Similar findings have been described for other
species and other inducers of mixed function oxidases. Thus Abdelsalam et
al. (1982) have reported that dteldrin, as does phenobarbital, potentiates
chloroform-induced hepatotoxicity of male Nubian goats.
Steroids appear to play a role in the potentiation of chloroform
toxicity, especially in the kidney as seen from the sex-related differences
in the response of mice (Eschenbrenner and Miller, 1945a; Deringer et al.,
1953) and by experiments involving testosterone administration to castrated
male mice (Eschenbrenner and Miller, 1945b; Culliford and Hewitt, 1957; Hill,
1978) discussed previously. Clemens et al. (1979) further studied this
phenomenon in castrated male and intact female mice. Dose-dependent
testosterone sensitization of renal tubules to a fixed dose of chloroform was
observed in castrated males with the response ranging from kidney dysfunction
to death at high doses. Medroxyprogesterone acetate, an androgenic
progestin, enhanced chloroform-induced kidney damage in both castrated males
and intact females. Progesterone or hydrocortisone potentiated chloroform
5-53
-------
toxicity in DBA/2J castrated male mice, but not in the C57BL/J6 strain males
nor in any of the females. The mechanism by which the androgens exerted
their potentiation may have been mediated through strain specific androgen
receptors of the proximal convoluted tubular cells. The mechanism for the
potentiating action of the other steroids was less clear.
The potentiation of chloroform toxicity by various ketone and ketogenic
substances such as isopropanol, acetone, 2-hexanone and the insecticide
chlordecone (Kepone) has been studied extensively in recent years. Hewitt et
al. (1979) and Cianflone et al. (1980) have shown that while pretreatment of
mice with chlordecone enhanced the liver damage caused by chloroform
exposure, the non-ketonic structural analog mirex did not. Animals
administered chloroform alone, when compared to those pretreated with
chlordecone plus chloroform, have different histological pattern of damage
(Hewitt et al., 1979), despite similar rate of chloroform bioactivation
(Cianflone et al., 1980). Thus it has been postulated that chlordecone
pretreatment may not only bioactivate chloroform metabolism but also perturb
the macromolecular localization of the chloroform-derived reactive
intermediates, thereby changing the evolution of the toxic process. Hewitt
et al. (1983b) compared the pattern of chloroform binding to hepatocyte
macromolecular constituents, at various time points, in animals pretreated
with chlordecone or its nonpotentiating analog mirex. Macromolecular
distribution of chloroform-derived reactive intermediate(s) was assessed by
determining irreversibly bound 14c in the protein, lipid and acid soluble
fractions after administering l^C-chloroform. Rats and mice were given
chlordecone and mirex by gavage 18 hr before 14C-chlorofrom challenge dose
and killed 1 hr later. A marked increase of total l^C-hepatic binding
occurred with chlordecone but not with mirex. Chlordecone binding was
5-54
-------
distinguished by a significant increase of protein and lipid binding with a
concomitant decrease in acid-soluble fraction binding, i.e., an alteration in
binding distribution in hepatocytes. Since mirex is also a known inducer of
MFO enzymes (like chlordecone), but lacks the chloroform potentiating
ability, these investigators suggest that these two chemicals induce
catalytically different forms of cytochrome P45Q with different substrate
specificities (Kaminsky et al., 1978), as well as altering the cellular
distribution of chloroform-derived toxic metabolites. lijima et al. (1983)
have noted from morphologic measurements that pretreatment of rats with
chlordecone caused a dose-dependent increase in number of cells (area)
affected by chloroform (independently of severity), as well as an enhanced
severity of cellular change for a given chloroform dose.
Other ketones have also been compared for their ability to enchance the
hepatotoxic and nephrotoxic action of chloroform in rats with the following
results: methyl n-butyl ketone (MBK) and 2,4-hexanedione were the most
potent enhancers, followed by acetone and n-hexane (a ketogenic chemical)
(Hewitt et al., 1980). Hewittt et al. (1983a) have noted a positive
correlation betweeen ketone carbon chain length (3 to 7C ) and the severity
of the potential hepatotoxic response to chloroform in rats, even though the
ketones at the dose levels given did not themselves produce significant
hepatotoxicity. Jernigan and Harbison (1982) speculated that perhaps female
mice have greater microsomal enzyme activities, different membrane
properties, or perhaps produce a different reactive metabolite of chloroform
than do males.
The mechanism of ketone potentiation of chloroform-induced hepato- and
nephrotoxicity was also investigated by Branchflower and Pohl (1981) using
MBK. Male rats were pretreated with MBK followed by chloroform
5-55
-------
administration. The metabolism of chloroform by liver and kidney microsomal
enzymes and the toxicity to these organs were examined. Control experiments
were conducted in which rats were either not pretreated, not given
chloroform, or given deuterated chloroform (CDC13) instead of chloroform.
MBK increased cytochrome P-450 levels and NADPH-dependent cytochrome
reductase activity in liver microsomes, while having no effect on renal
levels of these microsomal components. MBK pretreatment doubled the rate of
metabolism of chloroform to diglutathionyl dithiocarbonate (GSCOSG) in
microsomal preparations, and more GSCOSG was excreted into the bile of the
pretreated animals when compared with rats receiving only chloroform. The
amount of GSCOSG in bile was less in MBK-COCla- treated animals. GSH levels
were significantly decreased by MBK treatment and this-decrease was enhanced
following chloroform exposure, and to a lesser extent, following CDCls
exposure. Rats pretreated with MBK followed by chloroform had greatly
elevated levels of SGPT associated with liver necrosis and signifipantly
greater BUN levels associated with renal cortical tubule lesions over the
control groups. A mechanism was proposed whereby MBK, by increasing
cytochrome P-450 levels, enhanced the metabolism of chloroform to phosgene.
Furthermore, according to the hypothesis, the phosgene was converted to
GSCOSG through GSH, levels of which were diminished by MBK, because the more
phosgene formed, the more GSH was depleted in the reaction. The results with
CDC13 indicated that C-H bond was involved in the mechanism. Although MBK
also potentiated chloroform toxicity to the kidney, a different mechanism may
have been involved since renal cytochrome P-450 and renal GSH levels were not
affected.
More recently, Branchflower et al. (1983) have shown that in rat liver
MBK potentiates toxicity at least in part by altering the composition of MBK-
5-56
-------
induced P45Q isocytochromes that metabolize chloroform to phosgene.
Microsomes from MBK and phenobarbital-treated male rats were isolated and
subjected to gel electrophoretic and anion-exchange chromatography analysis.
The qualitative character of the liver P45Q cytochromes formed by the two
inducers appeared similar. Isolation and reconsititution of the major P45Q
isocytochrome induced by MBK and by phenobarbital showed a parallel ability
to form phosgene from chloroform and chloroform-induced hepatotoxicity.
However, MBK-induced microsomes were significantly more active than
phenobarbital-induced microsomes in the formation of phosgene from chloroform
as expressed in terms of nanomoles of GSCOSG formed per nanomole of
cytod,hrome P450- Cowlen et al. (1984) have also investigated the mechanisms
of ketone-potentiated toxicity of chloroform. These workers oi^ally dosed
rats with 2-hexanone in corn oil and 18 hr later administered a challenge
dose of chloroform. The rats were killed 1, 2, and 6 hr later. Hepatic
activity of NADPH-succinate-dependent cytochrome C reductases were determined
as marker enzymes of endoplasmic reticulum and of mitochondrial membranal
function, respectively. Also measured were three indices of
lipoperoxidation: formation of conjugated dienes, depletion of unsaturated
fatty acids, and production of malondialdehyde. NADPH-dependent cytochrome C
reductase activity increased modestly (59%) one hr after chloroform challenge
(as compared to chloroform alone), but succinate-dependent cytochrome C
activity decreased (87%). Lipoperoxidation was not initiated at the doses of
hexanone and chloroform used, although liver damage was increased 30-fold by
the ketone as assessed by other measures. These investigators concluded that
2-hexanone potentiation of chloroform hepatotoxicity could not be attributed
solely to potentiated lipid peroxidation, but that the changes in succinate-
5-57
-------
dependent cytochrome C reductase activity suggests the existence of an
additional mechanimsm independent of the bioactivation of chloroform.
5.4.2. Factors that Decrease the Toxicity
As discussed above, the experiments of Sato et al. (1980, 1981) indicate
that ethanol is both a stimulator and an inhibitor of microsomal enzymes, and
hence, of chloroform metabolism and toxicity. depending upon the length of
time after ingestion.
Disulfiram and its metabolites have also been studied with respect to
their protective effects of chloroform-induced hepatotoxicity (Scholler,
1970; Masuda and Nakayama, 1982). Disulfiram is used in treating chronic
alcoholism and is metabolized to carbon disulfide and diethyldithiocarbamate
(a herbicide) (IARC, 1976). Disulfiram, a known inhibitor of the microsomal
drug metabolizing enzymes, given to rats prior to chloroform anesthesia
completely prevented the elevated SGPT activity and liver necrosis observed
in rats administered chloroform alone (Scholler et al., 1970). More
recently, Masuda and Nakayama (1982, 1983) studied the effects of
diethyldithiocarbamate (DTC) and carbon disulfide (C$2) pretreatment (given
30 min, orally) in male mice challenged with chloroform (i.p., 0.25 ml/kg).
DTC (mediated through C$2 production in stomach) and C$2 had protective
effects on both liver and renal toxicity as measured 24 hr later by SGPT
activity, phenolsulfonphalein clearance, liver and kidney calcium
accumulation, renal PAH accumulation, and cytopathology. DTC and C$2
decreased P45Q content and drug metabolizing enzyme activities in both liver
and kidney cortex (but i_n vitro only in the presence of NADPH), indicating
that these compounds must first be metabolized before exerting their
inhibitory effect on chloroform metabolism. Gopinath and Ford (1975) also
found that DTC or C$2 protected against chloroform hepatotoxicity in rats,
5-58
-------
and the effect was presumed to be due to inhibition of drug metabolizing
enzymes.
Dietary components can also alter the toxicity of chloroform. It is a
widely held opinion that low protein content of the diet decreases microsomal
enzyme activities, while a high protein diet increases the activities (McLean
and McLean, 1969; Nakajima et al., 1982). If this is the case, a low protein
diet should protect against chloroform hepatotoxicity by inhibiting the
enzymes responsible for chloroform metabolism. It was found, however, that
protein depletion did not alter the toxicity of chloroform in rats given a
single oral dose (McLean and McLean, 1969; McLean, 1970). If pretreated with
phenobarbital or NDDT to induce microsomal enzymes, rats maintained on a
standard diet "were no more susceptible to chloroform-induced liver damage
than were pretreated, protein-depleted rats (McLean, 1970).
More recently, Nakajima et al. (1982) studied the individual effects of
protein, fat, and carbohydrate on the metabolism of chloroform in relation to
its toxicity in male rats. Test diets were varied with respect to
carbohydrate, protein, or fat while maintaining isocaloric contents.
Microsomal enzymes were prepared and chloroform was added as a substrate.
The following results were obtained: decreased food intake increased liver
microsomal enzyme activities; decreased sucrose content in the diet increased
the metabolic rate; varying the protein and fat content, while holding the
sucrose content constant, had no effect on the metabolic rate; a
carbohydrate-free diet, which contained high protein and high fat,
accelerated the rate of chloroform metabolism almost as much as 1 day of food
deprivation. The authors concluded that it is a high carbohydrate content,
rather than a low protein content, which is responsible for the decreased
microsomal enzyme metabolism of chloroform and, hence, its toxicity.
5-59
-------
5.5. SUMMARY: CORRELATION OF EXPOSURE AND EFFECT
The purpose of this section is to delineate dose-response relationships
for the systemic toxicity of chloroform.
5.5.1. Effects of Acute Inhalation Exposure
The adverse effects on humans of inhaling high concentrations of
chloroform have been well documented in the course of its use as an
anesthetic. Studies that define the threshold region of exposure for such
effects in humans are, however, sparse at best. Experiments involving
subchronic exposure of several species of animals give some information on
toxicity thresholds for renal and hepatic effects, but little for CNS and
none for cardiovascular effects.
The only experimental studies conducted with humans (Lehmann and
Hasegawa, 1910; Lehmann and Schmidt-Kehl, 1936) involved relatively short
exposures and subjective responses. The results of these studies indicate
that the odor of chloroform can be perceived at about 200 ppm. Subjective
CNS effects (dizziness, vertigo) apparently did not occur at 390 ppm during a
30-minute exposure but were perceived at about 900 ppm after 2-3 minutes of
exposure. Subjects exposed to 1400 ppm for 30 minutes experienced tiredness
and headache in addition to the above CNS symptoms. The threshold for "light
intoxication" was about 4300 ppm (20 minutes). An exposure duration of
30 minutes or less is insufficient to achieve pulmonary steady state (or
total body equilibrium, Section 4.2.3). Hence, longer exposures at these
concentrations would be expected to cause more severe effects.
Chloroform concentrations used for the induction of anesthesia ranged
from about 20,000-40,000 ppm (NIOSH, 1974; Adriani, 1970) and for the
maintenance of anesthesia ranged from 1500 ppm (light anesthesia) to
15,000 ppm (deep anesthesia) (Goodman and Gilman, 1980). Continued exposure
5-60
-------
to 20,000 ppm could result in respiratory failure, direct depression of the
myocardium, and death (Section 5.1.1). Levels of exposure sufficient to
produce anesthesia have also caused cardiac arrhythmias and extrasystoles
(Kurtz et al., 1936; Orth et al., 1951) and hepatic necrosis and fatty
degeneration (Goodman and Gilman, 1980; Wood-Smith and Stewart, 1964).
Data from acute animal exposures tend to show similar CNS effects at
roughly the same levels of exposure that produced these effects in humans
(Lehmann and Flury, 1943). In addition, some data on the threshold for
hepatic effects have been obtained for mice. Kyi in et al. (1963), in
experiments with female mice of an unspecified strain, found that single,
4-hour exposures to chloroform produced mild hepatic effects (increased
incidence of moderate fatty infiltration) at 100 ppm. At 200 ppm, in
addition to fatty infiltration, hepatic necrosis and increased serum
ornithine carbamyl transferase activity occurred. (An elevation in serum
levels of this enzyme indicates liver damage according to Divincenzo and
Krasavage, 1974.) Further increases in fatty infiltration, necrosis, and
serum enzyme activity were observed at 400 and 800 ppm. These effects
appeared to be reversible because the extent of change was less severe 3 days
after exposure than it was 1 day after exposure.
Damage to the kidneys of male mice of sensitive strains (e.g., C3H) has
occurred at exposure levels as low as 5 mg/L (1025 ppm) for 1 hour (Deringer
et al., 1953). The damage consisted of necrosis of the epithelium of the
proximal tubules.
5.5.2. Effects of Acute Oral Exposure
Dose-response data for acute oral exposure of humans to chloroform are
limited to case reports. A fatal dose of as little as 1/3 ounce (10 ml) was
reported (Schroeder, 1965).
5-61
-------
A variety of dose-response data Is available for acute oral
administration of chloroform to animals. Single doses that were sufficient
to adversely affect kidney function (measured as excessive loss of glucose
and/or protein in the urine in male mice) ranged from 89-149 mg/kg in
sensitive and relatively insensitive strains (Hill, 1978). At single oral
doses of 1071 mg/kg, but not 756 or 546 mg/kg chloroform, increases in organ
weights and mild to moderate lesions were observed in the livers and kidneys
of Sprague-Dawley rats (Chu et al., 1982a). In male B6C3F1 mice, renal
necrosis occurred after 20 mg/kg, and focal tubular regeneration occurred
after 60 or 240 mg/kg but not after 15 mg/kg (Reitz et al., 1980). A low
observed adverse-effect level (LOAEL) for hepatic effects in mice can be
identified from the study of Jones et al. (1958), in which 30 mg/kg caused
midzonal fatty infiltration. Doses in the range of 133-355 mg/kg (Jones et
al., 1958; Reitz et al., 1980) represent a PEL (Frank-Effect-Level) for
hepatic damage (including centrilobular necrosis) in mice. According to
Torkelson et al. (1976), rats given "as little as" 250 mg/kg chloroform
"showed adverse effects" on liver and kidney as determined by gross
pathological examination. Reported oral LD^Q values for mice ranged from
119-1400 mg/kg, depending on sex, strain, and age (Kimura et al., 1971; Hill,
1978; Bowman et al., 1978). For rats, LD50 values of 908-2000 mg/kg have
been reported (Chu et al., 1980; Torkelson et al., 1976). The lethal dose
studies included both 24-hour and delayed deaths.
5.5.3. Effects of Dermal Exposure
Chloroform is irritating to the skin. It has been reported to cause
degenerative changes in the renal tubules of rabbits exposed dermally to high
doses under extreme conditions (1-3.98 g/kg body weight for 24 hours under an
5-62
-------
impermeable plastic cuff) (Torkelson et al., 1976). In humans, toxicity from
dermal exposure is probably not important in comparison with other routes.
5.5.4. Effects of Chronic Inhalation Exposure
Limited information on the effects of long-term intermittent exposure of
humans or animals to chloroform is available. A study involving a small
number of workers (Challen et al., 1958) indicates that long-term exposure to
20-71 ppm (98-346 mg/nr) for a 4-8 hour workday, with occasional brief
excursions to =1163 ppm, may represent a LOAEL for symptoms of CMS toxicity.
No evidence of liver damage or other organic lesion was detected by physical
examination and clinical chemistry tests. A single report linking liver
enlargement and viral hepatitis to occupational exposure to 10-200 ppm
chloroform (Bomski et al., 1967) is flawed by the apparent lack of suitable
controls. The available data do not define a NOAEL (no-observed-adverse-
effect-level) or NOEL (no-observed-effect-level) for humans.
Experiments with several species of animals (Torkelson et al., 1976)
give some information regarding the threshold region for hepatic and renal
effects of inhalation exposure to chloroform (Table 5-3). The animals were
exposed to chloroform 5 days/week for 6 months. Exposure to 25 ppm of
chloroform for up to 4 hours/day had no adverse effects in male rats as
judged by organ and body weights and probably the gross and microscopic
appearance of at least the liver and kidneys, although the authors were not
explicit about the latter. Exposure to 25 ppm for 7 hours/day, however,
produced histopathologic changes in the livers and kidneys of male but not
female rats. These changes were characterized as lobular granular
degeneration and focal necrosis throughout the liver and cloudy swelling of
the kidneys. The hepatic and renal effects appeared to be reversible because
rats exposed in the same way, but given a 6-week "recovery period" after the
5-63
-------
exposure period, appeared normal by the criteria tested. Increasingly
pronounced changes were observed in the livers and kidneys of both sexes of
rats exposed to 50 or 85 ppm. Hematologic, clinical chemistry, and
urinalysis values, tested at the two higher levels of exposure, were "within
normal limits."
Similar experiments with guinea pigs and rabbits gave somewhat
inconsistent results. Histopathological lesions were observed in liver and
kidneys of both species at 25 ppm but not at 50 ppm in either species and not
in guinea pigs even at 85 ppm (Torkelson et al., 1976).
The experiments of Torkelson et al. (1976) indicate that subchrom'c
exposure to 25 ppm (123 mg/m^), 4 hours/day, 5 days/week represents a NOAEL
and exposure to 25 ppm, 7 hours/day, 5 days/week respresents a LOAEL for
rats. Guinea pigs and rabbits may be slightly less sensitive.
5.5.5. Effects of Chronic Oral Exposure
Few dose-response data for oral exposure of humans to chloroform appear
to be available (Chapter 5). A single controlled study has been performed.
In this study, subjects were exposed to 70 or 178 mg of chloroform/day (=1
or 2.5 mg/kg/day assuming 70 kg body weight) for at least 1 year (DeSalva et
al., 1975). Neither liver runction tests nor blood urea nitrogen
determinations (a measure of kidney function) revealed statistically
significant differences between exposed and control subjects. Case reports
involving abuse of medicines containing chloroform (Wallace, 1950; Conlon,
1963) are not adequate for risk assessment because of the small numbers of
patients, exposure to other agents, and imprecise estimates of intake.
Subchrom'c and chronic toxicity experiments with rats, mice, and dogs,
when considered together (Table 5-4), do not clearly establish a NOAEL or
NOEL. Although no adverse effects were observed in four strains of mice
5-64
-------
given 17 mg/kg/day of chloroform, 6 days/week for 2 years (Roe et al., 1979),
at the lowest dose level tested (i.e., 15 mg/kg/day, 6 days/week for 7.5
years) in dogs, chloroform treatment was associated with an elevation in SGPT
in some but not all of the other tested clinical chemistry indices of hepatic
damage (Heywood et al., 1979). The livers of dogs treated with chloroform at
this dosage level had larger and more numerous "fatty cysts" than were found
in controls. These fatty cysts consisted of aggregations of vacuolated
histiocytes. No effect on survival, growth, organ weights, gross and
histological appearance of other organs, or hematologic or urinalysis values
was observed at this dosage level. Hence 15 mg/kg/day (6 days/week)
represents a LOAEL for dogs for effects on the liver. Chronic oral
administration of 60 mg/kg/day of chloroform (6 days/week) was associated
with slight hepatic changes in rats (Palmer et al., 1979) and with increased
incidences of moderate to severe renal disease in male mice of sensitive
strains (Roe et al., 1979).
None of the three species tested in long-term experiments appeared to be
markedly more sensitive to the toxicity of chloroform than any other; dogs
may have been slightly more sensitive. There were considerable differences
among strains of mice in the sensitivity of the males to chloroform
nephrotoxicity, as had also been observed in acute toxicity experiments.
As would be expected, dosages that produced little or no histologic or
clinical chemistry evidence of toxicity when given subchronically (15 and
30 mg/kg/day; rats, dogs) resulted in greater evidence of toxicity when given
for longer periods of time (Palmer et al., 1979; Heywood et al., 1979). The
response to chloroform in the long-term studies may have been modified by the
presence or absence of intercurrent respiratory and renal disease, but no
consistent pattern is obvious from an inspection of the data in Table 5-4.
5-65
-------
5.5.6. Target Organ Toxicit.y
Target organs characteristic of the acute toxicity of chloroform are the
central nervous system, liver, kidney, and heart. For chronic exposure to
chloroform, characteristic target organs are the liver and kidney, and
possibly the central nervous system. Some dose-response data are available
for the toxicity of chloroform to the liver, kidney, and central nervous
system; these data are summarized in Table 5-5 by target organ. The studies
from which these data are drawn are discussed more fully elsewhere in
Chapter 5, but a comparison on the basis of end point (target organ) was also
considered to be useful.
Manifestations of liver damage include centrilobular necrosis,
vacuolization, disappearance of glycogen, fatty degeneration and swelling
(Groger and Grey, 1979). In the kidney, chloroform exposure produces
necrosis of the proximal and distal convoluted tubules (Eschenbrenner and
Miller, 1945a). The mechanism by which chloroform produces these effects has
been extensively studied in experimental animals. From the studies
summarized in Section 5.3.1 (Brown et al., 1974; Ilett et a!., 1973; Docks
and Krishna, 1976; Ekstrom et al., 1981; Lavigne and Marchand, 1974; McMartin
et al., 1981; Masuda et al., 1980; Harris et al., 1982; Stevens and Anders,
1981), it appears that chloroform is first metabolized in the target organ by
microsomal drug metabolizing enzymes to a reactive intermediate, probably
phosgene, which in turn can react by various pathways, depending on
glutathione levels, one of which is the covalent binding to liver proteins
resulting in necrotic lesions. A similar mechanism may or may not occur in
the kidney (Kluwe and Hook, 1981).
5-66
-------
Table 5-5. TARGET ORGAN TOXICITY OF CHLOROFORM
Target
organ
Route and
type of
exposure
Species
Dose or
exposure
Effect on
target organ
Reference
1 iver
inhalation, acute
(surgical anesthesia)
human
en
i
liver
liver
inhalation, acute
inhalation, acute
liver inhalation, chronic
liver inhalation, chronic
liver inhalation, chronic
liver oral, acute
mice
mice
rats
rats
rats
mice
induction =
20,000-40,000 ppm x
a few minutes, plus
maintenance = 1500-
15,000 ppm x variable
duration
100 ppm x 4 hours,
single exposure
200 ppm x 4 hours,
single exposure
25 ppm, 4 hours/day,
5 days/week x
6 months
25 ppm, 7 hours/day,
5 days/week x
6 months
50 or 85 ppm, 7 hours/
day, 5 days/week x
6 months
30 mg/kg bw;
single dose
necrosis, fatty
degeneration in
some patients
fatty infiltration
necrosis, fatty
infiltration,
increase in SOCT
no effect
lobular granular
degeneration, focal
necrosi s
marked centrilobular
granular degeneration
fatty infiltration
NIOSH, 1974;
Goodman and
GiIman,
1980;
Wood-Smith
and Stewart,
1964
Kyi in et al.,
1963
Kyi in et al.,
1963
Torkelson
et al., 1976
Torkelson
et al., 1976
Torkelson
et al., 1976
Jones et al.,
1958
-------
Table 5-5. (continued)
Route and
Target type of
organ exposure Species
liver oral, acute mice
liver oral, acute mice
liver oral, acute mice
en
i
Oi
00 liver oral, acute mice
liver oral (drinking water) rats
subchronic
liver oral (drinking water) mice
subchronic
liver oral (drinking water) mice
subchronic
Dose or
exposure
133 mg/kg, single
dose
355 mg/kg, single
dose
60 mg/kg, single
dose
240 mg/kg, single
dose
20, 38, 57, 81, or
160 mg/kg/day x
90 days
64 or 97 mg/kg/day
x 90 days
145 or 190 mg/kg/day
x 90 days
Effect on
target organ
massive fatty
infiltration and
severe necrosis
massive fatty
infiltration and
severe necrosis
no effect
hepatocellular necrosis
and swelling; inflamma-
tion
transient hepatosis
(at 30 and 60, but
not at 90 days)
transient centrilobular
fatty change (at 30
and 60 but not at
90 days)
fatty change
Reference
Jones et al . ,
1958
Jones et al . ,
1958
Reitz et al . ,
1980
Reitz et al . ,
1980
Jorgenson
and
Rushbrook,
1980
Jorgenson
and
Rushbrook,
1980
Jorgenson
and
Rushbrook,
1980
-------
Table 5-5. (continued)
(Jl
CTi
Target
organ
liver
liver
liver
liver
Route and
type of
exposure
oral (gavage)
subchronic
oral (capsule),
subchronic
oral (capsule),
subchronic
oral (capsule),
Species
rats
dogs
dogs
dogs
Dose or
exposure
410 mg/kg/day, 6 days/
week x 13 weeks
30 mg/kg/day, 7 days/
week x 13 weeks
45 mg/kg/day, 7 days/
week x 13 weeks
60 mg/kg/day, 7 days/
Effect on
target organ
fatty change and
necrosi s
no effect
slight fatty change
fatty degeneration,
Reference
Palmer et al . ,
1979
Heywood
et al., 1979
Heywood
et al., 1979
Heywood
liver
subchronic
oral (capsule),
subchronic
liver oral (gavage),
chronic
liver oral (gavage),
chronic
liver oral (capsule),
chronic
dogs
rats
mice
dogs
week x 18 weeks
120 mg/kg/day, 7 days/
week x 12 weeks
60 mg/kg/day, 6 days/
week x 80 weeks
60 mg/kg/day, 6 days/
week x 80 weeks
15 or 30 mg/kg/day,
6 days/week x 7.5
years
increase in SGOT and
SGPT
fatty degeneration,
jaundice, increase in
SGOT, SGPT, bilirubin
minor histological
changes and decrease
in relative liver weight
no effect
et al., 1979
Heywood
et al., 1979
Palmer et al.,
1977
increases in SGPT and
other serum indicators
of hepatic damage,
increase in size and
number of fatty cysts
(vacuolated histiocytes)
Roe et al.,
1979
Heywood et
al., 1979
-------
Table 5-5. (continued)
Target
organ
Route and
type of
exposure
Dose or
Species exposure
Effect on
target organ
Reference
en
i
liver oral, chronic
liver dermal, acute
kidney inhalation, acute
kidney inhalation, chronic
kidney oral, acute
kidney oral, acute
kidney oral, acute
kidney oral, acute
humans
rabbits
mice, males
of sensitive
strains
rats
mice, males
of sensitive
strains
mice, males
of sensitive
strains
2.5 mg/kg/day for
^1 year
3.98 g/kg x 24 hours
under plastic cuff,
single exposure
5000 ppm, 1 hour,
single exposure
25, 50, or 85 ppm,
7 hours/day, 5 days/
week x 6 months
89 mg/kg, single dose
no effect
no macroscopic
pathologic changes
necrosis and calcifi-
fication of tubular
epithelium
cloudy swelling of
tubular epithelium
loss of glucose or
protein in urine
149 mg/kg, single dose loss of glucose or
protein in urine
mice, male 15 mg/kg, single dose no effect
mice, male 60 mg/kg, single dose
focal tubular
epithelial
regeneration
De Salva
et al., 1975
Torkelson
et al., 1976
Deringer
et al., 1953
Torkelson
et al., 1976
Hill, 1978
Hill, 1978
Reitz et al.,
1980
Reitz et al.,
1980
-------
Table 5-5. (continued)
Target
organ
kidney
kidney
kidney
kidney
kidney
Route and
type of
exposure Species
oral, acute mice, male
oral (drinking water) rats
subchronic
oral (drinking water) rats
subchronic
oral (drinking water) mice
subchronic
oral (capsule), dogs
subchronic
Dose or
exposure
240 mg/kg, single dose
160 mg/kg/day x
90 days
= 300 mg/kg/day x
90 days
290 mg/kg/day x
90 days
120 mg/kg/day,
7 days/week x
12 weeks
Effect on
target organ
severe diffuse
cortical necrosis,
focal tubular
epithelial regeneration
no effect
no effect
no effect
no effect
Reference
Reitz et al . ,
1980
Jorgenson
and
Rushbrook,
1980
Chu et al . ,
1980b
Jorgenson
and
Rushbrook,
1980
Jorgenson
and
Rushbrook,
1980
kidney oral, chronic
kidney oral (gavage),
chronic
humans
rats
2.5 mg/kg/day,
7 days/week, for
>1 year
200 mg/kg/day,
5 days/week x
78 weeks
no effect on BUN
no effect
De Salva
et al., 1975
NCI, 1976
-------
Table 5-5. (continued)
en
l
^-j
Target
organ
kidney
kidney
kidney
kidney
kidney
Route and
type of
exposure
oral (gavage),
chronic
oral (gavage),
chronic
oral (gavage)
chronic,
oral (gavage)
chronic
oral (capsule),
chronic
Species
mice
mice, males
of sensitive
strains
mice, males
of sensitive
strains
mice, males
of insensi-
tive strains,
females
dog
Dose or
exposure
138 or 227 mg/kg/day,
5 days/week x
78 weeks
17 mg/kg/day,
6 days/week x
80 weeks
60 mg/kg/day,
6 days/week x
80 weeks
60 mg/kg/day,
6 days/week x
80 weeks
15 mg/kg/day
6 days/week x
Effect on
target organ
decreased incidence
of renal disease
no effect
increased incidence
of moderate to severe
renal disease
no effect
no effect
Reference
NCI, 1976
Roe et al . ,
1979
Roe et al . ,
1979
Roe et al . ,
1979
Heywood et
al., 1979
kidney oral (capsule),
chronic
kidney dermal, acute
dog
rabbits
7.5 years
30 mg/kg/day,
6 days/week x
7.5 years
1.0, 2.0, and
3.98 g/kg x 24 hours
under plastic cuff,
single exposure
increase in fat
deposition in
glomeruli
degenerative
changes in tubules
Heywood
et al., 1979
Torkelson
et al., 1976
-------
Table 5-5. (continued)
Target
organ
Route and
type of
exposure
Species
Dose or
exposure
Effect on
target organ
Reference
en
CO
central
nervous
system
(CNS)
CNS
inhalation, acute
humans
CNS
CNS
CNS
CNS
inhalation, acute
inhalation, acute
inhalation, acute
inhalation, acute
inhalation, acute
humans
humans
humans
humans
mice
900-1400 ppm for
^30 minutes, single
exposure
dizziness, tiredness,
headache
4300-5100 ppm x
20 minutes, single
exposure
1500-2000 ppm,
single exposure
15,000 ppm,
single exposure
20,000-40,000 ppm x
a few minutes,
single exposure
2500 ppm x 12 hours,
single exposure
dizziness, light
intoxication
maintenance of light
anesthesia (after
induction)
maintenance of heavy
anesthesia (after
induction)
induction of
anesthesia
no obvious effects
Lehmann
and
Hasegawa,
1910;
Lehmann and
Schmidt-Kehl,
1936
Lehmann and
Hasegawa,
1910;
Lehmann and
Schmidt-Kehl,
1936
Goodman and
Gillman, 1980
Goodman and
Gillman, 1980
NIOSH, 1974
Adriani, 1970
Lehmann and
Flury, 1943
-------
Table 5-5. (continued)
en
—i
Target
organ
CNS
CNS
CNS
CNS
CNS
CNS
CNS
CNS
Route and
type of
exposure
inhalation, acute
inhalation, acute
inhalation, acute
inhalation, acute
inhalation, acute
inhalation, acute
inhalation, acute
inhalation, chronic
Species
mice
mice
cats
cats
cats
cats
cats
humans
Dose or
exposure
3100 ppm x 1 hour,
single exposure
4100 ppm x 0.5 hours,
single exposure
7200 or 21,500 ppm x
5 minutes, single
exposure
7200 ppm x 60 min
single exposure
7200 ppm x 93 min
single exposure
21,500 ppm x 10 min
single exposure
21,500 ppm x 13 min
single exposure
20-71 ppm (with
Effect on
target organ
light narcosis
deep narcosis
disturbance of
equilibrium
light narcosis
deep narcosis
light narcosis
deep narcosis
tiredness
Reference
Lehmann and
Flury, 1943
Lehmann and
Flury, 1943
Lehmann and
Flury, 1943
Lehmann and
Flury, 1943
Lehmann and
Flury, 1943
Lehmann and
Flury, 1943
Lehmann and
Flury, 1943
Challen
excursions to
1163 ppm lasting
1.5-2 min) for
4-8 hours/day,
5 days/week
et al., 1958
-------
Table 5-5. (continued)
Target
organ
Route and
type of
exposure
Species
Dose or
exposure
Effect on
target organ
Reference
CNS
inhalation, chronic
humans
en
CNS
CNS
CNS
CNS
CNS
oral, acute
rats
oral (drinking water), rats
subchronic
oral (drinking water), mice
subchronic
oral (drinking water) rats
subchronic
oral (gavage), rats
subchronic
77 to 237 ppm
(with excursions
to =1163 ppm
lasting 1.5-2 min)
for 4-8 hours/day,
5 days/week
350 mg/kg
single dose
20-160 mg/kg/day x
90 days
32-290 mg/kg/day x
90 days
= 300 mg/kg/day x
90 days
60 mg/kg/day
6 days/week x
80 weeks
tiredness, depression,
occasional silliness
or staggering during
the workday
minimum narcotic dose
(MND5Q)
dose-related signs
of depression during
1st week only
dose-related signs of
depression during 1st
week only
no histopathologic
changes in brain
no effect on gross
or histologic
appearance of brain
Challen
et al., 1958
Jones et al.,
1958
Jorgenson
and
Rushbrook,
1980
Jorgenson
and
Rushbrook,
1980
Chu et al.,
1982b
Palmer et al.,
1977
-------
5.5.7. Factors that Modify the Toxicity of Chloroform
Several substances alter the toxicity of chloroform, most probably by
modifying the metabolism of chloroform to a reactive intermediate (see
Section 5.4). These substances are of interest because humans may be
accidentally or intentionally exposed to them. Factors that potentiate the
toxic effects induced by exposure to chloroform include ethanol (Kutob and
Plaa, 1962; Sato et al., 1980, 1981), polybrominated biphenyls (Kluwe and
Hook, 1978), steroids (Clemens et al., 1979), and ketones (Hewitt et al.,
1979; Jernigan and Harbison, 1982; Branchflower and Pohl, 1981). Disulfiram
and its metabolites (Scholler et al., 1970; Masuda and Nakayama, 1982;
Gopinath-and Ford, 1975) and high carbohydrate diets (Nakajima et al., 1982),
appear to protect against chloroform toxicity.
5-76
-------
5.6 REFERENCES FOR CHAPTER 5
Abdelsalam, E.B.; Adam, S.E.I.; Tartour, G. (1982) Modification of the
hepatotoxicity of carbon tetrachloride and chloroform in goats by
pretreatment with dieldrin and phenobarbitone. Zbl. Vat. Med. A.
29:142-148.
Adriani, J. (1970) The pharmacology of anesthetic drugs. Springfield,
IL: Charles C. Thomas; pp. 57-60.
Ahmadizadeh, M.; Echt, R.; Kuo, C.; Hook, J.T. (1984a) Sex and strain
differences in mouse kidney: Bowman's capsule morphology and
susceptibility to chloroform. Toxicol. Lett. 20:161-172.
Ahmadizadeh, M.; Kuo, C.H.; Echt, R.; Hook, J.B. (1984b) Effects of
polybrominated biphenyls, beta-naphthoflavone and phenobarbital on aryl
hydrocarbon hydroxylase activities and chloroform-induced nephrotoxicity
and hepatotoxicity in male C57BL/6J and DBA/2J mice. Toxicology 31:343-
352.
Bailie, M.B.; Smith, J.F.; Newton, J.H.; Hook, J.B. (1984) Mechanism of
chloroform nephrotoxicity. IV Phenobarbital potentiation of in vivo
chloroform metabolism and toxicity in rabbit kidneys. Toxicol. Appl.
Pharmacol. 74:285-292.
Bomski, H.; Sobolweska, A.; Strakowski, A. (1967) [Toxic damage of the
liver by chloroform in chemical industry workers.] Arch. Gewerbepathol.
Gewerbehy. 24:127-134. (Ger.) (Cited in Utidjian, 1976; NIOSH, 1974.)
Bowman, F.J.; Borzelleca, J.; Munson, A.E. (1978) The toxicity of some
halomethanes in mice. Toxicol. Appl. Pharmacol. 44:213-215.
Branchflower, R.V.; Pohl, L.R. (1981) Investigation of the mechanism of the
potentiation of chloroform-induced hepatotoxicity and nephrotoxicity by
methyl, n-butyl ketone. Toxicol. Appl. Pharmacol. 61:407-413.
Branchflower, R.V.; Schulick, R.D.; George, J.W.; Pohl, L.R. (1983)
Comparison of the effects of methyl-N-butylketone and phenobarbital on
rat liver cytochromes P-450 and the metabolism of chloroform to
phosgene. Toxicol. Appl. Pharmacol. 71:414-421.
Branchflower, R.V.; Nunn, D.S.; Highet, R.J.; Smith, J.H.; Hook, J.B.;
Pohl, L.R. (1984) Nephrotoxicity of chloroform: metabolism to
phosgene by the mouse kidney. Toxicol. Appl. Pharmacol. 72:159-168.
Brown, B.R., Jr.; Sipes, I.G.; Sagalyn, A.M. (1974) Mechanisms of acute
hepatic toxicity: chloroform, halothane and glutathione.
Anesthesiology 41:554-561.
5-77
-------
Butler, T.C. (1961) Reduction of carbon tetrachloride in vivo and reduction
of carbon tetrachloride and chloroform i_n vitro by tissues and tissue
constituents. J. Pharmacol. Exp. Ther. 134:311-319.
Challen, P.J.R.; Hickish, D.E.; Bedford, J. (1958) Chronic chloroform
intoxication. Br. J. Ind. Med. 15:243-249.
Chu, I.; Secours, V.; Marino, I.; Villeneuve, D.C. (1980) The acute
toxicity of four trihalomethanes in male and female rats. Toxicol.
Appl. Pharmacol. 52:351-353.
Chu, I.; Villeneuve, D.C.; Secours, V.E.; Becking, G.C. (1982a) Toxicity of
trihalomethanes: I. The acute and subacute toxicity of chloroform,
bromodichloromethane, chlorodibromomethane and bromoform in rats.
J. Environ. Sci. Health. B 17:205-224.
Chu, I.; Villeneuve, D.C.; Secours, V.E.; Becking, G.G.; Valli, V.E. (1982b)
Toxicity of trihalomethanes: II. Reversibility of toxicological
changes produced by chloroform, bromodichloromethane,
chlorodibromomethane and bromoform in rats. J. Environ. Sci. Health.
B 17:225-240.
Cianflone, D.J.; Hewitt, W.R.; Villeneuve, D.C.; Plaa, G.L. (1980) Role of
biotransformation in the alterations of chloroform hepatotoxicity
produced by Kepone and Mirex. Toxicol. Appl. Pharmacol. 53:140-149.
Clemens, T.L.; Hill, R.N.; Bullock, L.P.; Johnson, W.D.; Sulatos, L.G.;
Vesell, E.S. (1979) Chloroform toxicity in the mouse: role of genetic
factors and steroids. Toxicol. Appl. Pharmacol. 48:117-130.
Condie, L.W.; Smallwood, C.L.; Laurie, R.D. (1982) Comparative renal and
hepatotoxicity of halomethanes: bromodichloromethane, bromoform,
chloroform, dibromochloromethane and methylene chloride. Drug Chem.
Toxicol. 6:563-578.
Conlon, M.F. (1963) Addiction to chlorodyne. Br. Med. J. 2:1177-1178.
Cowlen, M.S.; Hewitt, W. R.; Schroeder, F. (1984) Mechanisms in 2-hexanone
potentiation of chloroform hepatotoxicity. Toxicol. Lett. 22:293-299.
Culliford, D.; Hewitt, H.B. (1957) The influence of sex hormone status on
the susceptibility of mice to chloroform-induced necrosis of the renal
tubules. J. Endocrinol. 14:381-393.
Danni, 0.; Brossa, 0.; Burdino, E.; Milillo, P.; Ugazio, G. (1981) Toxicity
of halogenated hydrocarbons in pretreated rats: an experimental model
for the study of integrated permissible limits of environmental poisons.
Int. Arch. Occup. Environ. Health. 49:105-112.
Deringer, M.K.; Dunn, T.B.; Heston, W.E. (1953) Results of exposure of
strain C3H mice to chloroform. Proc. Soc. Exp. Biol. Med. 83:474-479.
5-78
-------
DeSalva, S.; Volpe, A.; Leigh, G.; Regan, T. (1975) Long-term safety
studies of a chloroform-containing dentifrice and mouth-rinse in man.
Food Cosmet. Toxicol. 13:529.
Divincenzo, G.D.; Krasavage, W.J. (1974) Serum ornithine carbamyl
transferase as liver response test for exposure to organic solvents.
Am. Ind. Hyg. Assoc. J. 35:21-29.
Docks, E.L.; Krishna, G. (1976) The role of glutathione in chloroform-
induced hepatotoxicity. Exp. Mol. Pathol. 24:13-22.
Ekstrom, T.; Hoegbers, J.; Jernstroem, B. (1982) Inhibition of hepatic,
gamma-glutamyl-cysteine synthetase by chloroform. Biochem. Pharmacol.
31:3510-3513.
Eschenbrenner, A.B.; Miller, E. (1945a) Induction of hepatomas in mice by
repeated oral administration of chloroform, with observations on sex
differences. J. Natl. Cancer Inst. 5:251-255.
Eschenbrenner, A.B.; Miller, E. (1945b) Sex differences in kidney
morphology and chloroform necrosis. Science 102:302-303.
Goodman, L.S.; Gilman, A. (1980) The pharmacological basis of therapeutics.
6th ed. New York: MacMillan and Co.
Gopinath, C.; Ford, E.J.H. (1975) The role of microsomal hydroxylases in
the modification of chloroform hepatotoxicity in rats. Br. J. Exp.
Pathol. 56:412-422.
Groger, W.K.L.; Grey, T.F. (1979) Effect of chloroform on the activities of
liver enzymes in rats. Toxicology 14:23-28.
Harris, R.N.; Ratnayake, J.H.; Garry, V.F.; Anders, M.W. (1982) Interactive
hepatotoxicity of chloroform and carbon tetrachloride. Toxicol. Appl.
Pharmacol. 63:281-291.
Heller, H.; Smirk, F.H. (1932) The influence of anaesthetics and
hypnotics on the absorption and excretion of water. J. Physiol.
(London) 76:292-302.
Hewitt, W.R.; Miyajima, H.; Cote, M.G.; Plaa, G.C. (1979) Acute alteration
of chloroform-induced hepato- and nephrotoxicity by mirex and Kepone.
Toxicol. Appl. Pharmacol. 48:509-527.
Hewitt, W.R.; Mujajima, H.; Cote, M.G.; Plaa, G.L. (1980) Acute alteration
of chloroform-induced hepato- and nephrotoxicity by n-hexane, methyl
n-butyl ketone and 2,5-hexanedione. Toxicol. Appl. Pharmacol.
53:230-248.
Hewitt, W.R.; Brown, E.M.; Plaa, G.L. (1983a) Relationship between the
carbon skeleton length of ketonic solvents and potentiation of
chloroform-induced hepatotoxicity in rats. Toxicol. Lett. 16:297-304.
5-79
-------
Hewitt, L.A.; Hewitt, W.R.; Plaa, G.L. (I983b) Fractional hepatic
localization of 14CHC13 in mice and rats treated with chlordecone or
micrex. Fund. Appl. Toxicol. 3:489-495.
Heywood, R.; Sortwell, R.J.; Noel, P.R.B.; Street, A.E.; Prentice, D.E.; Roe,
F.J.C.; Wardsworth, P.F.; Worden, A.N.; Van Abbe, N.J. (1979) Safety
evaluation of toothpaste containing chloroform. III. Long-term study
in beagle dogs. J. Environ. Pathol. Toxicol. 2:835-851.
Hill, R.N. (1978) Differential toxicity of chloroform in the mouse.
Ann. NY Acad. Sci. 298:170-175.
Hjelle, J.J.; Gordon, A.S.; Peterson, D.R. (1982) Studies on carbon tetra-
chloride-ethanol interactions in mice. Toxicol. Lett. 10:17-24.
IARC (International Agency for Research on Cancer). (1976) Disulfirams.
In: Some carbamates, thiocarbamates and carbazides. IARC Monographs on
the Evaluation of the Carcinogenic Risk of Chemicals to Man. Lyon,
France: WHO, IARC; pp. 85-96.
lijima, M.; Cote, M.G.; Plaa, G.L. (1983) A semiquantitative morphologic
assessment of chlorodecone-potentiated chloroform hepatotoxicity.
Toxicol. Lett. 17:307-314.
Ilett, K.F.; Reid, W.D.; Sipes, I.G.; Krishna, G. (1973) Chloroform
toxicity in mice: Correlation of renal and hepatic necrosis with
covalent binding of metabolites to tissue macromolecules. Exp. Mol.
Pathol. 19:215-229.
Jernigan, J.D.; Harbison, R.D. (1982) Role of biotransformation in the
potentiation of halocarbon hepatotoxicity by 2,5-hexanedione.
J. Toxicol. Environ. Health 9:761-781.
Jones, W.M.; Margolis, G.; Stephen, C.R. (1958) Hepatotoxicity of
inhalation anesthetic drugs. Anesthesiology 19:715.
Jorgenson, T.A.; Rushbrook, C.J. (1980) Effects of chloroform in the
drinking water of rats and mice: ninety-day subacute toxicity study.
EPA-600/1-80-030. Menlo Park, CA: SRI International. Available as
PB80-219108. Springfield, VA: National Technical Information Service.
Kaminsky, L.S.; Piper, L.J.; MacMartin, D.N.; Fasco, M.S. (1978) Induction
of hepatic microsomal cytochrome P-450 by mirex and kepone. Toxicol.
Appl. Pharmacol. 12:131-139.
Kimura, E.T.; Ebert, D.M.; Dodge, P.W. (1971) Acute toxicity and limits of
solvent residue for sixteen organic solvents. Toxicol. Appl. Pharmacol.
19:699-704.
Klaassen, C.D.; Plaa, G.L. (1966) Relative effect of various chlorinated
hydrocarbons on liver and kidney function in mice. Toxicol. Appl.
Pharmacol. 9:139-151.
5-80
-------
Kluwe, W.M.; Hook, J.B. (1978) Polybrominated biphenyl-induced potentiation
of chloroform toxicity. Toxlcol. Appl. Pharmacol. 45:861-869.
Kluwe, W.M.; Hook, J.B. (1981) Potentiation of acute chloroform nephro-
toxicity by the glutathione depletion diethyl maleate and protection by
the microsomal enzyme inhibitor piperonyl butoxide. Toxicol. Appl.
Pharmacol. 59:457-466.
Knocker, P.; Mandelstam, J. (1944) Rhythmic changes in the rat produced by
a single injection of chloroform. Nature (London) 154:148-149.
Krantz, J.C., Jr.; Carr, C.J. (1965) Pharmacologic principles of medical
practice. 6th ed. Baltimore: Williams and Wilkins.
Krus, S.; Zaleska-Rutczynski, Z. (1970) Morphological counterparts of the
genetically determined resistance of mice to chloroform. Experientia
26: 101-102.
Kurtz, C.M.; Bennett, J.H.; Shapiro, H.H. (1936) Electrocardiographic
studies during surgical anesthesia. J. Am. Med. Assoc. 106:434-441.
Kutob, S.D.; Plaa, G.L. (1962a) The effect of acute ethanol intoxication on
chloroform-induced liver damage. J. Pharmacol. Exp. Teratol.
135:245-251.
Kutob, S.D.; Plaa, G.L. (1962b) A procedure of estimating the
hepatotoxicity potential of certain industrial solvents. Toxicol. Appl.
Pharmacol. 4:354.
Kylin, B.; Reichard, H.; Sumegi, I.; Yllner, S. (1963) Hepatotoxicity of
inhaled trichloroethylene, tetrachloroethylene and chloroform. Single
exposure. Acta Pharmacol. Toxicol. 20:16-26.
Lavigne, J.G.; Marchand, C. (1974) The role of metabolism in chloroform
hepatotoxicity. Toxicol. Appl. Pharmacol. 29:312-326.
Lavigne, J.G.; Belanger, P.M.; Dore, F.; Labrecque, G. (1983) Temporal
variations in chloroform-induced hepatotoxicity in rats. Toxicology
26:267-273.
Lehmann, K.B.; Flury, F.F. (1943) Chlorinated hydrocarbons. In: Lehmann,
K.B.; Flury, F.F., eds. Toxicology and hygiene of industrial solvents.
Baltimore: Williams and Wilkins Co.; pp. 138-145, 191-196.
Lehmann, K.B.; Hasewaga. (1910) [Studies of the absorption of chlorinated
hydrocarbons in animals and humans.] Arch. Hyg. 72:327-342. (Ger.)
(Cited in NIOSH, 1974.)
Lehmann, K.B.; Schmidt-Kehl, L. (1936) [The thirteen most important chlor-
inated aliphatic hydrocarbons from the standpoint of industrial
hygiene.] Arch. Hyg. 116:131-200. (Ger.) (Cited in NIOSH, 1974.)
Lucas, G.H.W. (1928) A study of the fate and toxicity of bromine and
chlorine containing anesthetics. J. Pharmacol. Exp. Ther. 34:223.
5-81
-------
Mansuy, D.; Beaune, P.; Cresteil, T.; Lange, M.; Lerous, J.P. (1977)
Evidence for phosgene formation during liver microsomal oxidation of
chloroform. Biochem. Biophys. Res. Commun. 79:513-517.
Masuda, Y.; Nakayama, N. (1982) Protective effect of diethyl-
dithiocarbamate and carbon disulfide against liver injury induced by
various hepatotoxic agents. Biochem. Pharmacol. 31:2713-2725.
Masuda, Y.; Nakayama, N. (1983) Protective action of diethyl-
dithiocarbonate and carbon disulfide against renal injury induced by
chloroform in mice. Biochem. Pharmacol. 32:3127-3135.
Masuda, Y.; Yano, I.; Murano, T. (1980) Comparative studies on the
hepatotoxic actions of chloroform and related halomethanes in normal and
phenobarbital-pretreated animals. J. Pharmacobio-Dyn. 3:53-64.
McLean, A.E.M.; McLean, E.K. (1969) Diet and toxicity. Br. Med. Bull.
25:278-281.
McLean, A.E.M. (1970) The effect of protein deficiency and microsomal
enzyme induction by DDT and phenobarbitone on the acute toxicity of
chloroform and a pyrrolizidiue alkaloid, retrorsine. Br. J. Exp.
Pathol. 51:317-321.
McMartin, N.D.; O'Connor, Jr., J.A.; Kaminsky, L.S. (1981) Effects of
differential changes in rat hepatic and renal cytochrome P-450
concentrations on hepatotoxicity and nephrotoxicity of chloroform. Res.
Commun. Chem. Pathol. Pharmacol. 31:99-110.
Nakajima, T.; Koyama, Y.; Sato, A. (1982) Dietary modification of
metabolism and toxicity of chemical substances with specific reference
to carbohydrate. Biochem. Pharmacol. 31:1005-1011.
National Cancer Institute (NCI). (1976) Report on carcinogenesis bioassay
of chloroform. Available as PB-264-018. Springfield, VA: National
Technical Information Service.
National Institute for Occupational Safety and Health (NIOSH). (1974)
Criteria for a recommended standard - occupational exposure to
chloroform. Available as PB-246-695. Springfield, VA: National
Technical Information Service.
Oettel, H. (1936) Einwirkung organische Flussigkeiten auf die haut. Arch.
Exp. Pathol. Pharmacol. 183:641-96.
Orth, O.S.; Liebenow, R.R.; Capps, R.T. (1951) III-The effect of chloroform
on the cardiovascular system. In: Waters, R.M., ed. Chloroform - a
study after 100 years. Madison: University of Wisconsin Press;
pp. 39-75.
Palmer, A.K.; Street, A.E.; Roe, F.J.C.; Worden, A.M.; Van Abbe, N.J.
(1979) Safety evaluation of toothpaste containing chloroform. II.
Long-term studies in rats. J. Environ. Pathol. Toxicol. 2:821-833.
5-82
-------
Paul, B.B.; Rubinstein, D. (1963) Metabolism of carbon tetrachloride and
chloroform by the rat. J. Pharmacol. Exp. Ther. 141:141-148.
Pereira, M.A.; Savage, R.E.; Guion, C.W.; Wernsing, P.A. (1984) Effect of
chloroform on hepatic and renal DNA synthesis and ornithine
decarboxylase activity in mice and rats. Toxicol. Lett. 21:357-364.
Phoon, W.H.; Goh, K.T.; Lee, L.T.; Tan, K.T.; Kwok, S.E. (1983) Toxic
jaundice from occupational exposure to chloroform. Med. J. Malaysia
30:31-34.
Piersol, G.M.; Tumen, A.J.; Kau, L.S. (1933) Fatal poisoning following the
ingestion of chloroform. Med. Clin. North Am. 17:587-601.
Pohl, L.R.; Bhooshan, B.; Whittaker, N.F.; Krishna, G. (1977) Phosgene: a
metabolite of chloroform. Biochem. Biophys. Res. Commun. 79:684-691.
Pohl, L.R. (1979) Biochemical toxicology of chloroform. Rev. Biochem.
Toxicol. 1:79-107.
Pohl, L.R.; George, J.W.; Satch, H. (1984) Strain and sex differences in
chloroform-induced nephrotoxicity. Different rate of metabolism of
chloroform to phosgene by the mouse kidney. Drug Metab. Dispos. 12:304-
308.
Reid, W.D.; Krishna, G. (1973) Centrilobular hepatic necrosis related to
covalent binding of metabolites of halogenated aromatic hydrocarbons.
Exp. Mol. Pathol. 18:80-95.
Reitz, R.H.; Quast, J.F.; Stott, W.T.; Watanabe, P.G.; Gehring, P.J. (1980)
Pharmacokinetics and macromolecular effects of chloroform in rats and
mice: implications for carcinogenic risk estimation. Water
Chlorination: Environ. Impact Health Eff. 3:983-993.
Reynolds, E.S.; Yee, A.G. (1967) Liver parenchymal cell injury. V.
Relationships between patterns of chloromethane ^4C incorporation into
constituents of liver i_n vivo and cellular injury. Lab. Invest.
16:591-603.
Roe, F.J.C.; Palmer, A.K.; Worden, A.N.; Van Abbe, N.J. (1979) Safety
evaluation of toothpaste containing chloroform. I. Long-term studies
in mice. J. Environ. Toxicol. 2:799-819.
Sato, A.; Nakajima, T. (1984) Dietary carbohydrate induced and ethanol-
induced alteration of the metabolism and toxicity of chemical
substances. Nutr. Cancer 6:121-132.
Sato, A.; Nakajima, T.; Koyama, Y. (1980) Effects of chronic ethanol
consumption on hepatic metabolism of aromatic and chlorinated
hydrocarbons in rats. Br. J. Indust. Med. 37:382-386.
Sato, A.; Nakajima, T.; Koyama, Y. (1981) Dose-related effect of a single
dose of ethanol on the metabolism in rat liver of some aromatic and
chlorinated hydrocarbons. Toxicol. Appl. Pharmacol. 60:8-15.
5-83
-------
Sax, N.I. (1979) Dangerous properties of industrial materials. 5th ed.
New York: Van Nostrand Reinhold Company, Inc.; p. 193.
Savage, R.E., Jr.; Westrich, C.; Guin, C.; Pereira, M.A. (1982) Chloroform
induction of ornithine decarboxylase activity in rats. Environ. Health
Persp. 46:157-162.
Scholler, K.L. (1966) Elektronmikroskopische untersuchungen an leberzellen
der Ratte nach narkosen mit verschiedenen inhlalationsnarkotica. Der
Anaesthetist. 15:145-148. (Ger).
Scholler, K.L. (1967) Effect of haloethane and chloroform on the fine
structure and protein synthesis of rat liver. Experientia 23:652-655.
(In German).
Scholler, K.L. (1970) Modification of the effects of chloroform on the rat
liver. Br. J. Anaesth. 42:603-605.
Schroeder, H.G. (1965) Acute and delayed chloroform poisoning. Br. J.
Anaesth. 37:972-975.
Sipes, I.G.; Krishna, G.; Gillette, J.R. (1977) Bioactivation of carbon
tetrachloride, chloroform, and bromotrichloromethane: role of
cytochrome P-450. Life Sci. 20:1541-1548.
Smith, J.H.; Hook, J.B. (1983) Mechanism of chloroform nephrotoxicity.
II. in vivo evidence for renal metabolism of chloroform in mice.
Toxicol. Appl. Pharmacol. 70:480-485.
Smith, J.H.; Hook, J.B. (1984) Mechanism of chloroform nephrotoxicity.
III. Renal and hepatic microsomal metabolism of chloroform in mice.
Toxicol. Appl. Pharmacol. 73:511-524.
Smith H.J.; Malta, K.; Sleight, S.D.; Hook, B. (1983) Mechanism of
chloroform nephrotoxicity. I. Time course of chloroform toxicity in
male and female mice. Toxicol. Appl. Pharmacol. 70:467-479.
Smith, J.H.; Malta, K.; Sleight, S.D.; Hook, J.B. (1984) Effect of sex
hormone status on chloroform nephrotoxicity and renal mixed-function
oxidases in mice. Toxicology 30:305-316.
Stevens, J.L.; Anders, M.W. (1981) Effect of cysteine, diethyl maleate, and
phenobarbital treatment on the hepatotoxicity of (1H)- and
(2H)-chloroform. Ch.em. Biol. Interact. 37:207-217.
Storms, W.W. (1973) Chloroform parties. J. Am. Med. Assoc. 225:160.
Testai, E.; Vittozzi, L. (1984) Different pathways of chloroform
metabolism. Arch. Toxicol., Suppl. 7:278-281.
Torkelson, T.R.; Oyen, F.; Rowe, V.K. (1976) The toxicity of chloroform as
determined by single and repeated exposure of laboratory animals. Am.
Ind. Hyg. Assoc. J. 37:697-705.
5-84
-------
Utidjian, H.M.D. (1976) Recommendations for a chloroform standard.
J. Occupat. Med. 18:253.
Van Dyke, R.A.; Chenoweth, M.B.; Poznak, A.V. (1964) Metabolism of volatile
anesthetics. I. Conversion in vivo of several anesthetics to 14C02 and
chloride. Biochem. Pharmacol. 13:1239-1247.
von Oettingen, W.F. (1964) The halogenated hydrocarbons of industrial and
toxicological importance. Amsterdam: Elsevier.
Wallace, C.J. (1950) Hepatitis and nephrosis due to cough syrup containing
chloroform. Calif. Med. 73:442.
Watrous, W.M.; Plaa, G.L. (1972) Effect of halogenated hydrocarbons on
organic ion accumulation by renal cortical slices of rats and mice.
Toxicol. Appl. Pharmacol. 22:528-543.
Winslow, S.G.; Gerstner, H.B. (1978) Health aspects of chloroform - a
review. Drug Chem. Toxicol. 1:259-275.
Wood, S.; Ward ley-Smith, B.; Halsey, M.J.; Green, C.J. (1982) Hydrogen
bonding in mechanisms of anesthetic tested with chloroform and
deuterated chloroform. Br. J. Anaesth. 54:387-391.
Wood-Smith, F.G.; Stewart, H.C. (1964) Drugs in anaesthetic practice.
Washington, DC: Butterworth, Inc.; pp. 131-135.
Yannai, S. (1983) Adrenocortical response to single and repeated doses of
chloroform in rats. Arch. Toxicol. 54:145-156.
5-85
-------
6. TERATOGENICITY AND REPRODUCTIVE EFFECTS
Schwetz et al. (1974) evaluated the effects of reagent grade chloroform
(lot no. 9649, Burdick and Jackson Laboratories, 99.3% chlGroform, 0.65%
carbon tetrachloride, and 0.03% unknown) on the maternal and fetal well-being
of Sprague-Dawley rats. Twenty female rats were exposed, by inhalation
(7 hours/day) to 30, 100, or 300 ppm chloroform on days 6-15 of gestation.
The authors analyzed the results statistically using Fisher's Exact
Probability Test, analysis of variance, Dunnett's test, or Tukey's test to
compare the frequency of anomalies, resorptions, maternal and fetal weights,
body lengths, liver weights, or serum glutamic pyruvic transaminase (SGPT)
activity in the exposed versus the control groups. The level of significance
was chosen at p < 0.05, and the litter was used as the experimental unit.
Animals exposed to the highest dose of chloroform (300 ppm) had
significant increases in the number of resorptions and a decrease in the
apparent conception rate (Schwetz et al., 1974). At the lower doses (30 and
100 ppm), no alterations in resorption rate, fetal body weight, conception
rate, number of implantations, or average litter size were observed. Fetal
crown-rump length was significantly decreased at 30 and 300 ppm, but not at
the 100 ppm level. At 100 ppm, increases in the incidence of acaudia
(absence of tail), short tail, imperforate anus, subcutaneous edema, missing
ribs, and delayed ossification of sternebrae were observed. At 300 ppm,
subcutaneous edema and abnormalities of the skull and sternum were observed,
but the incidence of these was not statistically significant. The authors
pointed out that small numbers of survivors in the 300 ppm group (4±7 versus
the control 10±4 live fetuses/litter) may have prevented adequate statistical
evaluation of this effect.
6-1
-------
In this study (Schwetz et al., 1974), chloroform produced evidence of
maternal toxicity, such as a decrease in the rate of maternal weight gain at
all dose levels and a decrease in food consumption during pregnancy at the
100 and 300 ppm level (Table 6-1). Other maternal effects include changes in
liver weight gain during pregnancy (no change in absolute liver weight gain
at 30 ppm, but an increase at 100 ppm, possibly due to the concomitant
anorexia at this dose). No significant changes in SGPT activity were
observed in groups exposed to chloroform at the 100 and 300 ppm levels.
Developmental effects observed in the 300 ppm group were associated with
anorexic effects in the mother, and to control for this effect a starvation
group was included in this study. The starvation control group was
restricted to a level of food consumption comparable to the 300 ppm
chloroform group. Animals on starvation diets (allowed 3.7 g/day of food on
days 6-15) had a significant decrease in the absolute weight of the liver and
an increase in the relative weight of the liver (see Table 6-1). The effects
of 300 ppm chloroform on the increase in the relative weight of the liver
were much greater than those of starvation alone. Additionally, exposure to
300 ppm chloroform resulted in a dramatic decrease in the number of animals
pregnant at sacrifice (15% pregnant versus 88% in air control), a decrease in
the number of live fetuses per litter (4 versus 10 live fetuses/litter), and
an increased percentage of litters with resorptions (100% versus 57%).
Examination of the uteri indicated that the conceptus had been completely
resorbed very soon after implantation. These effects appeared to result
primarily from chloroform exposure, and not from influences of maternal
stress, since anorexia and liver weight changes associated with starvation
were not accompanied by embryotoxic or teratogenic effects (see Table 6-1).
6-2
-------
TABLE 6-1. SUMMARY OF RESULTS OF THE SCHWETZ et al. (1974) STUDY
Chemical: Chloroform, reagent grade, lot no. 9649, Burdick and Jackson
Laboratories
Animal: Sprague-Dawley rats, 20 animals per group
Route of exposure: Inhalation, 30, 100, 300 ppm
Duration of exposure: 7 hours/day, days 6-15 of gestation
Summary of developmental effects
Chloroform
concentration,
(ppm)
Air Air control
control (starved) 30
100
300
% pregnancy
live fetuses/
litter
88(68/77) 100(8/8) 71(22/31) 82(23/28) 15(3/20)a
10±4 10±4 12+2 11±2 4±7a
% resorptions/
implantation 8(63/769) 7(6/87) 8(24/291) 6(16/278) 61(20/33)a
% litters with
resorptions 57(39/68) 25(2/8) 68(15/22) 52(12/23) 100(3/3)
fetal body
weight, g
5.69+0.36 5.19±0.29 5.55+0.20 5.59±0.24 3.42±0.02a
fetal crown-rump
length, mm 43.5+1.1 42.1±1.1 42.5±0.6 43.6+0.7 36.9±0.2a
aSignificantly different from controls.
SOURCE: Schwetz et al., 1974.
(continued on next page)
6-3
-------
TABLE 6-1. (continued)
Summary of maternal effects
Chloroform
concentration
Absolute liver
weight (g)
body weight
gain during
gestation
maternal
food
consumption
Air control
Air control (starved) 30
36±3
42±4
35±4
100
41+3
300
53+11
gestation
day
6
13
21
275±21
310117
389+28
274±13
223±13a
326±24a
266±14
280±l4a
381+23
274±17
274±18a
365±22
284±+9
192±9a
24l±29a
gestation
day
4-5
6-7
8-9
10-11
12-13
14-15
16-17
18-19
19±4
19±3
20±3
23±2
22±2
23±3
25±4
26±3
19±2
starved
starved
starved
starved
starved
21±2a
24±8a
20±3
3±3
18±1
18±2
20±1
21±2
27±3a
29+5
20±3
13+4a
15+2a
16+2a
15±2a
19+2a
30±3a
33+3a
18±1
l+±la
4±2
1±1
l±la
l±la
12±2a
aSignificant1y different from controls.
SOURCE: Schwetz et al., 1974.
6-4
-------
Murray et al. (1979) evaluated the effects of chloroform (spectral
grade, Mallinckrodt, lot CSZ, code 4434, purity not reported) administered by
inhalation (7 hours/day) in the mouse. Thirty-five animals per group were
exposed on gestation days 1-7 or 6-15; forty animals per group were exposed
on gestation days 8-15. Only one dose, 100 ppm, was tested during the three
different time periods (days 1-7, 6-15, or 8-15 of gestation) in CF-1 mice.
The various exposure periods were designed to evaluate the effects of
chloroform in very early pregnancy, organogenesis, and somewhat later in
pregnancy. Sodium sulfide staining of the uteri was used to detect
resorptions in very early pregnancies.
The authors (Murray et al., 1979) -analyzed the results statistically
using the Fisher's Exact Probability Test to evaluate pregnancy incidence;
the modified Wilcoxan test for fetal outcomes; the Mann-Whitney signed rank
test for SGPT activity; and one-way analysis of variance for fetal body
weights and body measurements, maternal body weights, liver weights, food
consumption, and number of implantations and resorptions. The authors chose
the level of statistical significance at p < 0.05.
Murray et al. (1979) reported that 100 ppm chloroform significantly
reduced the ability of the mice to maintain pregnancies when the animals were
exposed on days 1-7 or 6-15 of gestation but not on days 8-15 (see Table 6-2
for summary of data). No significant effect was reported on the average
number of implantation sites. In animals exposed on days 1-7 of gestation,
but not in those exposed on days 6-15 or 8-15, there were significant
increases in resorptions per litter. This effect was accounted for by the
loss of two entire litters. Mean fetal body weight and crown-rump length
were significantly decreased in the groups exposed on days 1-7 and 8-15, but
not in those exposed on days 6-15. Maternal toxicity (slight decrease in
6-5
-------
TABLE 6-2. SUMMARY OF RESULTS OF THE MURRAY et al. (1979) STUDY
Chemical: Chloroform, spectral grade, Lot CSZ, Code 4434, Mallinckrodt
Animal: CF-1 mice, 35 animals in groups exposed on days 1-7 and 6-15;
40 animals in groups exposed on days 8-15
Route of exposure: Inhalation, 100 ppm (one dose only)
Duration of exposure: 7 hours/day, days 6-15 of gestation
Days
Number Pregnant
Additional pregnancies
(sodium sulfide stain) Total pregnancies
1-7
6-15
8-15
Control
22/35(63%)
29/34(85%)
25/40(62%)
Exposed
11/34(32%)
13/35(37%)
18/40(45%)
Control
4
2
1
Exposed
4
2
6
Control
26/35(74%)
31/34(91%)
26/40(65%)
Exposed
15/34(44%)
15/35(43%)
24/40(60%)
1-7
Days of gestation
6-15
8-15
Fetal effects
resorptiona
fetal body weight
and crown-rump
lengthb
delayed skeletal
ossificationa
fetal body weight
and crown-rump
lengthb
cleft palate3
delayed skeletal
ossification*
Maternal effects body weight gain*3
body weight gain'5
liver weight3
SGPT (only one3
dose)
body weight gain15
liver weight3
3Significantly increased.
bSignificantly decreased.
SOURCE: Murray et al., 1979.
6-6
-------
body weight gain during pregnancy) was seen in groups exposed on days 6-15,
with a more severe decrease in groups exposed on days 1-7 and 8-15. Less
food and water were consumed in all experimental groups as compared to
controls. Absolute and relative weight of the liver were increased in groups
exposed on days 6-15 and 8-15, but not in those exposed on days 1-7. SGPT
activity was increased in mice exposed on days 6-15, which was the only time
period evaluated for this measurement.
A summary of this study (Murray et al., 1979) is presented in Table 6-2.
Chloroform caused a decrease in pregnancy maintenance. The authors concluded
that chloroform affected the stages either prior to, or during, early
implantation. However, because-^ef~the small numbers of animals and the lack
of dose-response evaluation (only one dose was tested, 100 ppm), this
conclusion must be considered tentative. Other results of the Murray et al.
(1979) study indicated that the incidence of cleft palates increased in pups
exposed i_n utero on days 8-15 of gestation, but not on days 1-7 or 6-15. The
authors suggested three possibilities to explain this result. The first was
that earlier exposure on days 6-15 prevented susceptible offspring from
implanting. The second possibility was that the number of litters available
(11 in the group exposed on days 6-15) was insufficient to detect this
effect. The third was that the teratogenic effect (cleft palate) did not
occur in offspring exposed on days 1-7 since they were exposed before
organogenesis. The number of offspring coming to term was consistently less
in all exposure groups than in the controls (days 1-7, 9 litters versus 22 in
control; days 6-15, 11 litters versus 29 in control; days 8-15, 18 litters
versus 24 in control). Since the pups with cleft palate were also retarded
in growth, the authors suggested that the ability of chloroform to cause
malformations was indirect embryotoxic effect and not direct teratogenic
-------
effect. However, this speculation was not supported by experimental results
and does not alter the conclusion that chloroform is a potential
developmental toxicant.
In summary, this study (Murray et a!., 1979) indicated that chloroform
administration (100 ppm) by inhalation (7 hour/day) produced teratogenic and
embryotoxic effects, interfered with pregnancy, and caused maternally toxic
effects (changes in liver weight and decreases in weight gain during
pregnancy). Exposure in the early stages of pregnancy appeared to produce a
decreased incidence of conception, but the results of this study did not
conclusively determine which days of pregnancy were most susceptible to the
effects of chloroform. To answer this question, it would be necessary to use
a greater number of doses and larger numbers of animals per dosage group.
Thompson et al. (1974) investigated the effect of chloroform
administered orally, using Sprague-Dawley rats and Dutch-Belted rabbits. The
rats were intubated with chloroform (Mai 1inckrodt, Batch ZJL dissolved in
corn oil, purity not reported) twice a day in divided doses of 20 to
516 mg/kg/day. The rabbits were intubated once a day in doses of 20 to
398 mg/kg/day. Each study was divided into two parts, a range-finding
portion designed to establish the proper dose range (six rats were
administered 79, 126, 300, 316, or 516 mg/kg/day of chloroform; five rabbits
were administered 63, 100, 159, 251, or 398 mg/kg/day), and a teratologic
study, using greater numbers of animals and three doses (25 rats were
administered 20, 50, or 126 mg/kg/day and 15 rabbits were administered 20,
30, and 50 mg/kg/day). The rats were exposed to chloroform on days 6-15 of
gestation, while the rabbits were exposed on days 6-18 of gestation.
Statistical evaluations of maternal body weight gains, food consumption,
implantations, corpus luteum, resorptions, litter size, and fetal weights
6 8
-------
were made by an analysis of variance and Dunne'.t's Two-Tailed Multiple Range
Test. Sex ratios and frequency of anomalies among the fetal population and
among litters were analyzed by the chi-square test. In all analyses, the
level of significance chosen was p < 0.05.
The data for the range-finding portion of this study (Thompson et al.,
1974) were not presented; however, the authors reported that rats treated
orally with greater than 126 mg/kg/day chloroform had signs of maternal
toxicity, such as a decrease in food consumption, acute toxic nephrosis,
hepatitis, and gastric erosion. Fetal development was adversely affected in
rats receiving 316 and 501 mg/kg/day with an increase in the number of
resorptions and a -decrease in fetal viability, litter size, and fetal weight
in dams given 316 mg/kg/day. Only two rats survived when given
501 mg/kg/day; one was not pregnant, the other had complete early
resorptions. In a teratologic study (Table 6-3), rats receiving 50 and 126
mg/kg/day displayed signs of maternal toxicity (lowered body weight gain,
lowered food consumption, and fatty changes in the liver). No overt toxic
effects were observed in animals given 20 mg/kg/day, and no malformation was
noted at any dose level.
In the range-finding portion of the study by Thompson et al. (1974)
using rabbits, there were signs of maternal toxicity such as severe acute
hepatitis and nephrosis, which were observed in animals given 63 mg/kg/day
and higher. No overt signs of toxicity were observed at the 25 mg/kg/day
level. In the two surviving dams given 100 mg/kg/day, one had four
resorption sites with no viable offspring, while the other was not pregnant.
No other embryotoxic or teratogenic effect was observed. In the teratologic
study of rabbits (Table 6-3), maternal toxicity (depressed weight gain) was
observed at the 50 mg/kg/day level. In the fetus, mean body weight was
6-9
-------
TABLE 6-3. SUMMARY OF EFFECTS OF THE THOMPSON et al. (1974) STUDY
Chemical: Chloroform NF, Mai 1inckrodt Batch ZJL
Animal: Sprague-Oawley rats, 25 animals per group, Dutch-Belted rabbits,
15 animals per group
Route of exposure: Rats intubated with 0, 20, 50, 126 mg/kg/day
rabbits intubated with 0, 20, 35, 50 mg/kg day
Duration of exposure: Rats, 6-15 of gestation, rabbits 6-18 of gestation
Maternal effects, teratologic study
rats
rabbits
rats
rabbits
Fetal effects,
rats
rabbits
20 mg/kg/day
20 mg/kg/day
no overt toxic
effect
teratology study
20 mg/kg/day
20 mg/kg/day
50 mg/kg/day
35 mg/kg/day
body weight3
body weight3
50 mg/kg/day
35 mg/kg/day
126 mg/kg/day
50 mg/kg/day
food consumption3
126 mg/kg/day
50 mg/kg/day
rats
rabbits
no significant
effect
mean body weight3
no significant
effect
no significant
effect
mean body weight3
Significantly decreased.
SOURCE: Thompson et al., 1974.
depressed at the 20 and 50 but not 35 mg/kg/day levels. No treatment-related
effect or major abnormalities were observed although there was an increase in
the numbers of offspring with incomplete skeletal when analyzed statistically
among all fetuses, but not among litters at the 25 or 35 mg/kg/day level.
6-10
-------
In this study (Thompson et al., 1974), adverse effects such as skeletal
deformities and deficiencies in pregnancy maintenance reported by Schwetz et
al. (1974) and Murray et al. (1979), were not observed. Fetotoxicity in the
form of reduced birth weights was observed at the highest doses which were
also toxic to the dam. The authors suggested that the greater maternal
toxicity observed in this study compared to that of others (Schwetz et al.,
1974; Murray et al., 1979) might be attributable to differences in the route
or duration of exposure. In the study by Thompson et al. (1974), rats and
rabbits were exposed orally to chloroform once or twice a day, while Schwetz
et al. (1974) and Murray et al. (1979) administered chloroform by inhalation
7 hour/day. No data are available on the-influence of different routes of
exposures.
Burkhalter and Balster (1979) evaluated the potential of chloroform to
adversely affect behavior in developing ICR mice. This study was designed as
a preliminary screening study, with the parental generation of male and
female mice (five animals) exposed 21 days prior to mating, during mating,
and for an additional 21 days. The offspring were exposed starting on day 7
until day 21 after birth. Only one oral dose of chloroform, 31.1 mg/kg/day,
was administered by gavage to parental generation and offspring. Each litter
was reduced to eight pups, and three pups were randomly selected for
behavioral teratogenic testing. The chloroform (Mallinckrodt, nanograde
purity) was administered by gavage and delivered in a solution of 1 part
polyoxyethylated vegetable oil, Emulphor (EL-620, GAF Corp., New York), and 8
parts saline. A variety of behavioral responses were evaluated which
included righting reflex, forelimb placing response, forepaw grasp, rooting
reflex, cliff drop aversion, auditory startled response, bar-holding ability,
eye opening, motor performance, and learning ability. The scoring for these
6-11
-------
responses was based upon predesignated criteria. These criteria established
behavioral ability by measuring both objective standards (time to complete
test) and subjective standards ("weak" or "complete" grasp of paws).
Body weight and latency to enter the dark compartment in the passive
avoidance test were analyzed statistically by repeated measures of the
analysis of variance. Screen test latencies were analyzed by t-test. The
data from the neurobehavioral developmental scale were analyzed using a Mann-
Whitney U test. The level of statistical significance was chosen at
p < 0.05.
Burkhalter and Balster (1979) reported that the sizes of litters were
similar for both the control and experimental groups; however, fetal body
weight gain of pups during the 14 days of exposure (days 7-21 following
birth) was decreased. Forelimb placement response was reduced in the exposed
group on day 5 and 7 of birth, but not on day 9. The significance of this
reduction is not known, although the recovery on day 9 suggested that the
effect was not indicative of serious delays in behavioral response. The
other behavioral responses were not significantly different in the exposed
groups. Burkhalter and Balster (1979) concluded that 31.1 mg/kg/day of
chloroform produced no significant adverse behavioral effects in pups exposed
both j_n utero and after birth (days 7-21). However, since this study was
•
-------
reported). The animals were exposed by inhalation (20 + 1.2 g/m3/day) on
days 7-14 of gestation. Two lower concentrations were administered in the
study, but the doses were not reported in the abstract. Dilly reported that
chloroform increased fetal mortality and decreased fetal weight gain;
however, there were no malformations.
In another abstract, Ruddick et al. (1980) reported the results of a
study using 15 Sprague-Dawley rats administered 100, 200, and 400 mg/kg/day
of chloroform by gavage on days 6-15 of gestation. Chloroform (purity not
reported) was reported to cause maternal toxicity (changes in weight gain,
biochemical and hematological parameters, and liver or kidney changes).
Chloroform also produced adverse effects in fetal development (type not
specified), but the authors attributed these effects primarily to maternal
toxicity and not directly to chloroform exposure. Without the presentation
of this data, however it is not possible to fully evaluate these results.
6.1 SUMMARY
In summary, the results of four articles and two abstracts indicated
that under the conditions of the experiments, chloroform has the potential
for causing adverse effects in pregnancy maintenance, delays in fetal
development, and production of terata in laboratory animals. The adverse
effects on the conceptus were observed in association with maternal toxicity;
'•'It?
however, the type and severity of effects appeared to be specific to the
conceptus, affecting the fetus to a much greater degree than the mother
(Schwetz et al., 1974; Murray et al., 1979). The results of a preliminary
screening study indicated that a single dose of 31.1 mg/kg/day chloroform has
no significant effect on neonatal behavior (Burkhalter and Balster, 1979).
Thompson et al. (1974) reported that chloroform does not cause adverse fetal
effects except at maternally toxic levels. The two abstracts did not contain
6-13
-------
enough detail for critical scientific review (Dilley et al., 1977; Ruddick et
al., 1980).
The studies in which chloroform was administered by inhalation for
7 hours/day (Schwetz et al., 1974; Murray et al., 1974) reported more severe
outcomes than other studies which administered chloroform by intubation once
or twice a day. However, since the pharmacokinetic relationship associated
with route or duration of exposure has not been studied, it is not possible
to evaluate the importance of the route of exposure in causing adverse
reproductive outcome. To evaluate more fully the influence of these factors,
additional investigations would have to be conducted.
6.2 REFERENCES FOR CHAPTER 6
Burkhalter, J.; Balster, R.L. (1979) Behavioral teratology evaluation
of chloroform in mice. Neurobehavioral Toxicol. 1:199-205.
Dilley, J.V.; Chernoff, N.; Kay, D.; Winslow, N.; Newell, E.W. (1977)
Inhalation teratology studies of five chemicals in rats. Toxicol. Appl
Pharmacol. 41:196.
Murray, F.A.; Schwetz, B.A.; McBride, J.G.; R.E. Staples. (1979) Toxicity
of inhaled chloroform in pregnant mice and their offspring. Toxicol.
Appl. Pharmacol. 50:151-522.
Ruddick, J.A.; Villenouve, O.C.; Chu, I.; Balli, V.E. (1980)
Teratogenicity assessment of four halomethanes. Teratology 21:66A.
Schwetz, B.A.; Leong, B.K.J.; Gehring, P.O. (1974) Embryo- and
fetotoxicity of inhaled chloroform in rats. Toxicol. Appl. Pharmacol.
28:442-451.
Thompson, D.J.; Warner, S.D.; Robinson, V.B. (1974) Teratology studies on
orally administered chloroform in the rat and rabbit. Toxicol. Appl.
Pharmacol. 29:348-357.
6-14
-------
REFERENCES FOR CHAPTER 6
Burkhalter, J.; Balster, R.L. (1979) Behavioral teratology evaluation
of chloroform in mice. Neurobehavioral Toxicol. 1:199-205.
Dilley, J.V.; Chernoff, N.; Kay, D.; Winslow, N.; Newell, E.W. (1977)
Inhalation teratology studies of five chemicals in rats. Toxicol. Appl
Pharmacol. 41:196.
Murray, F.A.; Schwetz, B.A.; McBride, J.G.; R.E. Staples. (1979) Toxicity
of inhaled chloroform in pregnant mice and their offspring. Toxicol.
Appl. Pharmacol. 50:151-522.
Ruddick, J.A.; Villenouve, D.C.; Chu, I.; Balli, V.E. (1980)
Teratogenicity assessment of four halomethanes. Teratology 21:66A.
Schwetz, B.A.; Leong, B.K.J.; Gehring, P.J. (1974) Embryo- and
fetotoxicity of inhaled chloroform in rats. Toxicol. Appl. Pharmacol.
28:442-451.
Thompson, D.J.; Warner, S.D.; Robinson, V.B. (1974) Teratology studies on
orally administered chloroform in the rat and rabbit. Toxicol. Appl.
Pharmacol. 29:348-357.
6-15
-------
7. MUTAGENICITY
7.1. INTRODUCTION
The mutagenic potential of chloroform has been assessed on the basis of
several i_n vitro bacterial studies, one host-mediated assay with Salmonella
as the indicator organism, one yeast study, one Drosophila sex-linked
recessive lethal test, one mammalian cell culture mutagenicity assay, two
sperm head abnormality tests, four cytogenetic studies, and seven ONA damage
studies (sister chromatid exchange and unscheduled DNA synthesis). Several
assays from a recently published screening study, in which 42 chemicals were
tested in various short-term protocols, are also briefly discussed.
Information on the binding of metabolically activated chloroform to cellular
macromolecules is presented before the sections on genetic damage. Most of
the mutagenicity tests have been negative, but there is some evidence that
chloroform may be mutagenic. Suggestions are made for further testing.
7.2. COVALENT BINDING TO MACROMOLECULES
As mentioned in Chapter 4, the primary reactive metabolite of chloroform
is phosgene, COC12. The toxicity and carcinogenicity of chloroform may be
related to its metabolism to phosgene, which is a crosslinking agent and may
bind to macromolecules covalently and crosslink them. Several studies on the
binding potential of metabolically activated chloroform have been described
in Chapter 4, section 4.6. Other studies and additional information on
studies summarized in Chapter 4 are covered below.
Diaz Gomez and Castro (1980a) assessed the potential of purified rat
liver nuclei to activate chloroform by measuring covalent binding to nuclear
protein and lipid. The results were compared to results obtained from
similar incubation mixtures containing microsomes instead ">f purified nuclei.
7-1
-------
The mixtures containing either nuclei (1.3 mg protein/ml) or microsomes
(1.56 mg protein/ml) were incubated for 30 min i.n 6.4 nM ^CHClj (5.4 Ci/mol)
and an NADPH generating system. The extent of binding to protein in the
nuclear preparation was approximately 40% of that in the microsomal
preparation (nuclei, 27 pmol/mg; microsomes, 68 pmol/mg). Thus, isolated
nuclei were less efficient than microsomes in metabolizing chloroform, but
the results were within the same order of magnitude.
This study suggests that metabolism of chloroform to a reactive
intermediate(s) can occur in association with nuclear membranes, as may be
the case with other xenobiotics (Weisburger and Williams, 1982). However,
contamination of the nuclear preparations with trace-amounts of endoplasmic
reticulum may have been sufficient to explain at least part of the nuclear
activation observed.
In a subsequent study, Diaz Gomez and Castro (1980b) exposed mouse liver
DNA or RNA to ^CHClg in vivo or i_n Vitro. For the i_n vivo experiments, the
animals were pretreated with phenobarbital or 3-methylcholanthrene and
exposed to 14CHC13 by intraperitoneal injection either once daily for 4 days
or twice weekly for 2 weeks. The specific activity of the 14CHC13 sample
injected was 13.15 pCi/mmol (10% in olive oil) administered at 5 mg/kg. The
animals were killed 6 hr after the last injection and the livers were
analyzed for binding of l^CHCls to DNA and RNA. For the jn vitro study, the
investigators added 2.5 x 10.5 dpm 14QHC13, microsomes (9 mg of microsomal
protein), cofactors, and 4 mg of mouse liver DNA to 3 ml of reaction mixture.
The reaction mixtures were incubated for 30 min at 37°C. It is not clearly
stated whether microsomes from chemically pretreated animals were used in the
ijn vitro study. The background counts for both the in vivo and j_n vitro
7-2
-------
experiments were 150-160 dpm. The authors observed no significant binding of
14c to the nucleic acids in both the in \QyjD_and j_n vitro experiments.
DiRenzo et al. (1982) studied i_n vitro covalent binding of chloroform to
calf thymus DNA following bioactivation by liver microsomes isolated from
phenobarbital-treated rats. ^CHC^ (5.4 Ci/mol, New England Nuclear) was
mixed with unlabeled chloroform and diluted in ethanol so that 1 pmol
contained 2-4 x 105 dpm (A.J. Gandolfi, personal communication). The ^CHC^
(2 jamol or 4-8 x 1Q5 dpm) was incubated with 2 mg of microsomal protein, 5 mg
of calf thymus DNA, and an NADPH-generating system in 2.5 ml of 0.05 M Tris-
HC1, pH 7.4. After 60 min, the reaction was stopped with cold ethanol and
DNA was isolated by pronase digestion- followed by extraction with
chloroform/isoamyl alcohol (24:1) and precipitation with cold ethanol/sodium
chloride. Background was 40-50 dpm (A.J. Gandolfi, personal communication).
Protein as analyzed by the Coomassie technique was < 10 yg/mg of DNA.
Chloroform bound to DNA at 0.46 + 0.13 nmol/mg DNA/hr. The majority of the
radiolabel cosedimented with native DNA and putative partially hydrolized DNA
in CsCl gradients.
When the two i_n vitro binding studies (Diaz Gomez and Castro, 1980b;
DiRenzo et al., 1982) are compared, it is clear that DiRenzo and coworkers
optimized binding conditions, which may be the reason they obtained positive
results. DiRenzo et al. (1982) added 2- to 3-fold more dpm to the reaction
mixtures, incubated the mixtures twice as long, and had about 25% of the
background levels, when compared with Diaz Gomez and Castro (1980b). In
addition, DiRenzo et al. (1982) used microsomes from chemically pretreated
animals, whereas it was not clearly stated whether the microsomes Diaz Gomez
and Castro (1980b) used in their i_n vitro studies were from chemically
pretreated animals. The signifiance of i_n vitro binding for in vivo exposure
7-3
-------
is unclear. If glutathione were present in the reaction mixtures to more
nearly simulate i_n vivo conditions, DiRenzo and coworkers may not have
observed binding. Also, binding at the low levels observed could have been
due to a reactive impurity in the radiochemical sample.
In summary, binding of metabolically activated chloroform to liver
microsomal and nuclear protein and lipid has been observed. The only
conclusion that can be made from the available studies on the DNA binding
potential of metabolically activated chloroform is that, if DNA binding does
occur, it is probably at a very low level.
7.3. MUTAGENICITY STUDIES IN BACTERIAL TEST SYSTEMS
Uehleke et al. -(-1977) tested chloroform for mutagenicity .in suspension
assays with S. t.yphimurium strains TA1535 and TA1538. No mutagem'c activity
was observed. About 6-9 x 10^ bacteria were incubated for 60 min under N^ in
tightly closed test tubes with 5 mM chloroform and microsomes (5 mg protein)
plus an NADPH generating system. The mutation frequencies (His+ colony
forming units/10° his" colony forming units) were less than 10 for both
strains and the spontaneous mutation frequencies were 3.9 + 3.7 for strain
TA1535 and 4.4 + 3.5 for strain TA1538. At this concentration of chloroform,
survival of the bacteria was at least 90%. Higher concentrations should also
have been tested, because mutagenicity is sometimes observed only at higher
toxicities. Dimethylnitrosamine (50 mM). cyclophosphamide (0.5 mM), 3-
methylcholanthrene (0.1 mM), and benzo(a)pyrene (0.1 mM) were the positive
controls.
Studies demonstrating that metabolically activated chloroform binds to
protein and lipid in the presence of rabbit microsomes were also described in
the paper by Uehleke et al. (1977) and are mentioned in Chapter 4 of this
document. However, it is not clear whether rat, mouse, or rabbit microsomes
7-4
-------
were used in the mutagenicity studies in this paper. If mouse or rat
microsomes were used for the mutagenicity tests, it cannot be assumed that
chloroform was sufficiently activated, because only rabbit microsomes
activated ^CHCl^ sufficiently for binding to macromolecules in this study.
Another deficiency in this study is that Salmonella strains TA98 and TA100
were not used; these strains contain the R factor plasmid pKMlOl, which
increases the sensitivity of the tester strains to certain mutagens.
The mutagenicity of chloroform was tested in a study on the mutagenic
potential of 71 chemicals identified in drinking water (Simmon et al., 1977).
Chloroform was tested at 10% by volume (1.24 M) in a suspension assay with
Salmonella strains TA1535, TA1537-, TA1538, TA98, and TA100; this.
concentration of chloroform exceeds its solubility. Metabolic activation was
provided by S9 mix prepared from Aroclor 1254-treated rats. Mutagenic
activity was not observed, but no information on toxicity was provided.
Simmon et al. (1977) also tested chloroform in a desiccator to assess
mutagenicity due to vapor exposure. Agar plates were placed uncovered in a
desiccator above a glass petri dish containing the chloroform. The
desiccator contained a magnetic stirrer that acted as a fan to promote
evaporation of the measured amount of chloroform and to maintain an even
distribution of the vapors. Plates //ere exposed to the vapors for 7-10 hr
and then removed from the desiccators, covered, ind incubated approximately
40 hr before scoring. As in the suspension assay, mutagenic activity was not
observed and no information on toxicity was provided.
The study of Simmon et al. (1977), although lacking in details of the
chloroform assay, clearly identifies other trihalomethanes (CHBr3, CHBr2Cl,
CHBrCl2) as mutagens in the vapor assay in desiccators. Methyl bromide,
methyl chloride, methyl iodide, and methylene chloride were also found to be
7-5
-------
mutagenic in the desiccator assay. However, these seven halogenated
compounds did not require metabolic activation to exhibit mutagenic activity.
Kirkland et al. (1981) studied the mutagenicity of chloroform in
Escherichia coin strains WP2p and WP2uvrA~p, using reversion to tryptophan
prototrophy as the end point. The bacteria were treated with chloroform in
plate incorporation and preincubation tests both with and without rat liver
microsomes (plus cofactors) prepared from Aroclor 1254-induced rats. The
concentration of protein in the microsomal suspension was not given.
Chloroform was added at 10,000, 1,000, 100, 10, 1, or 0.1 ^g/plate. Negative
results were obtained in both tests. In the plate incorporation test, the
procedure used to prevent loss of chloroform may have been inadequate;
chloroform was added to bacterial suspensions in molten agar, and each
mixture was mixed rapidly on a Whirlimixer and poured onto agar plates. Loss
of chloroform could have occurred during this mixing procedure. The plates
were incubated in gas-tight containers, but it is unlikely that this would
prevent evaporation of chloroform from the plates. 2-Aminoanthracene was the
positive control requiring metabolic activation, and N-methyl-N'-nitro-N-
nitrosoguanidine was the direct-acting positive control.
In a subsequent study from the same laboratory (Van Abbe et al., 1982),
two experiments were carried out: the standard plate procedure in which no
precautions were taken to control evaporation of chloroform and a vapor-phase
experiment in which the bacteria were exposed to chloroform vapors for 2-8
hr. Because chloroform is associated with the development of liver tumors in
mice and kidney tumors in rats (NCI bioassay, 1976), the standard plate
assay, which usually employs only rat liver S9 mix when activation is
desired, included Aroclor-induced S9 mixes from mouse and rat livers and
kidneys. Strains TA1535, TA1537, TA1538, TA98, and TA100 and chloroform
7-6
-------
doses of 0-10 mg/90-mm plate were tested in the standard plate incorporation
assay. Negative results were obtained at all doses, both in the absence of
metabolic activation or in the presence of S9 mixes from all sources.
Toxicity (incomplete bacterial lawn) was observed at 10 mg/plate. All of the
positive controls tested were mutagenic under similar conditions.
In the vapor phase study (Van Abbe et al., 1982), the protocol differed
from the standard procedure. The bacterial suspension (TA1535 or TA1538) and
(where appropriate) the rat liver S9 mix were spread over the surface of the
bottom agar and the cofactors were incorporated into the top agar. The
plates were exposed to a stream of chloroform vapor in an anaerobic jar at 32
ml per hr in triplicate runs. The positive controls, ethyl methanesulfonate
(2.5%) and 2-acetylaminofluorene (50 jag/plate) were applied as discs.
Negative results were obtained and toxicity was observed after 6-8 hr of
exposure to the chloroform vapor. Because the vapor phase study was carried
out using a nonstandard procedure, its validity cannot be evaluated. In
addition, the positive controls were not volatile and were therefore not
tested in the same way as chloroform. Also, only rat liver S9 mix was used
for metabolic activation in the vapor phase study, even though the
investigators emphasized the advisability of including mouse and rat liver
and kidney S9 mixes for the standard protocol.
Gocke et al. (1981) assessed the mutagenicity of 31 chemicals (including
chloroform) used as ingredients in European cosmetics. Chloroform was
studied in three test systems: the Salmonel1 a/microsome test, the sex-linked
recessive lethal test in Drosophila, and the micronucleus test in mice. The
latter two tests will be discussed in the following sections. At least five
doses, usually up to 3.6 mg/plate for nontoxic and soluble compounds, were
tested in the Salmonella/microsome assays. Salmonella strains TA1535, TA100,
7-7
-------
TA1538, TA98, and TA1537 were used with and without activation by S9 mix
prepared from Aroclor-pretreated rats. Three halogenated aliphatic
hydrocarbons were tested (1,1,1-trichloroethane, dichloromethane, and
chloroform). Gocke and his coworkers placed the plates containing the
bacteria, metabolic activation system, and various amounts of the halogenated
aliphatic hydrocarbons in airtight desiccators for 8 hr and stated that the
desiccators were used because these three substances are volatile. However,
this procedure is not likely to prevent evaporation of the volatile test
compounds from the plates. The composition and purity of the chemicals were
not specified, and details of the protocols were not given. Dichloromethane
and trichloroethane exhibited mutagenic activity with and without metabolic
activation, but chloroform was inactive. However, information as to whether
the cells were exposed to adequate doses of chloroform was not provided,
because it is not known whether doses up to a toxic level were tested.
In a screening study of 42 chemicals (de Serres and Ashby, 1981),
chloroform was evaluated in 38 i_n vivo and ijn vitro short-term tests for
potential genotoxicity. Results from bacterial assays carried out in 18
laboratories using Salmonella (Ames reversion test) or E± coli (forward
mutation test) were essentially negative. These results are inconclusive,
however, because the studies were carried out blind and, therefore, the
investigators could not take into account the physicochemical properties
(such as volatility) of the substances they were testing when designing their
protocols.
Agustin and Lim-Sylianco (1978) investigated the mutagenicity of
chloroform in a host-mediated assay in which the indicator organisms were
Salmonella strains TA1535 and TA1537 injected into male and female mice. The
authors found that male mice metabolized chloroform into a mutagen that was
7-8
-------
active in strain TA1537. However, only ratios of mutation frequencies for
treated and control animals were given, rather than actual colony counts.
The ratio of mutation frequencies (tested/control) for strain TA1537 was
36.75 in male mice and 2.30 in female mice. The mutation frequencies for
strain TA1535 were 0.61 and 0.12, respectively. The results with strain
TA1535 indicate that the mutation frequency of the unexposed bacteria was
greater than that of the exposed bacteria. This strange result suggests that
the data may be extremely variable. Details of the procedures, such as doses
of chloroform, numbers of bacteria injected and recovered, routes of
exposure, and times of exposure before the animals were killed, were not
presented. Therefore, despite the possibility of a positive response, firm
conclusions cannot be reached because of inadequate data presentation,
insufficient details about the protocols, and variability of the data.
Agustin and Lim-Sylianco (1978) also studied the mutagenicity of ether-
extracted urine concentrates from 10 male mice in a bacterial spot test in
strain TA1537- The mice were exposed to chloroform at 700 mg/kg. Urine
concentrates from chloroform-treated mice yielded 302 revertant colonies and
a zone of inhibition of 29 mm, whereas urine from control animals yielded 10
revertant colonies with no zone of inhibition. Details of the ether
extraction procedure were not provided, but the likelihood of a false
positive result due to the presence of histidine in the extracted urine is
low, because the urine concentrate from the control animals yielded only 10
colonies and was presumably subjected to the same extraction procedure.
In summary, the results of bacterial tests of chloroform are
predominately negative, indicating that chloroform is not a mutagen in
bacteria. However, false negative results could have been obtained due to a
number of factors, including:
7-9
-------
1. The activation systems, although adequate for metabolism of the
standard positive controls, may have been inadequate for metabolism
of chloroform.
2. Phosgene, the primary reactive metabolite of chloroform, is
unstable and highly reactive (Kirk-Othmer, 1971). Because
exogenous activation systems (i.e., S9 mixes) were used in these
negative studies, any phosgene generated may have been scavenged by
microsomal protein or lipid before reaching the bacterial DMA.
However, the positive result in the urine spot test (Augustin and
Lim-Sylianco, 1978), if confirmed, would argue against this. The
positive urine spot test suggests that the mutagenic species is
stable.
3. Adequate exposure to chloroform may not have occurred if
appropriate precautions were not taken to prevent the evaporation
of chloroform. The use of a' volatile positive control compound
would be helpful in this regard. Many of the above studies did not
include volatile mutagenic chemicals, either as controls or as
additional test compounds. A toxic response at the higher
concentrations of chloroform tested would be an indication that the
bacteria were adequately exposed.
The positive result in the host-mediated assay cannot be evaluated
independently because the details of the procedures and appropriate data were
not reported and the data were variable. The results from the urine spot
test are suggestive of a positive response. These results indicate a need
for additional testing.
7.4. MUTAGENICITY STUDIES IN EUCARYOTIC TEST SYSTEMS
Callen et al. (1980) studied the genetic activity of chloroform in
7-10
-------
strain D7 of Saccharomyces cerevisiae, which contains an endogenous
cytochrome P-450 dependent monooxygenase metabolic activation system. By
using yeast, Callen and his coworkers eliminated the need for the exogenous
type of metabolic activation system used in the bacterial studies. Three
different genetic end points can be examined in strain D7: mitotic gene
conversion at the trp5 locus, mitotic crossing over at the ade2 locus, and
reversion at the ilvl locus. The effect of chloroform on these end points
was measured in cells exposed in suspension to 2.5, 5.0, and 6.3 g of
chloroform per liter of buffer (21 mM, 41 mM, and 54 mM, respectively). The
purity of the chloroform sample (from J.T. Baker) was not stated. Escape of
volatilized chloroform is not expected to have occurred to any significant
extent, because the incubations were carried out in screw-capped glass tubes.
Results of the study by Callen et al. are presented in Table 7-1. A 1-hr
treatment of cells with 54 mM chloroform resulted in an increased ratio of
convertants to survivors and a marginal increase in the observed number of
convertant colonies. Similar results were obtained for mitotic crossing over
and gene reversion. Toxicity at this concentration was high (6% survival).
At the lower concentrations of chloroform (21 mM and 41 mM), a small
dose-related increase (1.2-fold and 2.7-fold, respectively) in convertants
was observed. In addition, a 9-fold increase in the frequency of genetically
altered colonies, which are due primarily to gene conversion and mitotic
crossing over, was observed at 41 mM chloroform. For gene reversion, a 2-
fold increase was observed at 41 mM chloroform. Toxicity was low at these
levels. These results strongly suggest a positive response, but additional
studies are needed before it can be stated conclusively that chloroform
causes genetic effects in yeast.
7-11
-------
TABLE 7-1. GENETIC EFFECTS OF CHLOROFORM ON STRAIN D7 OF S. CEREVISIAE
Concentration, mM
ro
ade2 locus (mitotic crossing over)
Total twin spots
Mitotic cross-overs/10^ survivors
Total genetically altered colonies
Total genetically altered colonies/
10^ survivors
ilvl locus (gene reversion)
Total revertants
Revertants/10° survivors
21 41
54
Survival
Total colonies
% of control
trp5 locus (gene conversion)
Total convertants
Convertants/10^ survivors
1423
100
246
1.7
1302
91
274
2.1
982
69
450
4.6
84
6
278
33.1
1124
1.6 1.7 4.1 44.8
6 11 43 47
1.0 1.9 8.9 52.7
61 46 81 50
4.3 3.5 8.2 60.0
-------
Sturrock (1977) tested the mutagenicity of chloroform at the
hypoxanthine- guanine phosphoribosyl transferase (HGPRT) locus in Chinese
hamster lung fibroblasts (V-79 cells) in culture. The cells were grown to a
monolayer and exposed for 24 hr to an atmosphere containing 1 to 2.5%
chloroform. Cells were then plated onto media with or without 8-azaguanine.
After incubation, theplates were examined for mutations and survival. No
significant increase in the frequency of mutants was observed in treated
cultures as compared with untreated controls. However, no provision was made
for metabolic activation, so the test must be regarded as incomplete.
Gocke et al. (1981) evaluated the mutagenicity of chloroform in the sex-
linked recessive lethal test in Drosophila. The flies were exposed by the
adult feeding method to 25 mM chloroform. Three successive broods (3-3-4
days) of flies were examined for sex-linked recessive lethal mutations. Over
4000 chromosomes per brood were tested. In two of the broods, small
increases in mutations were observed. Frequencies of sex-linked recessive
lethals were as follows: Brood 1, 20/4616 chromosomes (0.43%); Brood 2,
13/4349 chromosomes (0.29%); Brood 3, 15/4249 chromosomes (0.35%). Controls
were 0.27%, 0.14%, and 0.39%, respectively. Although there may be an
increase in the frequency of lethals in Brood 1, this increase was not
significant, as determined by the Kastenbaum-Bowman test (Kastenbaum and
Bowman, 1970).
Because sperm head abnormalities may be caused by mutations in genes
that control spermatogenesis (Wyrobek and Bruce, 1978), assays for
morphological sperm abnormalities have been used as indicators of the
mutagenic potential of chemicals. In a study of 54 chemicals, Topham (1980)
reported that chloroform did not induce sperm head abnormalities in (CBA X
BALB/c)F;L mice. Groups of five male mice received five daily intraperitoneal
7-13
-------
injections of corn oil alone (5 ml/kg/day) or chloroform in corn oil at
0.025, 0.05, 0.075, 0.1, and 0.25 mg/kg/day. Topham (1980) reported that the
highest of these doses (0.25 mg/kg/day) was lethal. Five weeks after the
last treatment, caudal sperm smears were examined for morphological
abnormalities. Topham (1980) reported that chloroform induced sporadic small
increases in abnormal sperm heads at a high dose and that this result could
not be repeated. Raw data for the experiments with chloroform were not
presented.
A sperm head abnormality test was conducted by Land et al. (1981) in
order to determine whether certain anesthetics affect mouse sperm morphology.
Groups of five male mice (11 weeks old) were exposed by inhalation to
chloroform at 0.04 and 0.08% (vol/vol) for 4 hr/day for 5 days in glass
exposure chambers. Control mice (N = 15) were exposed to compressed air
under similar conditions. Twenty-eight days after the first exposure, the
nine survivors from each exposure level were killed and the caudal sperm were
examined for abnormalities. The results were reported as % abnormal sperm (±
SEN) and were as follows: control, 1.42 + 0.08; 0.04% chloroform, 2.76 +
0.31; 0.08% chloroform, 3.48 + 0.66. The authors concluded that exposure of
mice to chloroform resulted in a significant increase in sperm head
abnormalities compared to the control (P < 0.01). Significance was
calculated by the t-test and the F test, but these tests may not be
appropriate because of the nonhomogeneity of the variance in chloroform-
treated and control groups (Dr. Chao Chen, Carcinogen Assessment Group, U.S.
EPA, personal communication). Although this study suggests a positive
response, a more appropriate statistical analysis of the data is needed.
Unfortunately, the data necessary to carry out the analysis were not
provided.
7-14
-------
Testing of the mutagenic potential of chloroform in eucaryotic systems
was carried out in the same screening study edited by de Serres and Ashby
(1981) that was discussed in the previous section for bacterial assays.
Seven yeast assays, two in vitro mammalian DMA damage assays (unscheduled DMA
synthesis and sister chromatid exchange), and three whole-animal tests
(Drosophlla sex-linked recessive lethal, mouse bone marrow micronucleus, and
mouse sperm abnormality) were reported for chloroform. The DNA damage
studies will be discussed in the next section of this chapter.
The seven yeast assays measured forward and reverse mutations, mitotic
crossing over, mitotic gene conversion, and aneuploidy in mitotic cells. A
positive result was obtained only in-a-forward mutation assay in
Schizosaccharomyces pombe. In the reverse mutation assay, chloroform was
tested only in stationary phase cells of S. cerevisiae in the presence of rat
S9 mix. Exposure was for 24 hr. Growing cells are more sensitive to the
mutagenic effects of several chemicals than are stationary cells, possibly
because log-phase yeast cells contain an endogenous cytochrome P-450
metabolizing system (Callen et al., 1980). The negative results in the
mitotic gene conversion assay, which was carried out in S. cerevisiae strain
D7, are in conflict with the positive results reported for chloroform in
strain D7 by Callen et al. (1980). The three whole-animal tests on
chloroform yielded negative results (de Serres and Ashby, 1981). Chemical
Work Group C in this volume edited by de Serres and Ashby recommended that
chloroform be tested further in in vivo short-term tests.
In summary, the results in eucaryotic test systems are primarily
negative but inconclusive. More studies, particularly with organisms
possessing endogenous activation, are needed before a definite conclusion on
7 15
-------
the mutagenicity of chloroform can be reached. Suggested studies are
described later in this chapter.
7.5. OTHER STUDIES INDICATIVE OF DMA DAMAGE
Two types of DNA damage studies, sister chromatid exchange (SCE) and
unscheduled DNA synthesis (UDS), are described in this section.
Sister chromatid exchange (SCE) is widely used as an indicator of
induced DNA damage. White et al. (1979) studied the induction of SCEs by
chloroform and other anesthetics. Information about the purity of the
compounds was not given. Exponentially growing Chinese hamster ovary cells
were exposed to the gases in the presence and absence of S9 mix (10% by
volume) prepared from Aroclor-induced rat livers. -Exposure .was at 0.71%
(vol/vol) chloroform (88 mM) for 1 hr in closed screw-capped culture flasks.
This exposure was 1 human MAC (maximum allowable concentration). After the
chloroform was removed from the culture flasks, the cells were incubated for
24 hr in medium containing 10 pM 5-bromo-2'-deoxyuridine (BrdUrd). Numbers
of SCEs per chromosome were 0.544 + 0.018 for chloroform-exposed cells and
0.536 + 0.018 for the controls. Positive results were observed for
anesthetics containing vinyl groups. Although the rat liver S9 was
sufficient for activation of vinyl compounds to derivatives (presumably
epoxides) that induce SCE, it may not be adequate for activation of
chloroform and other haloalkanes. Also, the exposure (1 hr) was short, only
one dose was tested, and information on toxicity was not provided.
Another SCE study was carried out by Kirkland et al. (1981) in human
lymphocytes. The cells were treated with chloroform at 25, 50, 75, 100, 200,
and 400 ng/ml (0.2, 0.4, 0.6, 0.8, 1.6, and 3.3 mM, respectively) for 2 hr in
the presence of S9 mix from Aroclor-induced rats. Metaphase spreads from
approximately 100 cells per treatment were examined. Acetone was the
7-16
-------
negative control, and a positive control was not included in the assay
because the same donor's lymphocytes had previously shown a dose-related
increase in SCE after treatment with benzo(a)pyrene in the presence or
absence of S9 mix. A small increase in SCE occurred at 50 ng of chloroform
per ml, but no dose-dependent relationship was observed. The study has
several shortcomings. First, the positive control was not concurrent.
Second, there is no indication that evaporation and escape of chloroform was
prevented. Third, the maximal dose was only 3.3 mM. Fourth, information
about the toxicity of chloroform for the lymphocytes was not provided.
In contrast to the negative results for the above two SCE tests,
Morimoto and Koizumi (1983) found that:chloroform and other trihalomethanes
induced SCEs in human lymphocytes i_n vitro and mouse bone marrow cells j_n
vivo. For the in vitro experiment, the lymphocytes were exposed for 72 hr to
0-50 mM chloroform dissolved in dimethyl sulfoxide in the presence of 20 pM
Brdllrd. Colcemid (0.2 pM final concentration) was added 3 hr before
fixation. Thirty-five second-division cells were scored for each point. A
dose-related increase above the control value of 8 SCEs per cell was
observed. At 2 mM, 10 mM, and 50 mM chloroform, the numbers of SCEs per cell
were about 8.5, 10.5, and 14.7, respectively. The lowest chloroform
concentration that caused a significant increase in SCE (Student's t test, P
< 0.05) was 10 mM.
In the in vivo experiment, male mice (45-50 g) ingested 0-200 mg of
chloroform per kg for 4 days. The chloroform was dissolved in olive oil.
Immediately after the last administration, infusion with BrdUrd for 24 hr was
initiated. Colcemid (1 mg/kg) was injected intraperitoneally 2 hr before
termination of the experiment. The mice were killed and bone marrow cells
were removed, fixed in methanolrglacial acetic acid (3:1), and air dried on
7-17
-------
micros!ides. The chromosomes were stained by a modification of the
fluorescence-plus-Giemsa method. SCEs were analyzed in 25 second-division
cells for each animal. A dose-related increase in SCEs per cell was
observed. At 0, 25, 50, 100, and 200 mg/kg/day, the mean SCEs per cell were
about 5.8, 5.5, 6.8, 7.2, and 9.2, respectively. The lowest chloroform
concentration that resulted in a significant increase in SCE (Student's t
test, P < 0.05) was 50 mg/kg/day.
Two i_n vitro mammalian DMA damage studies on chloroform (UDS and SCE)
were described in the volume edited by de Serres and Ashby (1981). The SCE
assay (Chapter 51) utilized an exogenous activation system and yielded
negative results. Chinese hamster ovary cells, were exposed to chloroform
(0.084 to 84 urn) in the presence of rat liver S9 mix for 1 hr. This
treatment time may be insufficient, particularly since a positive response
for 2-acetylaminofluorene in the presence of S9 was obtained after a 2-hr
exposure but not after a 1-hr exposure. In addition, the cells were exposed
to very low levels of chloroform and there was no indication that precautions
were taken to prevent evaporation and loss of chloroform from the culture
flasks. Thus, the negative results obtained in this study are inconclusive.
Unscheduled DNA synthesis (UDS), which is a measure of ONA repair, can
be used as an indicator of chemically induced DNA damage. Mirsalis et al.
(1982) measured UDS in primary rat hepatocyte cultures following i_n vivo
treatment of adult male Fischer-344 rats (175-275 g) with chloroform at 40
and 400 mg/kg by gavage. Control rats received corn oil (the vehicle for
chloroform) by gavage. At 2 or 12 hr after treatment, the livers were
perfused in situ and hepatocytes were isolated. Approximately 6 x 105 viable
cells were seeded in 35-mm culture dishes containing coverslips and allowed
to attach to the coverslips for about 90 min. After the coverslip cultures
7-18
-------
were washed, they were incubated in a medium containing 10 pCi [^H]thymidine
(40-50 Ci/mmol) per ml for 4 hr. The cultures were washed again and
incubated in medium containing 0.25 mM cold thymidine for 14-16 hr. The
extent of UDS was assessed by autoradiography. Net grains/nucleus were
calculated as the silver grains over the nucleus minus the highest grain
count of three adjacent nuclear-sized areas over the cytoplasm.
Cells from negative control animals (given vehicle only) ranged from
-3.0 to -5.1 net grains/nucleus. Several chemicals were positive in this
assay (>5 net grains/nucleus was considered positive), including methyl
methanesulfonate, dimethylnitrosamine, 2-acetylaminofluorene, and benzidine.
Chloroform at 40-and 400 mg/kg^yielded a_negative response (-2.7 to -4.4 net
grains/nucleus). However, rats are not susceptible to chloroform-induced
hepatocarcinogenesis (NCI bioassay, 1976). Benzo(a)pyrene and
7,12-dimethylbenz(a)anthracene, carcinogens that, like chloroform, are not
rat liver carcinogens, were also negative in this assay. These chemicals
were positive, however, in the in vitro rat hepatocyte UDS assay (Williams,
1981). This discrepancy suggests that the i_n vitro test may be more
sensitive than the in vivo assay. Chloroform has not been tested in the j_n
vitro rat hepatocyte UDS assay. Since the mouse is susceptible to
chloroform-induced liver tumors (NCI bioassay, 1976), it may be a more
appropriate test animal than the rat for the i_n vivo UDS assay.
It is uncertain whether measuring UDS by subtraction of cytoplasmic
grain counts from nuclear grain counts would allow for detection of a weak
response. In a recent article discussing the validity of the
autoradiographic procedure for detecting UDS in rat hepatocytes, Lonati-
Galligani et al. (1983) describe some potential problems with this method.
First, they found that it is difficult to obtain hepatocyte preparations of
7-19
-------
reproducible quality. Since preparations can differ in their metabolic
capabilities, they suggest that test chemicals should be studied in
conjunction with a potent UOS-inducing analog and that negative results be
accepted only in tests in which the analog is strongly positive. No known
positive analog of chloroform was tested in the study of Mirsalis et al.
(1982). Second, the cytoplasmic layer covering the nucleus is thinner than
the cytoplasmic area next to the nucleus. Therefore, a variable
overcorrection is probably applied, as witnessed by the usually higher
cytoplasmic than nuclear counts observed in control cells. This effect would
tend to obscure a weakly positive UDS response. Lonati-Galligani et al.
(1983) suggest that the grains.over the nucleus and over a cyfceplasmic area
should be scored and dose-response curves plotted separately instead of
subtracting cytoplasmic grain counts from nuclear grain counts. Both dose-
response curves should be considered before a decision is reached on whether
exposure to a certain chemical results in UDS.
Results of an additional DMA repair study were published by Reitz et al.
(1980a). Mice were exposed orally to chloroform at 240 mg/kg, and DNA repair
in the livers was assayed. Negative results were obtained. The methodology
used to assay for DNA repair was provided in a paper on vinylidine chloride
from the same laboratory (Reitz et al., 1980b). The difference in the
control values in the vinylidine chloride paper was an order of magnitude in
some cases, indicating that the assay methodology was not sensitive enough to
pick up repair DNA synthesis. Neither paper included information on the
length of time the mice were exposed to chloroform. Some compounds require a
longer exposure than others, as evidenced by the results with
2-acetylaminofluorine in Mirsalis et al. (1982). The possibility of false
negative results, as discussed above, exists in this study as well.
7-20
-------
The UDS assay discussed in Chapter 48 of the de Serres and Ashby (1981)
volume was carried out with HeLa cells, which do not contain a cytochrome
P-450 activation system. An exogenous rat liver S9 activation system was
employed. Although chloroform was positive in this assay in the absence of
the activation system, the discussion of this assay in the de Serres and
Ashby book suggests that the result may be misleading because of inadequate
statistical evaluation. In the presence of rat liver S9, chloroform was
negative.
In summary, many of the above studies on the DNA-damaging potential of
chloroform suffer from various deficiencies. Three negative SCE studies were
inconclusive. Positive results for i_n vitro and i_n vivo SCE were reported by
Morimoto and Koizumi (1983). Three negative UDS studies were inconclusive.
Increases in mitotic gene conversion and mitotic crossing over in yeast
suggest that chloroform damages DNA (Callen et al., 1980). The weight of the
available evidence suggests that chloroform may damage DNA. Additional
studies will be required, however, before firm conclusions can be reached.
7.6. CYTOGENETIC STUDIES
Kirkland et al. (1981) studied the ability of chloroform to induce
chromosome breakage in cultured human lymphocytes. The cells from one donor
were treated with chloroform at 50, 100, 200, and 400 ng/ml for 2 hr in the
presence of an S9 activation system derived from Aroclor 1254-induced rats.
The positive control compound, benzo(a)pyrene, in a separate experiment with
the same donor's lymphocytes induced chromosome breakage with or without S9
treatment. The response of this donor's lymphocyte chromosomes to chloroform
was a random variation around the control value. The highest breakage level
was at 200 pg/ml with 8 breaks/100 cells compared with 5.5 breaks/100 cells
in the control. However, this difference was not significant in a chi-square
7-21
-------
test. The same problems discussed in the previous section for the SCE study
carried out by Kirkland et al. (1981) apply to their study on chromosome
breakage.
According to Schmid (1976), the bone marrow micronucleus test can be
used to detect clastogens and spindle poisons. Micronuclei are small
nucleus-like elements that contain either chromosomal fragments that
originated from clastogenic events or whole chromosomes that were segregated
by malfunction of the spindle apparatus. Gocke et al. (1981) tested
chloroform in a micronucleus test in mice. The animals were treated with
chloroform at 0 and 24 hr, and bone-marrow smears were prepared at 30 hr.
The purity of the chloroform (purchased from Merck)..was.not stated. Four
mice (two males and two females) were used for each of three doses of
chloroform and for an untreated control group. The animals were each given
two intraperitoneal injections of chloroform for individual treatment dosages
of 238, 476, and 952 mg/kg (2, 4, and 8 mmol/kg). Since the assay was
performed by the method of Schmid (1976). one would assume that the
treatments included the maximum tolerated dose. However, details of dosage
selection were not discussed. Slides were coded, and 1000 polychromatic
erythrocytes were scored per mouse.
The results were as follows (dose, number of micronucleated
polychromatic erythrocytes per 1000 polychromatic erythrocytes): 0 mg/kg,
1.2; 2 x 238 mg/kg, 2.2; 2 x 476 mg/kg, 2.6; 2 x 952 mg/kg, 2.2. Although
the values for the treated groups were higher than the control, the results
were not significant at P < 0.01 according to the Kastenbaum-Bowman test
(Kastenbaum and Bowman, 1970). Three halogenated alkanes were tested
(dichloromethane, 1,1,1-trichloroethane, and chloroform), and all yielded
negative results. Of 30 chemicals tested, only two (pyrogallol and
7-22
-------
hydroqulnone) yielded positive results in the micronucleus test. Positive
controls were not included in the assay, but the positive results for
pyrogallol and hydroquinone indicate that the assay system w,as working.
The micronucleus test was also used by Agustin and Lim-Sylianco (1978)
to study the clastogenic potential of chloroform. The authors tested seven
concentrations of chloroform up to 900 mg/kg in the mouse. The number of
mice used and their sex were not specified. The chloroform was purchased
from Mallinckrodt and was redistilled before use. For each slide, 1000
polychromatic erythrocytes were scored. The authors reported that chloroform
was clastogenic. Results were as follows (dose in mg/kg, number of
micronucleated polychromatic erythrocytes per 1000 polychromatic erythrocytes
+ SE): 0, 4 + 1; 100, 3 ± 1; 200, 5 ± 1; 400, 5 ± 1; 600, 9+2; 700, 17 +
4; 800, 9 + 2; 900, 10 + 2. The authors interpreted the nonlinearity of the
dose-response relationship as evidence that chloroform must be metabolized to
a clastogenic substance. The data, however, are not sufficient to reach any
conclusions regarding the metabolism of chloroform. In the same paper,
Agustin and Lim-Sylianco demonstrated that vitamin E administered 1 hr after
chloroform reduced the number of micronucleated cells observed at 700 mg of
chloroform/kg (17 + 4) to the control level (4 + 1). The significance of
this result is not clear.
The paper by Agustin and Lim-Sylianco gives insufficient details of the
experimental procedures for an independent evaluation of the results. For
example, the number and sex of the animals and positive control data are not
given. This study suggests that chloroform may cause chromosome damage, but
corroborative studies are needed to confirm or refute this interpretation.
A cytogenetic assay for mitotic poisons in grasshopper embryos indicates
that chloroform vapors induce a complete arrest of cells at metaphase (Liang
7-23
-------
et al., 1983). Grasshopper embryos (eggs) at 7-9 days of development were
suspended in 95-ml jars and chloroform (MCB Reagents) at 0.01 (30,000 ppm),
0.05, 0.1, and 0.2 ml per jar was placed on the bottom of the jars. The jars
were closed with screw caps and tightly sealed with petroleum jelly. Embryos
(12 total) exposed to air alone in jars served as untreated controls. Three
randomly selected embryos were exposed in each jar. The jars were incubated
at 24°C for 16 hr. During this time, the chloroform completely vaporized and
the embryos were exposed to the chloroform vapor through respiration. After
treatment, the embryos were made into squash preparations. In each squash
preparation, the mitotic index (MI) and the anaphase to metaphase ratio (A/M)
were determined. MI in each^embryo is the percentage of mitoses (metaphases
and anaphases) estimated by scanning 3000 cells, and A/M is the number of
anaphases present per 100 metaphases. Statistical significance was
determined by chi square analysis. If chloroform completely arrested mitosis
at metaphase, the A/M would be zero; if chloroform does not prevent
interphase cells from entering mitosis, the MI would be elevated compared to
the control.
The results shown in Table 7-2 indicate that chloroform arrested mitosis
at metaphase. In control embryos, all stages of mitosis were observed; the
average MI was 1.0 and the average A/M was 0.65. At a chloroform dose of
0.05 ml per jar, the average A/M was zero and the average MI was 11 times
that of the control. The arrested metaphases showed colchicine-like mitotic
effects (c-mitosis), such as lagging chromosomes and multipolar spindles.
Liang et al. (1983) concluded that chloroform interferes with spindle
microtubules, thereby causing mitotic arrest at metaphase.
7-24
-------
TABLE 7-2. MITOTIC INDEX (MI), ANAPHASE/METAPHASE (A/M), AND PRESENCE (+) OF
COMPLETE C-MITOSIS IN GRASSHOPPER EMBRYOS AFTER EXPOSURE TO CHC13 VAPOR
Amount of CHCls
per jar (ml)
Untreated control
0.01
0.05
0.1
0.2
MI
1.0
(0.7 - 1.6)a
1.2b
11.1*
6.4*
_c
Complete
A/M c-mitosis
0.65
(0.48 - 1.33)
0.49
0.* +
0.* +
-
^Average of 12 embryos (range).
bAverage of 3 embryos at each concentration.
CEmbryo death induced.
*Significantly different from control, p_ < 0.005.
SOURCE: Liang et al. (1983)
In summary, of the three available studies on the clastogenic effects of
chloroform, one was negative, one was negative but inconclusive, and one was
positive but inconclusive. More studies are needed before a conclusion is
reached as to whether chloroform is clastogenic. The data reported by Liang
et al. (1983) suggest that chloroform affects spindle microtubules.
7.7. SUGGESTED ADDITIONAL TESTING
Additional studies are needed to measure covalent binding of ^CHCl3 to
DNA. The ability of chloroform to cause DNA damage should be studied further
in UDS and SCE tests. Suggested testing includes measurement of UOS i_n vivo
in mice and i_n vitro in both rat and mouse hepatocytes. The issues raised by
Lonati-Galligani et al. (1983) should be taken into consideration when
choosing procedures for the UDS assays. Additional studies are needed to
corroborate or refute the study of Call en et al. (1980) in yeast. Further
testing for the ability of chloroform to cause changes in chromosome
structure and number is needed, particularly in i_n vivo systems.
7-25
-------
7.8. SUMMARY AND CONCLUSIONS
It has been demonstrated that chloroform can be metabolized in vivo and
In vitro to a substance (presumably phosgene) that interacts with protein and
lipid. The potential for metabolically activated chloroform to bind to DNA
cannot be determined from the available studies, but, if binding to DNA does
occur, it would be at a very low level.
The majority of the assays for genotoxicity have yielded negative
results. However, many of these results are inconclusive because of
inadequacies in the experimental protocols. The major problem has been the
use of exogenous activation systems (i.e., S9 mix). In none of the studies
was it shown that chloroform was activated or metabolized by the activation
system used. Metabolism of 2-aminoanthracene or vinyl compounds (used as
positive controls) is an inadequate indication that the activation system can
metabolize chloroform, because these substances are not halogenated alkanes.
A better indication that the activation system is sufficient for metabolism
of chloroform may be to show that it metabolizes ^CHClj to intermediates
that bind to macromolecules. A second problem in the use of exogenous
activation systems is the possibility that highly reactive metabolites may
react with microsomal or membrane lipid or protein before reaching the DNA of
the test organism. Another problem in in vitro tests is that adequate
precautions are sometimes not taken to prevent the escape of volatilized
chloroform.
On the basis of presently available data, no definitive conclusion can
be reached concerning the mutagenicity of chloroform. However, evidence from
studies measuring binding to macromolecules, DNA damage, and mitotic arrest
suggest that chloroform may be mutagenic. Alternatively, because recent
studies on the mechanism of action of tumor promoters suggest that promoters
7-26
-------
can damage DNA [see Marx (1983) for review], chloroform may promote
carcinogenesis rather than initiate it.
7.9 REFERENCES FOR CHAPTER 7
Agustin, J.S.; Lim-Sylianco, C.Y. (1978) Mutagenic and clastogenic
effects of chloroform. Bull. Phil. Biochem. Soc. 1:17-23.
Callen, D.F.; Wolf, C.R.; Phil pot, R.M. (1980) Cytochrome P-450
mediated genetic activity and cytotoxicity of seven halogenated
aliphatic hydrocarbons in Saccharom.yces cerevisiae. Mutat. Res.
77:55-63.
de Serres, F.J.; Ashby, J. eds. (1981) Evaluation of short-term tests
for carcinogens. Progress in Mutation Research, Vol. I. Elsevier/North
Holland.
Diaz Gomez, M.I.; Castro, J.A. (1980a) Nuclear activation of carbon
tetrachloride and chloroform. Res. Commun. Chem. Pathol. Pharmacol.
27:191-194.
Diaz Gomez, M.I.; Castro, J.A. (1980b) Covalent binding of chloroform
metabolites to nuclear proteins-no evidence for binding to nucleic
acids. Cancer Lett. 9:213-218.
DiRenzo, A.8.; Gandolfi, A.J.; Sipes, I.G. (1982) Microsomal
bioactivation and covalent binding of aliphatic halides to DNA.
Toxicol. Lett. 11:243-252.
Gocke, E.; King, M.-T.; Eckhardt, K.; Wild, D. (1981) Mutagenicity of
cosmetics ingredients licensed by the European communities. Mutat. Res.
90:91-109.
Kastenbaum, M.A.; Bowman, K.O. (1970) Tables for determining the
statistical significance of mutation frequencies. Mutat. Res.
9:527-549.
Kirkland, D.J.; Smith, K.L.; Van Abbe, N.J. (1981) Failure of
chloroform to induce chromosome damage or sister-chromatid exchanges in
cultured human lymphocytes and failure to induce reversion in
Escherichia coli. Fd. Cosmet. Toxicol. 19:651-656.
Kirk-Othmer Encyclopedia of Chemical Technology, Second Edition, Supplement
volume. (1971) Interscience Publishers; pp. 674-683.
Land, P.C.; Owen, E.L.; Linde, H.W. (1981) Morphologic changes in mouse
spermatozoa after exposure to inhalational anesthetics during early
spermatogenesis. Anesthesiology 54:53-56.
7-27
-------
Liang, J.C.; Hsu, T.C.; Henry, J.E. (1983) Cytogenetic assays for
mitotic poisons: The grasshopper embryo system for volatile liquids.
Mutat. Res. 113:467-479.
Lonati-Galligani, M.; Lohman, P.H.M.; Berends, F. (1983) The validity
of the autoradiographic method for detecting DMA repair systhesis in rat
hepatocytes in primary culture. Mutat. Res. 113:145-160.
Marx, J.L. (1983) Do tumor promoters affect DNA after all? Science
219:158-159.
Mirsalis, J.C.; Tyson, C.K.; Butterworth, B.E. (1982) Detection of
genotoxic carcinogens in the i_n vivo - in vitro hepatocyte DNA repair
assay. Environ. Mutagen. 4:553-562.
Morimoto, K.; Koizumi, A. (1983) Trihalomethanes induce sister
chromatid exchanges in human lymphocytes i_n vitro and mouse bone marrow
cells in vivo. Environ. Res. 32:72-79.
National Cancer Institute (NCI). (1976) Report on Carcinogenesis Bioassays
of Chloroform. NTIS PB-264-018. Springfield, VA: National Technical
Information Service.
Reitz, R.H.; Quast, J.F.; Stott, W.T.; Watanabe, P.G.; Gehring, P.J.
(1980a) Pharmacokinetics and macromolecular effects of chloroform in
rats and mice: Implications for carcinogenic risk estimation. In:
Jolley, R.L.; Brungs, W.A.; Cumming, R.B., eds. Water Chlorination:
Environmental Impact and Health Effects, Vol. 3, pp. 983-992.
Reitz, R.H.; Watanabe, P.G.; McKenna, M.J.; Quast, J.F.; Gehring, P.J.
(1980b) Effects of vinylidine chloride on DNA synthesis and DNA repair
in the rat and mouse: a comparative study with dimethylnitrosamine.
Toxicol. Appl. Pharmacol. 52:357-370.
Schmid, W. (1976) The micronucleus test for cytogenetic analysis. In:
Hollaender, A., ed. Chemical Mutagens. Vol. 4, pp. 31-53. New York:
Plenum Press.
Simmon, V.F.; Kauhanen, K.; Tardiff, R.G. (1977) Mutagenic activity of
chemicals identified in drinking water. In: Scott, D.; Bridges, B.A.;
Sobels, F.H., eds., Progress in Genetic Toxicology, pp. 249-258. New
York: Elsevier/North Holland Biomedical Press.
Sturrock, J. (1977) Lack of mutagenic effect of halothane or chloroform on
cultured cells using the azaguanine test system. Br. J. Anaesth.
49:207-210.
Topham, J.C. (1980) Do induced sperm-head abnormalities in mice
specifically identify mammalian mutagens rather than carcinogens?
Mutat. Res. 74:379-387.
7-28
-------
Uehleke, H.; Werner, T.; Greim, H.; Kramer, M. (1977) Metabolic
activation of haloalkanes and tests i_n vitro for mutagenicity.
Xenobiotica 7:393-400.
Van Abbe, N.J.; Green, T.J.; Jones, E.; Richold, M.; Roe, F.J.C. (1982)
Bacterial mutagenicity studies on chloroform j_n vitro. Fd. Chem. Toxic
20:557-561.
Weisburger, J.H.; Williams, G.M. (1982) Metabolism of chemical
carcinogens. In: Becker, F.F., ed. Cancer, a Comprehensive Treatise,
2nd Edition, vol. 1, pp. 241-333. New York: Plenum Press.
White, A.E.; Takehisa, S.; Eger, E.I.; Wolff, S.; Stevens, W.C. (1979)
Sister chromatid exchanges induced by inhaled anesthetics.
Anesthesiology 50:426-430.
Williams, G.M. (1981) Liver culture indicators for the detection of
chemical carcinogens. In: Short-term Tests for Chemical Carcinogens,
Stich, H.F.; and San, R.H.C., eds., pp. 275-289. New York: Springer
Verlag.
Wyrobek, A.; Bruce, W.R. (1978) The induction of sperm-shape
abnormalities in mice and humans. In: Hollaender, A., ed. Chemical
Mutagens. Vol. 5, pp. 257-285. New York: Plenum Press.
7-29
-------
Van Abbe, N.J.; Green, T.J.; Jones, E.; Richold, M.; Roe, F.J.C. (1982)
Bacterial mutagenicity studies on chloroform i_n vitro. Fd. Chem. Toxic,
20:557-561.
Weisburger, J.H.; Williams, 6.M. (1982) Metabolism of chemical
carcinogens. In: Becker, F.F., ed. Cancer, a Comprehensive Treatise,
2nd Edition, vol. 1, pp. 241-333. New York: Plenum Press.
White, A.E.; Takehisa, S.; Eger, E.I.; Wolff, S.; Stevens, W.C. (1979)
Sister chromatid exchanges induced by inhaled anesthetics.
Anesthesiology 50:426-430.
Williams, G.M. (1981) Liver culture indicators for the detection of
chemical carcinogens. In: Short-term Tests for Chemical Carcinogens,
Stich, H.F.; and San, R.H.C., eds., pp. 275-289. New York: Springer
Verlag.
Wyrobek, A.; Bruce, W.R. (1978) The induction of sperm-shape
abnormalities in mice and humans. In: Hollaender, A., ed. Chemical
Mutagens. Vol. 5, pp. 257-285. New York: Plenum Press.
7-30
-------
8. CARCINOGENICITY
8.1. ANIMAL STUDIES
The carcinogenicity of chloroform has been evaluated in mice, rats, and
dogs. Evidence for carcinogenic activity by chloroform includes induction of
renal epithelial tumors, mostly malignant, in male Osborne-Mendel rats
(National Cancer Institute [NCI], 1976 and Jorgenson et al.,1985),
hepatocellular carcinomas in male and female B6C3F1 mice (NCI, 1976), kidney
tumors in male ICI mice (Roe et a!., 1979), and hepatomas in female strain A
mice (Eschenbrenner and Miller, 1945) and NIC mice (Rudali, 1967). Capel et
al. (1979) demonstrated an ability of chloroform to promote growth and
metastasis of murine tumors. Chloroform was not shown to be carcinogenic in
(C57 x DBA2 Fl) mice (Roe et al., 1968), female Osborne-Mendel rats (NCI,
1976), female ICI mice and male mice of the CBA, C57BL, and CF/1 strains (Roe
et al., 1979), female B6C3F1 mice when administered in drinking water
(Jorgenson et al., 1985), male and female Sprague-Dawley rats (Palmer et al.,
1979), and male and female beagle dogs (Heywood et al., 1979). Chloroform
was negative in a pulmonary tumor induction bioassay in male strain A/St mice
(Theiss et al., 1977). Chloroform in liquid solution did not induce
transformation of baby Syrian hamster kidney cells (BHK-21/C1 13) i_n vitro.
Under the conditions of the carcinogenicity bioassays showing carcinogenic
activity for chloroform specifically in kidney and liver of mice and rats,
the conclusion can be made, by applying either the EPA proposed Carcinogen
Risk Assessment Guidelines or the IARC classification approach for
carcinogens, that there is sufficient evidence for the carcinogenicity of
chloroform in experimental animals.
8-1
-------
8.1.1. Oral Administration (Gavaqe): Rat
8.1.1.1. National Cancer Institute (1976)—A carcinogenesis bioassay on
chloroform in Osborne-Mendel rats was reported by the NCI (1976). The
chloroform product (Aldrich Chemical Company, Milwaukee, Wisconsin) was
analyzed at the carcinogenesis bioassay laboratory and was shown by gas-
liquid chromatography, flame ionization detection, and infrared spectrometry
to be 98 percent pure chloroform and 2 percent ethyl alcohol (stabilizer).
fresh chloroform solutions in corn oil were prepared each week and stored
under refrigeration.
Fifty animals of each sex were randomly assigned to each of two
chloroform dose test groups. Treated animals were compared with matched
vehicle-control groups (20 males and 20 females) and with a vehicle colony
control group (99 males and 98 females). This latter group included the
matched control group and three other control groups put on study within
3 months of the matched control group. Matched control and treated animals
were housed in the same room, and colony controls were housed in two
different rooms.
Doses selected for the long-term carcinogenicity study in rats were
estimated as the doses maximally tolerated by the animals and one-half the
maximally tolerated doses based on survival, body weights, clinical signs,
and necropsy examinations in a preliminary toxicity test in each sex. The
chloroform was given by g'avage for 6 weeks followed by an observation period
of 2 weeks without treatment. The chronic study began with 52-day-old rats
and ended with sacrifice of the survivors at 111 weeks. Chloroform was
administered to the rats in corn oil by gavage 5 days each week during the
initial 78 weeks. Doses of 90 and 180 mg/kg/day were administered to male
rats throughout the chronic study; however, since initial doses of 125 and
8-2
-------
250 mg/kg/day were reduced to 90 and 180 mg/kg/day at 22 weeks, doses given
to female rats were expressed as time-weighted averages of 100 and
200 mg/kg/day.
Body weights and food consumption were monitored weekly for the first 10
weeks and monthly, thereafter. Animals were observed twice daily. Decedents
and survivors were necropsied, and tissues and organs were examined
microscopically.
In matched control and both dose groups, at least 50 percent of the male
and female rats survived as long as 85 and 75 weeks, respectively. Seven
matched control, 24 low-dose, and 14 high-dose males, and 15 matched control,
22 low-dose, and 14 high-dose females survived until the end of the study.
Only one control male rat died before 90 weeks; the increase in death rate of
control males after 90 weeks was, according to the NCI (1976) report,
"probably due to respiratory and renal conditions." Overall survival was
reduced in treated females and high dose males relative to controls
(Figure 8-1).
Appearance and behavior among groups were generally similar, but
hunching, urine stains on the lower abdomen, redness of eyelids, and wheezing
were noted in treated animals early in the study. Food consumption was
reported as slightly less in treated animals, but data were not provided.
Decreased body weight gain was evident in both treatment groups in both sexes
of rats. Initial mean body weights for all groups were about 175 g for
females and 250 g for males. By 50 weeks, mean body weights were
approximately 400 g in control, 350 g in low-dose, and 330 g in high-dose
females; by 100 weeks, mean body weights were about 375 g in all groups of
females. In males, mean body weights were about 640 g in the control group,
8-3
-------
CO
1.00
0.90
0.80
_i
^ 0.70
0.60
oc
D
LL
O
Ł 0.50
m
g 0.40
OC
0_
0.30
0.20
0.10
0.00
n Control
O Low Dose
A High Dose
Male Rats
I
-I'l-
0 10 20 40 60 80 100
TIME ON STUDY (Weeks)
120
1.00
0.90
0.80
> 0.70
cc
D
w 0.60
LL.
O
Ł 0.50
m
<
2 0.40
O
cc
Q.
0.30
0.20
0.10
0.00
D Control
O Low Dose
A High Dose
Female Rats
0 10 20 40 60 80 100 120
TIME ON STUDY (Weeks)
Figure 8-1. Survival curves for Fischer 344 rats in a carcinogenicity bioassay on chloroform.
Source: NCI (1976).
-------
550 g in the low-dose group, and 500 g in the high-dose group by 50 weeks; by
100 weeks, mean body weights were approximately 500 g in all groups.
A statistically significant (P < 0.05) dose-related increase in renal
epithelial tumors of tubular cell origin was found in treated male rats
(Table 8-1). The epithelial tumors were described as follows: Of 13 tumors
in high-dose males, 10 were carcinomas and three were adenomas; two
carcinomas and two adenomas comprised the tumors found in four low-dose
males; one renal epithelial carcinoma and one squamous cell carcinoma from
renal pelvic transitional epithelium were noted in two high-dose females.
One low-dose male had both a malignant mixed tumor and a tubular cell adenoma
in the left kidney, and a high-dose male had a tubular cell carcinoma and a
tubular cell adenoma in the right kidney. Renal epithelial carcinomas were
large and poorly circumscribed, and they infiltrated surrounding normal
tissue. Renal epithelial adenomas were circumscribed and well-
differentiated. Additional kidney tumors included malignant mixed tumors in
two low-dose and two colony control males and hamartomas in one low-dose
male, one high-dose male, and one colony control male.
Although a statistically significant (P < 0.05) increase in thyroid
tumors was reported in both treatment groups of female rats as compared with
colony controls, the toxicologic significance of this finding is
controversial because C-cell tumors and follicular cell tumors, which have
different embryonic origins and different physiologic functions, are combined
in the incidences described in Table 8-2; the majority of tumors were
adenomas; the spontaneous incidence of thyroid tumors in Osborne-Mendel
females is variable as stated, in the NCI (1976) bioassay report although
without presentation of historical data; and the increased incidence of
thyroid tumors in treated females is not statistically significant (P > 0.05)
8-5
-------
00
CTi
TABLE 8-1. EFFECT OF CHLOROFORMS ON KIDNEY EPITHELIAL TUMOR INCIDENCE IN OSBORNE-MENDEL RATS
(NCI, 1976)
Control sb
Treatment Colony
Kidney tumor 0/99(0%)
incidence^
P valueh
Time to —
first tumor
(weeks)
Survival at 26%
terminal
sacrifice
(111 weeks)
Male
Female
Dose (mg/kg/day)c Controlsb Dose (mg/kg/day)c
Matched 90
0/19(0%) 4/50(8%)e
0.266
102
37% 48%
180 Colony
12/50(24%)f 0/98(0%)
0.0141
80
28% 51%
Matched 100 200
0/20(0%) 0/49(0%) 2/48(4%)9
0.495
102
75% 45% 29%
^Chloroform in corn oil administered by gavage 5 times per week for 78 weeks.
t>Colony controls consist of four vehicle-control groups, including matched controls, given corn oil.
CDoses are time-weighted averages.
dAnimals with tumor/animals examined.
eTwo with tubular cell adenocarcinoma and two with tubular cell adenoma.
fTen with tubular cell adenocarcinoma and two with tubular cell adenoma.
90ne with tubular cell adenocarcinoma and one with squamous cell carcinoma in the renal pelvis.
hFisher's Exact Test, compared with matched controls.
"iFor adenocarcinomas alone, P value is 0.03.
-------
CO
I
TABLE 8-2. EFFECT OF CHLOROFORM ON THYROID TUMOR INCIDENCE IN FEMALE OSBORNE-MENDEL RATS
(NCI, 1976)
Dosea.b
(mg/kg/day)
0 (matched)h
0 (colony)h
100
200
Follicular cell
tumorsc
Incidencef
1/19(5%)
1/98(1%)
2/49(4%)
6/49(12%)
C-cell
tumors^
Incidence
0/19(0%)
0/98(0%)
6/49(12%)
4/49(8%)
Incidence
1/19(5%)
8/49(16%)
10/49(20%)
Total
P valueg
0.216
0.121
tumors6
Time to first tumors
(weeks)
110
110
73
49
^Chloroform in corn oil administered by gavage 5 times per week.
bTime-weighted average doses.
CAdenomas except for carcinoma in one low-dose and two high-dose animals.
^Adenomas except for carcinoma in one high-dose animal.
eSee text.
fAnimals with tumors/animals examined.
gFisher's Exact Test, compared with matched controls.
hColony controls consist of four vehicle-control groups, including matched controls, given corn
oil.
-------
when compared with data for the matched controls only. No significant (P <
0.05) differences for other tumor types were apparent among the groups of
rats. Four rats were lost (missing or autolyzed) for pathology.
Non-neoplastic lesions described as treatment-related include necrosis
of liver parenchyma, epithelial hyperplasia in the urinary bladder, and
hematopoiesis in spleen. Inflammatory pulmonary lesions characteristic of
pneumonia were found in all groups, but the severity and incidence of these
lesions were stated (data not reported) to have been greater in treatment
groups.
Under the conditions of this bioassay. chloroform treatment
significantly (P < 0.05) increased the incidence of renal epithelial tumors
in male Osborne-Mendel rats. Although the number of matched vehicle controls
was low, the use of pooled colony controls gives additional support for
treatment-related effects. Moreover, historical control incidence of renal
epithelial tumors in Osborne-Mendel rats was reported as rare.
Lower survival rates and body weights in treated rats than in matched
controls provide evidence that the chloroform doses used may have been toxic
to the rat strain used in this study. A more precise estimate of dose-
response perhaps could have been obtained if additional lower doses had been
given, and if constant doses rather than time-weighted averages had been used
for the females as was done for the males. Treated animals were housed in
the same room as rats treated with other volatile compounds (1,1,2,2-
tetrachloroethane, 3-chloropropene, ethylene dibromide, carbon
tetrachloride); however, since controls were in the same room as treated
animals, and oral chloroform doses probably were much higher than ambient
levels of other volatiles, the likelihood that the other volatile compounds
were responsible for the observed results is considered to be low.
8-8
-------
Additionally, these other volatile compounds did not induce kidney tumors in
Osborne-Mendel rats (NCI, 1976; Weisburger, 1977). It should be noted that
ambient levels of volatiles in the animal quarters were not measured.
8.1.1.2. Palmer et al. (1979)--Palmer et al. (1979) reported carcinogenicity
studies on chloroform in Sprague-Dawley rats. The chloroform dose
preparations were concocted in toothpaste, as described in Table 8-3 herein
for the Roe et al. (1979) study, and administered by gavage. Dose levels of
15, 75, and 165 mg CHCls/kg/day were selected for the carcinogenicity study
based on results of a preliminary range-finding study suggesting the lowest
toxic dose, indicated by liver and kidney changes, as 150 mg/kg/day.
TABLE 8-3. TOOTHPASTE FORMULATION FOR CHLOROFORM ADMINISTRATION
(Roe et al., 1979)
Ingredient Percentage w/w
Chloroform 3.51
Peppermint oiia 0.25
Eucalyptoia 0.50
Glycerol 39.35
Carragheen gum 0.45
Precipitated calcium carbonate 48.53
Sodium lauryl sulphate 1.16
Sodium saccharin 0.03
White mineral oil 1.10
Water 5.12
Total 100.00
^Essential oil flavor components.
8-9
-------
An initial carcinogenicity study was conducted in which 25 rats of each
sex per group received one of the selected doses in toothpaste containing
essential oils (flavor components), indicated in Table 8-3, 6 days per week.
A concurrent control group of 75 males and 75 females was administered
toothpaste without chloroform or essential oils. A second carcinogenicity
study was conducted in which 50 male and 50 female specific pathogen-free
(SPF) Sprague-Dawley rats were dosed with 60 mg chloroform/kg/day in
toothpaste with essential oils 6 days per week, and 50 control rats of each
sex were given toothpaste without chloroform but with essential oils.
Body weights were measured weekly, and food consumption was recorded.
Body weights were initially 180 to 240 g for males and 130 to 175 g for
females. Blood and urine analyses were performed in the first study, and
serum and erythrocyte cholinesterase activities and other serum enzyme
activities were monitored in the second study. All animals were necropsied,
and tissues and organs were examined histopathologically. Adrenal glands,
kidneys, livers, lungs, and spleens were weighed.
Chloroform was not carcinogenic in these studies. Significant
(P < 0.05) body weight loss in high-dose males in the first study (data not
reported) and maximal body weight gain of approximately 370 g in control
males, 330 g in treated males, 220 g in control females, and 180 g in treated
females in the second study suggest an effect from chloroform treatment.
Other than a 40 percent reduction of plasma cholinesterase levels and slight
decreases in serum glutamic-pyruvic transaminase and serum alkaline
phosphatase in treated females, additional toxic effects from chloroform
treatment were not evident.
Low survival, attributed to respiratory disease, was apparent in both
studies. The initial study was terminated at 52 weeks; 50 percent of the
8-10
-------
animals in all groups had died by 52 weeks in the second study, which was
ended at 95 weeks. Except for 48 control females in the initial study, no
more than 18 animals were alive in each group at the conclusion of either
study. Although carcinogenic activity for chloroform was not observed, these
studies on Sprague-Dawley rats are weakened by the high early mortality in
both control and treated animals.
8.1.2. Oral Administration (Gavage): Mouse
8.1.2.1. National Cancer Institute (1976)—The carcinogenicity of chloroform
in B6C3F1 mice was evaluated by the NCI (1976). The chloroform product and
chloroform solutions in corn oil were those used in the NCI (1976)
carcinogenesis bioassay in rats previously described in Section 8.1.1.1.
Each of two dose treatment groups was composed of 50 males and 50
females. Treated mice were compared with matched vehicle control groups (20
males and 20 females) put on study 1 week earlier, and with vehicle colony
control groups (77 males and 80 females), which included the matched control
group and three other control groups put on study within 3 months of the
matched control group. All control mice were housed in the same room with
treated mice.
Maximally tolerated doses and one-half the maximally tolerated doses for
the long-term'carcinogenicity study in mice were estimated from a preliminary
toxicity test conducted as described for the NCI (1976) rat study. Mice were
35 days of age at the start of the chronic study, and the study was concluded
with the sacrifice of survivors at 92-93 weeks. Chloroform in corn oil was
administered by gavage 5 days per week during the first 78 weeks. Initial
dose levels of 200 and 100 mg/kg/day for males and 400 and 200 mg/kg/day for
females were raised to 300 and 150 mg/kg/day for males and 500 and 250
mg/kg/day for females at 18 weeks. Thus, doses expressed as time-weighted
8-11
-------
averages for the entire study were 138 and 277 mg/kg/day for males and 238
and 477 mg/kg/day for females.
Survival in mice was similar among groups except for high-dose females,
At least 50 percent of the animals in each group survived as long as 85
weeks. Ten matched control, 33 low-dose, and 30 high-dose males, and 15
matched control, 34 low-dose, and 9 high-dose females survived for the
duration of the study. All but two deaths in high-dose females occurred after
70 weeks.
Body weight gain among groups was comparable. Male and female mice
initially weighed about 18 and 15 g, respectively. Mean body weights at 50
weeks were approximately 35 g in males and 28 g in females, and these levels
were generally sustained throughout the remainder of the study. Food
consumption was stated to have been equivalent among groups. Appearance and
behavior among groups were similar except for bloating and abdominal
distension noted in treated animals beginning after 42 weeks of treatment.
Statistically significant (P < 0.05) dose-related increases in
hepatocellular carcinomas in both treatment groups in both sexes of mice were
observed (Table 8-4). Various histopathologic types of hepatocellular
carcinomas were observed. The hepatocellular carcinoma metastasized to the
lung in two low-dose males and two high-dose females, and to the kidney in
one high-dose male. Twenty animals were reported as missing or autolyzed,
and therefore were not included in the pathology report.
Nonneoplastic lesions in mice attributed to treatment were relatively
few. These include nodular hyperplasia of the liver in 10 low-dose males,
six low-dose females, and one high-dose female; and liver necrosis in one
low-dose male, four low-dose females, and one high-dose female. Nine high-
dose females with hepatocellular carcinoma had cardiac atrial thrombosis.
8-12
-------
TABLE 8-4. EFFECTS OF CHLOROFORMa ON HEPATOCELLULAR CARCINOMA INCIDENCE IN B6C3F1 MICE (NCI, 1976)
CD
I
Male Female
Controlsb Dose (mg/kg/day)c Controlsb
Treatment Colony
Hepato- 5/77(8%)
cellular
carcinoma
incidence^
P valuee
Time to first 72
tumor (weeks)
Survival at 48%
terminal
sacrifice
(111 weeks)
Matched 138 277 Colony Matched
1/18(6%) 18/50(36%) 44/45(98%) 1/80(1%) 0/20(0%)
0.011 3.13x10-13
72 80 54 90
50% 65% 65% 81% 75%
Dose (mg/kg/day)c
238
36/45(80%)
4x10-10
66
75%
477
39/41(95%)
3.7x10-14
67
20%
^Chloroform in corn oil administered by gavage 5 times per week for 78 weeks.
bColony controls consist of four vehicle-control groups, including matched controls, given corn oil.
CDoses are time-weighted averages.
dAnimals with tumor/animals examined.
^Fisher's Exact Test, compared with matched controls.
-------
Kidney inflammation was diagnosed in 10 matched control males, two low-dose
males, and one high-dose male.
Under the conditions of this bioassay, chloroform treatment
significantly (P < 0.05) increased the incidence of hepatocellular carcinoma
in male and female B6C3F1 mice. Although the number of matched vehicle
controls was low, the use of pooled colony controls gives additional support
for treatment-related effects. Moreover, historical control incidence of
hepatocellular carcinomas in B6C3F1 mice was reported as 5-10 percent in
males and 1 percent in females.
A more precise estimate of dose-response perhaps could have been
obtained if additional lower doses had been used and if constant doses rather
than time-weighted averages had been used. Treated animals were housed in
the same room as animals treated with other volatile compounds*; however,
since 1) controls were in the same room as treated animals, 2) oral
chloroform doses probably would have been much higher than ambient levels of
other volatiles, 3) the cages had filters to limit the amount of chemical
released into the ambient air, 4) the total room air was exchanged 10 to 15
times per hour, and 5) dosing was done in another room under a large hood,
the likelihood that the other volatile compounds were responsible for the
observed results is considered to be low. It should be noted that ambient
levels of volatiles in the animal quarters were not measured.
8.1.2.2. Roe et al. (19791—Roe et al. (1979) studied the carcinogenicity of
chloroform in toothpaste in four strains of mice (C57BL, CBA, CF/1, and ICI).
*l,l,2,2-tetrachloroethane, 3-chloropropene, chloropicrin,
1,1-dichloroethane, trichloroethylene, sulfolene, iodoform, ethylene
dichloride, methyl chloroform, 1,1,2-trichloroethane, tetrachloroethylene,
hexachloroethane, carbon disulfide, trichlorofluoromethane, carbon
tetrachloride, ethlene dibromide, dibrotnochloropropane.
8-14
-------
The toothpaste formulation used is presented in Table 8-3. The chloroform
product was described as British Pharmacopoeia grade which was not
contaminated with other haloalkanes or phosgene. Toothpaste was prepared
fresh each month. Chloroform in arachis oil vehicle was also tested in ICI
mice.
Dose levels for the carcinogenicity studies were selected based on
results of a 6-week preliminary range-finding study in male and female
Schofield mice which indicated moderate weight gain reduction at the lowest
dose considered toxic, which was 60 mg CHCl3/kg/day. Three different
carcinogenicity studies were conducted in which mice, initially no more than
10 weeks oW, were given chloroform by gavage 6 days per week for 80 weeks
followed by observation for 13 to 24 weeks. In one study, 52 male and 52
female ICI mice per dose group were given 17 or 60 mg/kg/day of chloroform in
toothpaste and compared with 100 ICI mice of each sex concurrently given
toothpaste without chloroform, peppermint oil, or eucalyptol. A second
study, confined to male ICI mice, included 52 untreated mice, 260 mice given
toothpaste alone without chloroform, eucalyptol, or peppermint oil, and
groups of 52 mice each given, in toothpaste, 60 mg CHCl3/kg/day, eucalyptol
up to 32 mg/kg/day, or peppermint oil up to 16 mg/kg/day; treatment with
chloroform, eucalyptol, or peppermint oil was performed in the absence of the
other two compounds. In the third study, groups of 52 male mice of each of
the C57CL, CBA, CF/1, and ICI strains were given 60 mg CHCl3/kg/day in
toothpaste and compared with concurrent vehicle-control groups of 52 mice
each, and with 100 untreated ICI mice. Fifty-two ICI male mice given 60 mg
CHCl3/kg/day in arachis oil, and concurrent control mice given arachis oil
alone, were also evaluated in the third study.
8-15
-------
Body weights were recorded in each study, and food consumption was
estimated in the second and third studies. In each study, the animals were
necropsied, and tumors and other lesions as well as routine tissues and
organs were examined histopathologically. Adrenal glands, kidneys, livers,
lungs, and spleens were weighed.
Although the authors stated (data were not reported) that body weight
gain was poorer in each treatment group than in controls in the third study
on the four mouse strains, differences in survival, body weights, and food
consumption between control and treatment groups were not statistically
(P < 0.05) significant, either as shown with data or as stated by the authors
without data. Median survival was > 73 weeks for all groups in jzhe two
studies on ICI mice alone; by terminal sacrifice in the study on four strains
(survival patterns were not reported), 52 to 79 percent of the C57BL and CBA
mice and 12 to 31 percent of the CF/1 and ICI mice were alive. Liver and
kidney weights were slightly lower (data not reported) in male ICI mice given
chloroform in toothpaste. The incidences of tumors and lesions between
control and chloroform-treated animals that were significantly (P < 0.05)
different included 1) increased kidney tumor incidences in treated male ICI
mice, as shown in Table 8-5, and 2) a significantly (P < 0.001, chi-square
test) higher incidence of moderate to severe kidney "changes" in treated CBA
and CF/1 males than in corresponding controls and of moderate to severe
kidney disease (P < 0.05, chi-square test) in ICI males given CHC13 in
arachis oil than in arachis oil controls, as described by the authors without
presentation of data. Results in Table 8-5 indicate more effective induction
of kidney tumors by chloroform in arachis oil than by chloroform in
toothpaste. Kidney tumors were not found in C57BL, CBA and female ICI mice,
and malignant kidney tumors were diagnosed in two control and one treated
8-16
-------
TABLE 8-5. KIDNEY TUMOR INCIDENCE IN MALE ICI MICE TREATED WITH CHLOROFORM
(Adapted from Roe et a!., 1979)
Number of mice with kidney tumors
Dose Group
Numbers of
mice examined
histologically Benign Malignant
Total
First study
Vehicle-control^
17 mg CHCl3/kg/dayb
60 mg CHCl3/kg/dayb
Second study
Untreated control
Vehicle-control^
60 mg CHCl3/kg/dayC
Third study
Untreated control
Vehicle-control^
60 mg CHCla/kg/daye
72
37
38
45
237
49
83
49
47
0
0
1
6
7h
0
1
2
0
0
0
0
21
0
0
3
0
0
8J
1
6
9J
0
1
5
Vehicle-control
60 mg CHCl3/kg/dayg
50
48
1
3
0
gh
0
12J
toothpaste base vehicle without chloroform, eucalyptol, and peppermint
oil.
^Chloroform given in toothpaste base with eucalyptol and peppermint oil.
cChloroform given in toothpaste base without eucalyptol and peppermint oil
dToothpaste base vehicle without chloroform.
eChloroform given in toothpaste base.
fArachis oil.
gChloroform given in arachis oil.
hstatistically significant versus vehicle-control (P < 0.05).
^Statistically significant versus vehicle-control (P < 0.01).
JStatistically significant versus vehicle-control (P < O.OOlJ).
CF/1 mice. Malignant kidney tumors were identified as hypernephromas, and
benign kidney tumors were characterized as cortical adenomas. The increased
incidences of both the malignant tumors and the benign tumors were each
separartely statistically significant, as well as when they were combined.
8-17
-------
Eucalyptol and peppermint oil were not toxic to male ICI mice in these
studies.
Results of the studies by Roe et al. (1979) show the ability of
chloroform to induce kidney tumors in male ICI mice. The stronger induction
of kidney tumors by chloroform in arachis oil compared with chloroform in
toothpaste may reflect an effect of the dosing vehicle on chloroform
absorption and resulting peak blood and tissue levels. Moore et al. (1982)
demonstrated greater severity of acute toxicity and regenerative changes in
kidneys of male CFLP mice given single gavage doses of 60 mg CHC^/kg when
corn oil rather than toothpaste was the dosing vehicle. Kidney pathology was
noted in treated animals in the study with four strains of mice; however,
although poorer body weight gain reported for treated mice in each of the
strains would suggest that a maximum tolerated dose was being approached, the
observation that survival, body weights, and other pathological changes
between control and treated mice in each of the four strains were not
significantly (P < 0.05) different also suggests that higher doses could have
been administered to more strongly challenge the mice for carcinogenicity.
Since mice were as old as 10 weeks at the start of the studies, it is evident
that treatment could have been started when the mice were younger to cover a
greater portion of their lifespan during growth. A fuller evaluation of
chloroform carcinogenicity could have been made if female mice had also been
included in each study.
8.1.2.3. Eschenbrenner and Miller (1945)—An early study on chloroform
hepatoma induction in mice was described by Eschenbrenner and Miller (1945).
Strain A mice, initially 3 months old, with a notably low historical
spontaneous hepatoma rate of <1 percent at 16 months of age were selected for
treatment. "Chemically pure" chloroform was used, but a chemical analysis
8-18
-------
was not indicated in the report. Test groups of five males and five females
each were treated by gavage with doses of 2.4, 1.2, 0.6, 0.3, or 0.15 g/kg of
chloroform in olive oil. Controls received olive oil alone.
In the study of hepatoma induction, mice were dosed once every 4 days
for a total of 30 doses. When 8 months old, the mice were examined for
hepatomas at one month after the last dose; however, these animals were given
an additional dose of chloroform 24 hours before necropsy. Tissues and
organs were examined histopathologically. Liver necrosis also was found in
mice given a single gavage treatment of one of the indicated doses of
chloroform (one male and two females per group) 24 hours before removal of
liver for microscopic evaluation.
Incidences of liver and kidney necrosis and hepatomas are shown in
Table 8-6. Liver necrosis was noted in both sexes in the three highest dose
groups. Males in all treatment groups developed kidney necrosis, whereas
kidney necrosis was not perceptible in females. No males in the three
highest-dose groups and no females in the highest-dose group survived the
study. All deaths occurred by 48 hours after the second administration of
chloroform. All surviving females dosed with 0.6 or 1.2 g CHCl3/kg had
hepatomas.
In the experiment to test the ability of a single dose of chloroform to
produce tissue necrosis, there was sharp distinction between normal and
necrotic cells in liver. Doses of 2.4 and 1.2 g/kg produced extensive
necrosis in all liver lobules, and the 0.6 g/kg dose produced necrosis in
some lobes. Mice given 30 doses of chloroform in the hepatoma study had
moderate liver cirrhosis and necrosis; however, animals given 30 doses that
did not result in necrosis had livers that appeared normal. Necrosis was not
found in hepatoms cells, and hepatomas contained cords of enlarged liver-like
8-19
-------
TABLE 8-6. LIVER AND KIDNEY NECROSIS AND HEPATOMAS IN STRAIN A MICE
FOLLOWING REPEATED ORAL ADMINISTRATION OF CHLOROFORM IN OLIVE OIL
(adapted from Eschenbrenner and Miller, 1945)
Dose (g/kg)
Observation
Liver necrosis
Kidney necrosis
Deaths^
Hepatomas in
surviving
animals receiving
30 dosesa
Sex 2.4
F +
M +
F 0
M +
F 5/5
M 5/5
F
M
1.2 0.6
+ +
+ +
0 0
+ +
1/5 2/5
5/5 5/5
4/4 3/3
--
0.3
0
0
0
+
0/5
2/5
0/5
0/3
0.15
0
0
0
+
0/5
0/5
0/5
0/5
Control
0
0
0
0
0/5
0/5
0/5
0/5
^Numerator is positive occurrences. Denominator is animals observed.
cells which formed disorganized anastamosing columns. The hepatomas did not
appear invasive, and metastasis was not found.
Renal necrosis in males was localized in the areas of the proximal and
distal tubules. Glomeruli and collecting tubules appeared normal. The
severity of renal necrosis was less with lower doses. The different kidney
responses by males and females to chloroform treatment may be due to the
unique lining of the Bowman's capsules with flat and cuboidal epithelium in
females and males, respectively (an anatomic sexual dimorphism in mice).
Although few animals were available for pathologic examination, the
Eschenbrenner and Miller study (1945) indicates that hepatomas in female mice
were induced at chloroform doses that also produced liver necrosis. Early
mortality precluded the development of hepatomas in all animals given
chloroform doses that produced liver necrosis. Hepatomas were not induced by
non-necrotizing doses of chloroform; however, this was not lifetime study and
8-20
-------
a lifetime study perhaps could have given a stronger indication of the
carcinogenic potential of chloroform at these lower doses. The observation
of kidney necrosis in males without tumor formation, and lack of necrosis in
hepatomas, suggests that liver in strain A mice was uniquely sensitive to
tumor induction at necrotizing doses, or that there might have been
additional factors in liver tumor formation besides necrosis. Furthermore,
since a dose of chloroform was given 1 day before sacrifice—a factor which
in itself could have been responsible for producing necrosis, as supported by
liver necrosis found in mice which died after one or two treatments with
chloroform--it is not clear what the extent of necrosis was during the last
month of observation, when the animals were untreated.
8.1.2.4. Rudali (1967)—Rudali (1967) reported a carcinogenicity study of
chloroform in NIC mice. Details such as age and sex of the mice were not
given. The mice received twice-weekly doses of 0.1 ml of a 40 percent
solution of chloroform in oil by force-feeding for an unspecified treatment
period. Twenty-four animals were initially on study, but only five "sound
mice" were evidently given a pathologic examination. An average survival
period of 297 days was reported, but it is not clear if this period applied
to the total group of 24 or to the smaller group of five. An observation
period for the study was not mentioned in the report. The use of a control
group was not indicated, nor was a chemical analysis of the chloroform sample
provided.
Three of the five mice examined in pathology were diagnosed with
hepatomas and hepatic lesions; however, details of the pathologic
observations were not reported. The study by Rudali (1967) gives evidence
for carcinogenic activity by chloroform in NIC mice, even though it is
8-21
-------
weakened by a lack of experimental details, the absence of a control group,
and the small number of animals examined in pathology.
8.1.3. Oral Administration (Drinking Water): Rat and Mouse
8.1.3.1. Jorqenson et al. (1985)--In order to further investigate the
reported positive response of experimental animals to chloroform, to validate
the studies performed by NCI, and to explore the carcinogenicity of
chloroform in drinking water, the route of exposure likely to be encountered
by humans, an investigation of the carcinogenicity of chloroform administered
in the drinking water of the strains and sexes of mice and rats showing a
positive response in the NCI study was undertaken at SRI International
(Jorgenson et al., 1985). In these studies chloroform was administered in
the drinking water of male Osborne-Mendel rats and female B6C3F1 mice at
concentrations of 0 (control), 200, 400, 900, and 1800 mg/L. A second
control group was included in the study with water intake restricted to equal
that of the high dose group. The animals, both rats and mice, were treated
for 104 weeks. The group sizes were larger in the lower level treatment
groups so as to increase the likelihood of detecting a carcinogenic response.
Because of questions arising from the conduct of previous studies, the
chloroform used in this study was redistilled to minimize the
diethylcarbonate, a contaminate noted in the previous study, and the
chloroform concentrations in the animal room and feed were monitored. In
addition, the blood concentrations of chloroform in the rats were also
monitored.
Water consumption was measured during the study. The water intake of
the rats was decreased as the chloroform concentration of the drinking water
increased. The water intake of the mice was minimally affected after the
initial week in which some animals refused to drink. The authors calculated
8-22
-------
the chloroform daily doses in terms of milligrams per kilogram of body
weight, based on the water consumption and concentrations of chloroform in
the water. These are tabulated below.
mg/L in water Rat, mg/kg Mouse, mg/kg
200
400
900
1800
19
38
81
160
34
65
130
263
In Table 8-7, the tumors that were statistically increased among the
rats have been tabulated. The increase in renal tumors among these rats
supports the previous finding from the NCI study. The nontumor renal
pathology, according to the authors, was high in all groups (91-100 percent).
Therefore, it was not possible to relate tumor pathology with nonneoplastic
lesions.
Among the female mice, tumor incidence was not increased. In
particular, the liver tumors (Table 8-8) were not increased, as had been
reported in previous investigations. The high dose in this study, when
expressed in terms of unit per kilogram of body weight, was 263 mg/kg/d and
therefore essentially the same as the lower dose given by gavage in the NCI
study. That similar time-weighted average dose in the NCI study produced an
80 percent response with regard to hepatocellular carcinomas.
It is probable, however, that the peak blood concentrations following a
single daily dose by oral gavage in either water or corn oil as the carrier
vehicle far exceed those following administration in the drinking water,
which results in several intermittent doses during a 24-hr dosing period. As
8-23
-------
TABLE 8-7. RELATIVE TUMOR INCIDENCE IN MALE OSBORNE-MENDEL RATS TREATED WITH CHLOROFORM IN DRINKING
WATER (Jorgenson et al., 1985)
00
I
ro
Tumor Type
All tumors
Neurof ibroma
All lymphomas and
leukemias
All circulatory system
tumors
All kidney tumors
Tubular cell adenoma
Tubular cell adenoma
and adenocarcinoma
Adrenal cortical
adenoma
Adrenal
pheochromocytoma
Thyroid c-cell adenoma
Thyroid c-cell adenomas
and carcinomas
Control
212/303&
(70)b
2/303
(1)
5/303
(2)
5/303
(2)
5/301
(2)
4/301
(1)
4/301
(1)
91/298
(31)
76/298
(26)
44/294
(15)
47/294
(16)
Matched
Control
39/50
(78)
1/50
(2)
1/50
(2)
0/50
(0)
1/50
(2)
0/50
(0)
1/50
(2)
16/50
(32)
8/50
(16)
9/49
(18)
12/49
(24)
Chloroform Concentration (mg/L)
200
227/316
(72)
2/316
(1)
19/316d
(6)
6/316
(2)
6/313
(2)
2/313
(1)
4/313
(1)
86/3116
(25)
71/311
(23)
33/3036
(11)
49/303
(16)
400
105/148
(71)
1/148
(1)
5/148
(3)
3/148
(2)
7/148
(5)
3/148
(2)
4/148
(3)
36/1446
(25)
25/144
(17)
18/148
(12)
27/148
(18)
900
38/48
(79)
0/48
(0)
2/48
(4)
3/48d
(6)
3/48
(6)
2/48
(4)
3/48
(6)
17/48
(35)
12/48
(25)
6/486
(13)
7/48
(15)
1800
34/50
(68)
3/50
(6)
3/50d
(6)
3/506
(6)
7/50d
(14)
5/50d
(10)
7/50d
(14)
11/506
(22)
5/50d
(10)
3/50d
(6)
7/50
(14)
Overall
p-value
0.0778C
0.0167
0.0368
0.0150
0.0001
<0.0001
O.0001
0.0085
0.0012
0.0233
0.0335
^Number of animals bearing indicated tumors/effective number of animals at risk (continuity corrected).
bFigure in parentheses represents percent of animals with the indicated tumor.
coverall p-value calculated using the Peto trend test, for continuity corrected and survival.
dlndividual treatment group statistically different from control group at p < 0.01.
eindividual treatment group statistically different from control group at p < 0.05.
-------
oo
TABLE 8-8. LIVER TUMOR INCIDENCE RATES IN FEMALE B6C3F1 MICE TREATED WITH CHLOROFORM IN
DRINKING WATER (Jorgenson et al., 1985)
Tumor Type
All tumors
Hepatocellular
Hepatocellular
Hepatocellular
carcinoma
adenoma
carcinoma
adenoma and
Control
225/423a
(53)b
19/415
(5)
2/415
(1)
21/415
(5)
Matched
Control
22/47
(47)
0/47
(0)
0/47
(0)
0/47
(0)
Chloroform Concentration (mg/L)
200
217/415
(52)
8/410
(2)
7/410
(2)
15/410
(4)
400
90/142
(63)
8/142
(6)
1/142
(1)
9/142
(6)
900
16/47
(34)
0/47
(0)
0/47
(0)
0/47
(0)
1800
24/44
(55)
0/44
(0)
1/44
(2)
1/44
(2)
^Number of animals bearing indicated tumor/effective number of animals at risk
(continuity corrected).
bFigure in () represents the percent of animals bearing the indicated tumor.
-------
indicated in Chapter 4, Withey et al. (1983) have shown that the
postabsorptive blood peak after 75 mg/kg of chloroform by gavage in water was
39 tig/ml and 6 pg/ml in corn oil. By comparison, Jorgenson et al. (1985)
indicated for 81 mg/kg (900 mg/L in drinking water) the mean blood
concentrations averaged only 75-81 pg/ml. These authors also indicate that
they felt the time of sampling (morning) was representative of a blood
chloroform concentration following a time when rats were actively drinking.
Thus the differences in absorption patterns depending upon the carrier
vehicle, the dosing regimen, and the resultant peak blood and tissue levels
may help to explain differences in outcomes of the carcinogenicity studies.
It has been postulated, but not shown, that corn oil itself may interact in
some way to promote the liver tumors in mice. However, the kidney tumors in
rats occur regardless of the carrier vehicle or dosing regimen.
8.1.4. Oral Administration (Capsules): Dog
8.1.4.1. Heywood et al. (1979)--The carcinogenicity of chloroform in
toothpaste was evaluated in beagle dogs by Heywood et al. (1979). The
toothpaste formulation used was that previously described in Table 8-3 except
for reduced amounts of carragheen gum and glycerol. Chloroform in toothpaste
was transferred from a syringe to gelatin capsules immediately before dosing.
Doses were selected from results of a preliminary range-finding study in
which one or two dogs of each sex per group were given oral chloroform doses
7 days per week for 13 (30 and 45 mg/kg/day), 18 (60 mg/kg/day). or 12 (90
and 120 mg/kg/day) weeks. Because 45 mg/kg/day (lowest toxic dose) produced
pathologic changes in the liver, dose levels of 0, 15, and 30 mg CHC.l3/kg/day
were chosen for the carcinogenicity study.
In the carcinogenicity study, chloroform was given orally in capsules 6
days per week for over 7 years. Eight males and 8 females were assigned to
8-26
-------
each treatment group and to an untreated control group, and 16 dogs of each
sex composed a vehicle-control group. The dogs were initially 18 to 24 weeks
old. All of the dogs were clinically examined before treatment, and had been
receiving medication annually for common diseases. Dogs were fed 200 g of
diet twice daily until week 300, when obese dogs received reduced daily
rations of 300 g. Body weights, food consumption, and water intake were
estimated during the study. Hematology, serum biochemistry, and urinalysis
were included in the evaluation of chloroform toxicity. Treatment was
stopped at 376 weeks, and survivors were sacrificed for macroscopic
examination at 395 to 399 weeks. Major organs were weighed. Tumors,
lesions, and routine tissues and organs were evaluated microscopically.
Liver and kidney specimens from control and high-dose dogs were also examined
by electron microscopy.
Survival, body weights, food and water consumption, and appearance of
the eyes were unaffected by chloroform treatment. Mean body weights
increased from 7 to 8 kg initially to a maximum of 14 to 15 kg; however,
reduction of diet portions for obese dogs complicated the body weight
results. Results of blood and urine analyses were unremarkable except for
dose-related increases in SGPT levels (Table 8-9), which could reflect liver
pathology.
No treatment-related carcinogenic effects were found in necropsy and
microscopic examination of tissues and organs. Nonneoplastic diagnoses
showed that fatty cysts in the livers of all groups were larger and more
numerous in treated dogs.
8-27
-------
TABLE 8-9. SGPTa CHANGES IN BEAGLE DOGS TREATED WITH CHLOROFORM (adapted from Heywood et al. 1979)
00
I
r\>
oo
Group mean SGPT (Mil/ml)
Treatment
(mg CHCl3/kg/day)
30 mg
15 mg
Vehicle-control
Untreated
Pretreatment
6
24 34b
22 29
22 29
24 30
26
58C
33
30
30
52
52C
32
29
27
Treatment stage (weeks)
104
64C
45
40
37
156
76C
46d
30
29
208
9ic
55d
40
30
260
147C
95C
33
32
312
128C
89C
47
50
Post-treatment (weeks)
372 14
102C 105d
66 53
51 56
50 53
19
111
48
128
56
aSerum glutamic-pyruvic transaminase.
bComparison with untreated group; P < 0.05.
comparison with untreated group; P < 0.01.
dComparison with untreated group; P < 0.001.
-------
The study by Heywood et al. (1979) did not show a statistically
significant carcinogenic effect of chloroform in toothpaste given to beagle
dogs, although an increased incidence of total number of neoplasms was
observed in treated dogs. Range-finding tests and SGPT and liver fatty cyst
diagnoses in the carcinogenicity study suggest that a maximally tolerated
dose was approached in the carcinogenicity study. It is not certain if 7
years was long enough for carcinogenicity testing with respect to the
lifespan of the beagle dog (13 to 14 years), but by 7 years spontaneous tumor
formation was becoming evident.
8.1.5. Intraperitoneal Administration: Mouse
8.1.5.1. Roe et al. (1968)--Roe et al. (1968) investigated the
carcinogenicity of chloroform in newborn (C57 x DBA2-F1) mice. Chloroform
was subcutaneously injected into mice of one group as a single 200 pg dose
when the animals were less than 24 hr old, and into mice of another group as
eight daily doses of 200 ^g each, beginning when the animals were 1 day old.
Control groups were given the dosing vehicle, arachis oil, alone. Survivors
were sacrificed for necropsy at 77 to 80 weeks.
No carcinogenic effect of chloroform was found. However, since the
study was reported as an abstract, experimental details were not provided.
Chloroform doses were relatively rather low, and the use of newborn mice
given one or a few doses of chloroform is not equivalent to lifetime
treatment of animals given doses as high as those maximally tolerated.
Additionally, there may be differences in chloroform metabolism between
newborn and adult (C57 x DBA2-F1) mice. Hence, it is concluded that this
study by Roe et al. (1968) does not present sufficient evidence for an
absence of carcinogenic activity by chloroform.
8-29
-------
8.1.5.2. Theiss et al. (1977)--The carcinogenicity of chloroform was
evaluated by Theiss et al. (1977) by means of the pulmonary tumor induction
bioassay in strain A mice.
Test animals were male strain A/St mice initially 6 to 8 weeks old.
Preliminary toxicity tests were performed for selection of maximum tolerated
doses; in these tests, mice received six intraperitoneal injections of
chloroform for 2 weeks and were observed for another 4 weeks. Results of the
preliminary tests were not reported. In the bioassay, chloroform doses in
tricaprylin were 80 or 200 mg/kg, administered 3 times weekly for a total of
24 intraperitoneal injections; or 400 mg/kg, which was injected only twice.
Fifty control mice were given tricaprylin alone. Each treatment group
contained 20 animals. Mice were sacrificed 24 hr after the last dose was
administered, and the lungs were removed for counting and examining surface
adenomas microscopically. The chloroform product given to the mice in this
study was reagent grade (Aldrich Chemical Company), but its chemical
composition was not reported. A positive control group of 20 mice was given
one injection of 1 g/kg of urethane in saline, and compared with 50 controls
given saline alone.
Chloroform treatment did not produce a pulmonary adenoma response in
this study. The average number of lung tumors per mouse was 0 to 0.39 in
each group, except for the positive controls, which had an average of 19.6
lung tumors. At least 90 percent of the mice in each group survived, except
for the mice given 400 mg CHCl3/kg, where there was only 45 percent survival.
However, since this type of bioassay is basically a screen for carcinogen
potential, a negative result does not necessarily indicate a lack of
carcinogenic activity. Evidence for the carcinogenic activity of chloroform
is available in other studies described in this document, and according to
8-30
-------
the authors, there is evidence for carcinogenic activity in other compounds,
e.g., 2-chloroethyl ether and hexachlorocyclohexane in liver, which also
tested negative in the Theiss et al. (1977) study. Carcinogenic effects of
chloroform have been shown in the liver and kidney, whereas the lung was
apparently not a target organ in the Theiss et al. (1977) study or other
chloroform cancer bioassay studies.
8.1.6. Evaluation of Chloroform Careinogem'city by Reuber (1979)
Reuber (1979) evaluated the carcinogenicity of chloroform based on his
review of slides in the NCI (1976) bioassay, and his review of data in other
carcinogenicity studies described in this document. Reuber concurred with
reported findings of rat kidney tumors and mouse hepatocellular carcinomas in
the NCI (1976) study, mouse hepatomas in the Eschenbrenner and Miller (1945)
and Rudali (1967) studies, and mouse kidney tumors in the Roe et al. (1979)
study. However, Reuber concluded that there was an increased incidence of
neoplasms in treated dogs in the Heywood et al. 1979 study. Reuber also
concluded that there were treatment-related neoplasms in the NCI study in
addition to those reported. In rats, Reuber concluded that chloroform
treatment induced liver tumors (hepatocellular carcinomas and neoplastic
nodules) and cholangiofibromas and cholangiocarcinomas in addition to kidney
tumors. Besides hepatocellular carcinomas, Reuber concluded that malignant
lymphoma was also induced by the chloroform treatment in mice. Reuber noted
that treated rats and mice did not exhibit liver cirrhosis, that treated rats
with thyroid tumors generally did not have liver or kidney tumors, and that
liver necrosis was apparent only in high-dose female mice. The differences
in histopathologic interpretation of tissue specimens in the NCI bioassay
between the Reuber study and the NCI report, outside of a difference of
opinion between pathologists, are not clear.
8-31
-------
8.1.7. Oral Administration (Drinking Water): Mouse: Promotion of
Experimental Tumors
8.1.7.1. Capel et al. (1979)—The effect of chloroform ingestion on the
growth of murine tumors was assessed by Capel et al. (1979). Redistilled
analar chloroform was used, but chemical analysis of the product was not
indicated. Test animals were male C57CL/105cSn/01a and male Theiller-
Original (TO) mice, 20 to 22 g body weight. A cage of 20 mice drank 80 to
100 ml of water each day; hence, chloroform was added to yield estimated
doses of 0.15 or 15 mg CHCl3/kg/day for two dose groups, with each mouse
drinking 4 ml water per day. Fresh chloroform solution was given daily and
was protected from light.
In one experiment, 100 TO mice in each dose group were divided into
three approximately equal subgroups. One subgroup (pretreated) was treated
with chloroform for 14 days before and after inoculation of Ehrlich ascites
tumor cells. Another subgroup (post-treated) was given chloroform only after
inoculation of tumor cells. The third subgroup, also inoculated with tumor
cells, served as untreated controls. Tumor cells had been maintained in the
peritoneal cavity of male TO mice by weekly passage of 106 cells. Peritoneal
fluid was collected 7 days after inoculation of cells and diluted with
buffered saline. All mice in the three subgroups were given intraperitoneal
injections of 0.1 ml diluent (106 cells). At the end of exponential growth
at 10 days following inoculation of cells, animals were sacrificed for
removal of peritoneal fluid. The peritoneal cavity was washed with
heparinized buffered saline. Fluid and washings were combined and diluted
with buffered saline. Cells were disrupted by sonication for estimation of
DMA levels per ml cell suspension as a measure of total cell content.
8-32
-------
A second experiment was done in which 100 C57BL mice in each dose group
were subdivided into three groups (pretreated, post-treated, and control),
each of which was treated with chloroform according to the corresponding
protocol in the first experiment. Each mouse received a subcutaneous
injection of 10° B16 melanoma cells suspended in 0.1 ml buffered saline.
Inoculum was obtained from a C57BL mouse that had received a transplant of
syngeneic 816 melanoma cells maintained by intramuscular passage every 14
days. Animals were sacrificed at 21 days after inoculation, and spleen,
mesenteric lymph nodes, and lungs were examined for metastases.
In the third experiment, Lewis lung tumor cells were maintained by
serial intramuscular transplantation in C57BL mice. A group of 100 mice was
divided into three approximately equal subgroups (pretreated, post-treated,
control) to investigate the effect of 15 mg CHCl3/kg/day on tumor growth and
spread according to the protocol used in the first experiment. Each mouse
received intramuscular thigh injections of 2 x 10^ cells suspended in 0.1 ml
buffered saline. Animals were killed 14 days after the administration of
tumor cells, and both the tumor-bearing and the normal thighs were skinned
and severed at the knee and hip. Tumor weight was estimated as the
difference between the weights of the thighs. Pulmonary tumor foci were also
counted.
For estimation of the effect of 0.15 mg CHC^/kg/day in the third
experiment, 100 mice were divided into subgroups of 20 animals each and were
pretreated with chloroform before (for 8, 6, 4, or 2 weeks) and after
injection of the Lewis cells. Mice were sacrificed at 16 days after
inoculation of tumor cells, and tumor weights and numbers of lung foci were
determined. In these animals, homogenates of primary tumors were prepared
for p-glucuronidase estimation and protein content.
8-33
-------
The results of these experiments are summarized in Tables 8-10, 8-11,
and 8-12. Body weights and survival were not affected by chloroform
treatment. Ehrlich ascites tumor cells, as equated with DNA content, were
significantly (P < 0.05) increased in high-dose mice, and slightly, though
not significantly, increased in low-dose animals. Invasion by B16 melanoma
cells, especially in the spleen, was augmented by both doses, and the numbers
of lung foci were also greater in both treatment groups. Metastasis of Lewis
cells was increased only by treatment with 15 mg CHCl^/kg. There was no
change in {3-glucuronidase levels based on tumor protein content in the low-
dose group. The increased tumor protein levels appear to reflect tumor
growth which was not evident by weighing.
The study by Capel et al. (1979) shows an ability of chloroform to
enhance the growth of three types of murine tumors in mice. A dose of 15 mg
CHCl3/kg was effective in each experiment, whereas a dose of 0.15 mg CHCl3/kg
was effective only in the test with B16 melanoma cells. Although this study
does not evaluate the ability of chloroform to induce primary tumors, it does
give evidence for a promoting effect of chloroform on the growth and spread
of experimental tumors at low doses. However, the mechanism by which
chloroform enhanced tumor growth in the study by Capel et al. (1979) is not
certain, and the relevance of this study to the overall evaluation of the
carcinogenic potential of chloroform is not clear.
8-34
-------
TABLE 8-10.
EFFECT OF ORAL CHLOROFORM INGESTION ON THE GROWTH OF EHRLICH ASCITE TUMORS
(Capel et al., 1979)
Dose
0.15 mg/kg/day
15 mg/kg/day
Treatment group
Control
Post-treated
Pretreated
Control
Post-treated
Pretreated
Number of
animals per
group
33
33
33
43
37
30
Average body
weight (g)a
38.3 ± 3.7
39.4 ± 2.9
37.9 ± 3.2
39.4 ± 3.4
37.5 ± 3.0
37.0 ± 3.9
Tumor DNA
(Hg/ml)a
661 ± 222
724 ± 254
770 ± 283
637 ± 221
1143 ± 324
827 ± 245
Significance
NS
NS
P < 0.001
P < 0.001
00
I
CO
en
SResults expressed are the mean ± S.D.
NS = Not significant; P > 0.05.
-------
TABLE 8-11. EFFECT OF ORAL CHLOROFORM INGESTION ON METASTATIC "TUMOR TAKES" WITH B16
MELANOMA (Cape! et al., 1979)
00
I
CO
Animals with B16 melanoma invasion in organics (%)
0.
15
Dose
15 mg/kg/day
mg/kg/day
Treatment
Control
Post-treated
Pretreated
Control
Post-treated
Pre-treated
Number of
animals per
group
26
31
28
30
32
32
Lung
Spleen
15
35
36
15
31
31
Mesenteric
lymph nodes
13
10
29
12
25
32
(a)a
12
10
18
6
19
20
(b)b
3
5
10
4
6
20
^Numbers in column (a) refer to the percentage of animals with tumor foci on the lungs.
^Numbers in column (b) refer to the average number of lung metastases.
-------
TABLE 8-12.
EFFECT OF ORAL CHLOROFORM INGESTION QN THE GROWTH AND SPREAD OF THE LEWIS LUNG TUMORS
(Cape! et al.t 1979)
CD
Number
of
animals Average
per body weight
Dose
0.15 mg/
kg/day
15 mg/
kg/day
Treatment
8d
6
4
2
0
(control)
Control
Post-treated
Pretreated
group
20
20
20
20
20
33
33
33
(g)
30.6
30.6
29.0
29.0
29.3
23.6
24.5
24.4
± 3.8
± 3.8
± 3.6
± 3.5
± 2.7
± 1.3
± 2.3
± 2.2
Tumor
Weight
(g)
3.5 ± 0.81
3.3 ± 0.72
3.3 ± 0.54
3.2 ± 0.72
3.1 ± 0.11
1.6 ± 0.31
1.7 ± 0.51
1.8 ± 0.12
Lung
metastases
165 ± 56
170 ± 41
154 ± 39
147 ± 44
142 ± 34
44 ± 26
57 ± 19
61 ± 19
B-glucuronidase
Significance
NS
NS
NS
NS
P < 0.05
P < 0.01
activityb
0.33
0.27
0.38
0.49
0.58
± 0.56
± 0.79
± 0.073
± 0.070
± 0.094
Protein
contentc
78.2 ±
66.8 ±
60.1 ±
60.3 ±
50.8 ±
4.2
2.7
5.1
4.7
6.2
^Results expressed are the mean ± S.D.
^Expressed as mole product/mg protein/min.
CMilligrams of protein after extraction mg/g wet weight.
dDuration of treatment (weeks).
NS = not significant, P>0.05.
-------
8.2. CELL TRANSFORMATION ASSAY
8.2.1. Styles (1979)
Styles (1979) reported an investigation of chloroform in a cell
transformation system with BHK cells, using growth in semi-solid agar as an
endpoint, as part of a larger study (Purchase et al. 1978) conducted to
screen chemicals for carcinogenic potential. The BHK-agar transformation
assay technique used has been previously described by Styles (1977) and by
Purchase et al.. (1978). In the chloroform study reported by Styles (1979),
baby Syrian hamster kidney (BHK-21/C1 13) cells were exposed to five
different doses of test substance in vitro in serum-free liquid tissue
culture medium in the presence of rat liver microsomal fraction and cofactors
(S-9 mix; Ames et al., 1975). The liver microsomal fraction was obtained
from Sprague-Dawley rats induced with Arochlor 1254.
Cells were grown and maintained in Dulbecco's modification of Eagle's
medium in an atmosphere of 20 percent C02 in air. Cells were maintained at
37°C until confluent, and then were trypsinized and resuspended in fresh
growth medium. Resuspended cells were grown until 90 percent confluent for
transformation assays or 100 percent confluent for stock. Only cells with
normal morphology were used for assays. To minimize spontaneous
transformation frequency, cells were obtained at low passage, grown to 90
percent confluency, and frozen in liquid nitrogen. Cells were thawed at 37°C
in growth medium for further use.
Test compounds were dissolved 1n DMSO or water as appropriate. Each
dose was tested 1n replicate assays. Cells Incubated until 90 percent
confluency were trypsinized and resuspended 1n Medium 199 at a concentration
of 106 cells/ml. Resuspended cells (106) were Incubated with test chemical
and S-9 mix at 37°C for 4 hours. After treatment, cells were centHfuged and
8-38
-------
resuspended in growth medium containing 0.3 percent agar. Survival after
treatment was estimated by incubating 1,000 cells at 37°C for 6 to 8 days
before counting colonies. Transformation was evaluated by counting colonies
after cells were plated and incubated for 21 days at 37°C. The dose-response
for transformation was compared with that for survival. Styles (1977)
accepted a fivefold increase in transformation frequency above control values
at the LCcjQ as a positive result. The spontaneous transformation frequency
of BHK cells (72 experiments) in this study was 50 + 16 per 106 survivors.
The suitability of the soft agar medium for colony growth was checked by
assays with polyoma-transformed BHK-21/C1 13 cells or Hela cells.
Cell transformation results were negative with exposure to chloroform
solution in OMSO added to culture medium in a dose range that included levels
at which toxicity was observed (Figure 8-2). Although chloroform doses high
enough to produce toxicity did not induce transformation, exposure of cells
to chloroform as a vapor could have provided a comparison of the
transformation potential of chloroform as a vapor and chloroform in liquid
solution.
The study by Purchase et al. (1978), which was done on 120 chemicals of
various classes, showed that the BHK-agar transformation assay system was
about 90 percent accurate in discriminating between compounds with
demonstrated carcinogenic or noncarcinogenic activity, and was in
approximately 83 percent agreement with the results of assays done by the
authors with S^ typhimurium (TA 1535, TA 1538, TA 98, TA 100). Styles
(1979) indicated, without presenting numerical data, that the results
obtained in Salmonella assays on chloroform in liquid solution were similar
to the findings of the transformation assays. Purchase et al. (1978) also
observed that metabolically activated agents transformed BHK cells more
8-39
-------
CO
oc
O
>
oc
z>
CO
CO
cc
O
cc
D
CO
oc
UJ
Q.
CO
oc
O
u.
CO
•z.
100
50
0
1100
900
Ł 700
500
300
100
CHCU
1
0.25 2.5 25 250
CONCENTRATION (/wl/ml)
ALSO AMES-VE
2500
Figure 8-2. Negative result in transformation assay of chloroform, which
was also negative in the Ames assay.
Source: Styles (1979).
8-40
-------
strongly in the presence of S-9 mix, thus suggesting that BHK cells have
limited intrinsic metabolic capability.
8.3. EPIOEMIOLOGIC STUDIES
In the last decade there has appeared in the literature a host of
epidemiologic and statistical studies of cancer and exposure to the
constituents of drinking water, of which chloroform is one (Harris, 1974;
Page et al., 1976; Tarone and Gart, 1975; Buncher, 1975; Vasilenko and Magno,
1975; De Rouen and Diem, 1975; McCabe, 1975; Kruse, 1977; Alavanja et al.,
1978; Rafferty, 1979; Kuzma et al., 1977; Harris et al., 1977; Salg, 1977;
Mah et al., 1977; Brenniman et al., 1978; Tuthill et al., 1979; Wilkins,
1978). These studies have been subjected to several critical reviews
(Wilkins et al., 1979; U.S. Environmental Protection Agency, 1979; National
Academy of Sciences, 1978) and have been discussed in some detail. Some very
general relationships have been noted by the reviewers. Of particular
importance is the appearance of some consistency in the finding of cancer of
the large intestine, rectum, and bladder associated with the constituents of
drinking water.
It must be emphasized that none of the studies discussed in this section
implicates chloroform directly as the sole or dominant constituent of
drinking water responsible for the excess of cancer at these sites. Over 300
volatile organic contaminants have been identified in drinking water, and
many of these have been identified as carcinogens (Wilkins et al., 1979).
However, chloroform at a peak concentration of 266 pg/L has been shown
to exceed peak concentrations of other detected carcinogens by levels 37
times higher than those of the next highest carcinogen, vinyl chloride
(Wilkins et al., 1979).
8-41
-------
Chloroform measurements appear to range largely between 1 and 112
according to a survey of 76 drinking water supplies (Cantor et al., 1978).
Although a direct association cannot be made, the possibility still
exists that since chloroform is apparently the predominant component in
chlorinated drinking water, it could be a contributing factor in the etiology
of the cancer associated with the consumption of drinking water.
Almost all of the above-referenced studies were ecological correlation
investigations, and only a few utilized case-control methods. The studies
varied by sample size, cancer sites considered, control variables, and the
types of end points used as indicators. Among the problems posed by the data
in these studies are the following: 1) a lack of data measuring the quantity
of chlorine and chloroform in drinking water; 2) the limited nature of
recently acquired data on the quality and quantity of organics in drinking
water; 3) the limited amount of information given regarding personal
consumption of drinking water; 4) the long latency periods associated with
most cancers (current cancer rates reflect exposures received decades
earlier); and 5) the demographic effects of migration, which adds another
dimension of difficulty to the quantification of personal consumption of
drinking water over time.
Since publication of the three reviews referred to above, several
additional studies of cancer and exposure to trihalomethanes have been
published. The following, pages discuss each of these studies in detail.
8.3.1. Young et al. (1981)
Young et al. (1981) conducted a case-control study in which cancer
deaths in 8,029 white females were matched with non-cancer deaths in some
8,029 white females for county of residence, year of death, and age recorded
on death certificates in 28 counties in the State of Wisconsin from 1972
8-42
-------
through 1977. Information about the chlorine content of the drinking water
of the 16,058 cases and controls was derived from mail-back questionnaires
recently submitted to the superintendents of 202 waterworks encompassing the
counties sampled. The questions pertained to prechlorination and
postchlorination dosages used over the past 20 years (average daily dose in
ppm). For 14% of the sample who were not served by a waterworks, decedents
were assigned chlorine dosages of zero. The assignment was on the basis of
water supplied to decedent's usual place of residence.
Odds ratios were calculated from a logistic regression model. This
model provided estimates of the relative risk of site-specific cancer deaths
for exposure of the previous 20 years to high, medium, and low chlorine
doses, as compared with no chlorination. Urbanicity, marital status, and
site-specific high-risk occupation were controlled in the model. Only colon
cancer showed a significant (P < 0.05) association with chlorine intake in
all three dosage categories. However, no gradient of increasing risk with
increasing dosage was apparent. For the high, medium, and low dosage
categories, the odds ratios were 1.51, 1.53, and 1.53, respectively. All
were significant at P < 0.05. In those counties where the drinking water
supplies were exposed to rural runoff, the odds ratios for colon cancer
increased to 3.43, 3.68, and 2.94 for high, medium, and low average daily
chlorine doses when controlled for water source depth and purification.
These were statistically significant at the P = 0.025 level. Colon cancer
mortality was not related to chlorination in counties not exposed to rural
runoff. This finding is consistent with the hypothesis that trihalomethanes
are formed through the action of chlorine on organic substances in drinking
water.
8-43
-------
Nonsignificant risks were evident at the remaining sites, i.e.,
esophagus, stomach, rectum, liver, pancreas, kidney, bladder, lung, brain,
and breast. The average daily chlorine dose categories were designated by
the authors as follows: none (less than 0.01 ppm), low (0.01-0.99 ppm),
medium (1.00-1.70 ppm), and high (1.71-7.00 ppm).
The authors made a number of assumptions regarding exposure of subjects
and controls to chloroform. They assumed that chlorine in drinking water
would represent a good surrogate for exposure of cases and controls to
chloroform, reasoning that trihalomethanes such as chloroform are believed to
result from the reaction of chlorine with naturally occurring organics in
water. Although drinking water at the tap was not analyzed for chloroform or
other trihalomethanes, the authors assumed that the measured levels of
chlorine at the respective waterworks would correlate well with presumed
exposure to chloroform in drinking water.
Such implicit assumptions appear questionable for several reasons.
First, the latent period for the development of several, if not most, of the
cancer sites is most probably greater than 20 years. This is longer than the
period covered by the exposure data on chlorination of water supplies used by
the authors.
Second, migration within and around the 28-county area could have masked
any real risk that was related to exposure. A diagnosis of colon cancer,
which has a 5-year survival rate of better than 46 percent, could have
induced victims to migrate to urban areas (where chlorine levels were higher)
in order to obtain better medical care, thus leading to a false positive
association.
Third, the amount of chloroform that is formed from the addition of
chlorine is a function of several important variables: the quantity of
8-44
-------
organics in the water supply, treatment practices, and chlorine dosages. The
quantity of organics in the water supply is, in turn, determined by the
nature of the water supply source. Surface water (rivers and streams)
receives large quantities of organics from land runoff, whereas groundwater
contains little or no organic material; hence, the likelihood of chloroform
formation from the addition of chlorine to a groundwater supply is minimal.
Fourth, liquid intake rates and amounts vary considerably from person to
person. It is clear that most people satisfy their liquid requirements
through a variety of drinks besides tap water, (e.g., milk, orange juice,
coffee, soda). It is conceivable that many may drink little water because of
these competing sources of liquid refreshment. Therefore, it is probable
that many persons who were ranked as having been exposed to chloroform may in
fact have had little exposure to it. The resulting misclassification of
cases and controls by exposure category would tend to mask any gradient of
increasing risk with exposure if one existed.
Another possibly confounding variable not controlled for in this study
is the dietary intake of meat and foods low in fiber content (Reddy et al.
1980), both of which have been hypothesized as being related to colon or
rectal cancer. The dietary intake of such foods, however, is not known to be
correlated with the quantity of chlorine in drinking water, although the
possibility of a spurious correlation cannot be ruled out. In more urbanized
counties where chlorine levels are higher, residents may consume a diet of
more meat and less fiber.
In summary, a definite association of chlorine or chloroform in drinking
water with an increased risk of colon cancer should not be made for the
reasons stated.
8-45
-------
8.3.2. Hogan et al. (1979)
Hogan et al. (1979) conducted an ecological study of site-specific
cancer rates based on NCI cancer mortality data by county for the years
between 1950 and 1969 (Mason and McKay, 1974) and on chloroform levels in
finished drinking water, as determined by the U.S. EPA in two separate
surveys (U.S. EPA, 1975). The first survey, known as the National Organics
Reconnaissance Survey (NORS), consisted of samples from 80 water treatment
facilities across the country. The second survey covered 83 utilities in the
states of Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin.
Linear multiple regression analyses were done for each set of data
separately. The dependent variable was county site-specific cancer
mortality. Weighted and unweighted regression coefficients were determined
for a number of independent variables selected by the author based on a study
by Hoover et al. (1976). A variety of demographic characteristics related to
cancer mortality were used in addition to the variable "chloroform levels" as
determined from the NORS and regional surveys to explain cancer mortality.
These characteristics were as follows: county population density, percent of
urbanization per county, percent of nonwhite people, percent of foreign-born,
county median family income, educational level, percent of workforce employed
in manufacturing, chloroform level in drinking water samples, and county
population. According to the authors, the weighting was based on the inverse
of the square root of the population of the race-sex county stratum, and was
done chiefly to improve the precision of the regression estimates.
Significant positive statistical correlations were found between
chloroform levels in treated drinking water and cancer mortality specific for
bladder, rectum, and large intestine in the "weighted" regression for white
females. On the other hand, only stomach cancer appeared to be positively
8-46
-------
correlated significantly with chloroform levels in white males. Without
weighting, cancers of the bladder, rectum, thyroid, and breasts were
significantly correlated with chloroform levels in white females. In white
males, cancers of the pancreas and rectum were significantly correlated with
chloroform without weighting. Only estimated regression coefficients were
provided with their corresponding P values. The study contained no
information on actual levels of chloroform observed in drinking water.
Nonwhites were not considered because of the small sizes of the populations
from which rates were derived.
Ecological studies such as this one are necessarily weak because their
information is based on aggregate rather than individual data. The evidence
for an association is indirect and definite conclusions cannot be drawn,
although hypotheses may be formulated. It is not certain whether a multiple
linear regression technique is the proper method for analyzing such data,
since the assumption of linearity implied in its selection may not be
warranted. Also, since the model contains no interaction terms, it is
implicit that the chosen control variables are independent of each other, and
such an assumption may also be unwarranted. Furthermore, as was mentioned in
the Young et al. (1981) study, these data are weakened because it was assumed
that the subjects were actually exposed to the levels of chlorine (or
chloroform) indicated. Another limitation is that since the chloroform data
were collected in 1975, the more relevant exposure data (assuming a general
cancer latency of 10 to 30 years) should be those of 1920 to 1959, given that
the site-specific cancer mortality data covered the period 1950-1969.
8.3.3. Cantor et al. (1978)
Cantor et al. (1978), in an ecological study of cancer mortality and
halomethanes in drinking water, used age-standardized cancer mortality rates
8-47
-------
by site and sex in whites for the years 1968-1971, but only in the 923 U.S.
counties that were more than 50% urban in 1970. This study was similar to
the Hogan et al. (1979) study with respect to its design; i.e., a weighted
linear regression model was used with sex- and site-specific cancer rates as
the dependent variable. The weight was directly proportional to the square
root of the counties' person-years at risk and thus inversely proportional to
the standard deviation of the estimated mortality rate. Chloroform (CHC13),
bromochloromethane (BTHM), and total trihalomethane (THM) levels were
obtained from the two EPA surveys (U.S. EPA 1975) used in the Hogan et al.
study. Demographic variables used in the regression model on a county-wide
basis were as follows: percent of urbanization (1970); median school years
completed by persons over age 25; population size (ratio of 1970 to 1950
population); percentage of the work force in manufacturing; and percentage of
foreign-born. Although a predicted, age-adjusted, site-specific cancer rate
was calculated for each county based on this regression technique, only the
data for 76 counties, where more than half of the population of the counties
was served by a sampled water supply, were actually used in this correlation
analysis of THM levels with residual mortality rates. Figure 8-3 gives a
frequency distribution of the chloroform levels in these 76 U.S. drinking
water supplies. The three indicators, chloroform, bromochloromethane, and
total trihalomethane, were highly correlated with one another.
Positive nonsignificant correlations with THM levels were evident with
respect to several forms of cancer, including lymphoma and kidney cancer in
males (Table 8-13). But according to the authors, bladder cancer mortality
rates gave the strongest and most consistent association with THM exposure
after controlling for differences in social class, ethnic group, urbanicity,
region, and extent of county industrialization (Table 8-14). However, the
8-48
-------
^.o
20
Ll 1
LU
O
z
IT
OC
§ 15h
0
u_
O
z 10
UJ
O
UJ
OC
LL.
n
—
l
i
1
i
0.001 0.01 0.1 1.0
MICROMOLES CHCI3/LITER
10.0
Figure 8-3. Frequency distribution of CHCU levels in 76 U.S. drinking
water supplies. The abscissa is linear in the logarithm of the level.
Source: Cantor et al. (1978).
8-49
-------
TABLE 8-13. CORRELATION COEFFICIENTS BETWEEN RESIDUAL MORTALITY RATES IN
WHITE MALES AND THM LEVELS IN DRINKING WATER BY REGION AND BY PERCENT OF
THE COUNTY POPULATION SERVED IN THE UNITED STATES (Cantor et al. 1978)
Site of
cancer
Kidney
Lymphoma
(non-
Hodgkins)
Correlation coefficients for regions
THM
Indicator
North
CHCls 0.11
(0.54)a
BTHM 0.06
(0.74)
South
-0.11
(0.73)
0.08
(0.79)
Mountain Pacific
0.66
(0.11)
0.05
(0.92)
of the U.S.
All regions
0.14
(0.33)
0.06
(0.70)
Correlation coefficients for counties in which
the percent of the population served was:
50-64% 65-84% 85-100% 50-100%
Kidney CHC13
Lymphoma BTHM
(non-
Hodgkins)
-0.16
(0.44)
-0.33
(0.11)
-0.11
(0.60)
-0.19
(0.36)
0.42
(0.04)
0.36
(0.08)
0.07
(0.55)
-0.08
(0.81)
aP value for two-tailed t-test is shown in parentheses.
TABLE 8-14. CORRELATION COEFFICIENTS BETWEEN BLADDER CANCER
MORTALITY RATES BY SEX AND BTHM LEVELS IN DRINKING
WATER BY REGION OF THE UNITED STATES
(Cantor et al. 1978)
Correlation coefficients by region
Bladder cancer
Number
Male white
Female white
North
31
0.523
(0.002)
0.30
(0.11)
South
13
0.04
(0.90)
0.20
(0.51)
Mountain
7
-0.02
(0.96)
0.63
(0.13)
Total
51
0.30
(0.03)
0.33
(0.02)
aP value for two-tailed t-test is shown in parentheses. Counties
with at least 65% of their populations served by one water supply
were included in this analysis.
8-50
-------
association appeared to be greatest with respect to BTHM and not chloroform.
The corresponding correlations for chloroform were positive but
nonsignificant. The authors report that although other sites appeared to be
positively correlated with THM levels, the inconsistencies "outweigh the
consistencies," thus casting doubt on the reliability of these correlation
coefficients; i.e., the direction and strength of the correlations bear
little relationship to the percent of population served by treated drinking
water and/or by region.
The authors noted an association of kidney cancer with chloroform
exposure that was restricted to males, but was significant only in counties
where at least 85 percent of the public was served by treated drinking water.
In counties where less than 85 percent was served by treated drinking water,
the correlation coefficients were actually negative. Combining all counties
with greater than 50 percent served by treated drinking water, the
correlation coefficient was nonsignificant and close to zero. One
interesting observation was that without controlling for ethnicity, the
authors found a "fairly strong" association of THM levels with colon cancer
and lung cancer rates in both sexes, and even a dose-response relationship
between these tumor sites and the proportion of the population exposed.
However, when ethnicity was added to the regression model, these
relationships disappeared.
Again, this is a descriptive study from which hypotheses can be
formulated only for future in-depth study. It cannot be concluded that even
the significant positive correlations in the study indicate any evidence of
real associations. As the authors point out, potential sources of error
(i.e., control of confounders such as cigarette smoking and diet) are
particularly difficult since no direct information is available on the
8-51
-------
individuals studied. The main problem with such studies, as mentioned
earlier, is that the data are aggregate rather than individual. Such data
frequently include large numbers of individuals who never received the
exposure in question. Associations derived from such data may be misleading
and are often unreliable.
8.3.4. Gottlieb et a!. (1981)
Gottlieb et al. (1981) completed a case-control study of the
relationship between Mississippi River drinking water and the risk of rectum
and colon cancer. The study was based on mortality data gathered from 20
parishes in southern Louisiana. Rectal and colon cancer deaths (692 and
1167, respectively) from 1969 to 1975 were matched one-to-one to noncancer
deaths by age at death, year of death, sex, and race, with respect to
industrial and urban-rural characteristics, which were defined so that each
parish included nearly equal populations using groundwater and surface water
supply sources, based on information from the 1970 census.
Three different estimators of exposure were used. The first,
"sourcelife," is defined as follows: "mostly surface" (birth and death in a
surface-water-using parish); "some surface" (some known surface water use at
birth or death); "possible surface" (death in a groundwater parish but had
either unknown or out-of-state birthplace); and "least surface" (birth and
death in a groundwater-using parish). Length of residence was also
considered, if known and for more than 10 years. The second index used was
chlorine level (none, low [less than 1.09 ppm], or high [greater than
1.08 ppm]). The third index was the level of organics in the drinking water
(low [less than 68 ppm] and high [greater than or equal to 68 ppm]).
Sourcelife could be determined for 99.2 percent of the entire group of 3718
cases and controls, but 51 percent had no data for length of residence or had
8-52
-------
lengths of residence of under 10 years. For those with lengths of residence
of less than 10 years, water sources during the possible carcinogenic period
were unknown. Chlorine values were available for 78.9 percent of the 3718
sources, while organic levels were available for only 50.1 percent of the
sources. The analyses using the latter two variables were equivocal,
possibly due to the lack of information on these parameters.
Colon cancer was found not to be related significantly to any water
variable, although the number of colon cancer cases in this study (1167) was
greater than the number of rectum cancer cases (692). The authors hint that
the earlier correlation found in ecological studies could have resulted from
confounding with urban lifestyles. Rectal cancer, on the other hand, was
found to be significantly elevated with respect to surface or Mississippi
River water consumption. Based on sourcelife, the odds ratio for rectal
cancer for those who were born and died using groundwater sources was 2.07
(95% confidence interval [C.I.] 1.49-2.88) based on a multidimensional
contingency table analysis. Chlorination was significantly associated with
rectal cancer, and for those who used river water, the risk decreased as the
distance from the mouth of the river increased. The odds ratio for cancer of
the rectum at a location below New Orleans versus one above the city was 1.82
(95 percent C.I. 1.01-3.26). The authors noted that both sexes were at
increased risk. With respect to controlling for the effect of chlorination
where adequate numbers existed, the surface water versus groundwater effect
on rectal cancer was of only borderline significance (P = 0.05), implying a
chlorine effect.
On the other hand, information on organics levels was available for over
48 percent of the rectal cancer group and their controls. The odds ratio
calculated based on these data was nonsignificant (Table 8-15), but was
8-53
-------
probably subject to some bias with respect to availability of exposure data
as a function of date of death.
TABLE 8-15. RISK OF MORTALITY FROM CANCER OF THE RECTUM ASSOCIATED WITH
LEVELS OF ORGANICS IN DRINKING WATER (Gottlieb et al., 1981)
High (> 68 ppm)
Low (< 68 ppm)
Total
Odds ratio
Cases
110
232
342
1.08
Controls
97
220
317
With respect to colon cancer, the authors felt that since they had
grouped the parishes according to industry and urban characteristics
(matching was done within the parish group), they successfully eliminated
urban lifestyle as a confounder in their evaluation of colon cancer and
drinking water.
The results of this study suggest that cancer of the rectum is linked to
the consumption of surface water, and since chlorination appears to be an
effect modifier altering the risk ratio to only borderline significance, it
would seem that chlorination does contribute to the risk of rectal cancer.
8.3.5. Alavan.la et al. (1978^
Alavanja et al. (1978) reported on a case-control study of 3446
gastrointestinal and urinary tract cancer deaths (1595 females and 1851
males) occurring during a three-year period from 1/1/68 to 12/31/70 in seven
counties of New York State. Some 3444 individually matched noncancer deaths
8-54
-------
were also selected. Independent variables were 1) residence in an urban or
rural area, 2) residence in an area served by chlorinated or nonchlorinated
water, 3) residence in an area served by surface water or groundwater, and
4) occupation. Cases were taken from computer tapes of New York State death
certificates and were individually matched with an equal number of non-cancer
deaths for the same year. Matching variables were age, race, sex, foreign-
versus United States-born, and county of usual residence. If potentially
confounding variables could not be controlled via the matching process, the
cases and controls were stratified by these confounding variables. The data
were analyzed by the chi-square test. A statistically significant excess of
gastrointestinal and urinary tract cancer mortality occurred among women in
the urban county of Erie (odds ratio [OR] - 2.08), with nonsignificant
excesses in Schenectady County (OR = 2.98) and Alleghany County (OR = 4.13).
Likewise, among men a statistically significant excess of gastrointestinal
and urinary tract cancer mortality occurred in Erie County (OR = 2.15) and
Rensselaer County (OR = 1.98), and a nonsignificant excess occurred in
Schenectady County (OR = 1.96) and Allegany County (OR = 2.85). Although the
study encompassed a seven-county area, almost two-thirds of the deaths
occurred in Erie County. The combined overall odds of dying from
gastrointestinal and urinary tract cancer for all seven counties combined
(including Erie), were only 1.79 based on 3446 cases, whereas in Erie County
alone they were 3.15 based on 2177 cases. The authors concluded that males
and females residing in the chlorinated water areas of the counties noted
above were at a greater risk of gastrointestinal and urinary tract cancer
mortality not due to age, race, ethnic distribution, urbanicity, occupation,
inorganic carcinogens (Cd, As, Be, Pb, Ni, N03), or surface/groundwater
difference. No environmental data are provided, however, to characterize
8-55
-------
quantities of chlorine (or chloroform) exposure. "Inadequate water quality
data" prevented the authors from making a "definitive claim that the process
of chlorination is directly or indirectly responsible for the greater risk of
gastrointestinal and urinary tract cancer mortality" in chlorinated cancer
areas. No description is given of how residence was classified into
chlorinated versus nonchlorinated water areas or surface water versus
groundwater areas through the use of water distribution maps, a practice
which can result in misclassification on the basis of exposure. Again,
because of the lack of individual dosage data on chloroform exposure and the
low significance of the risks described, this study can only be regarded as
suggestive for gastrointestinal and urinary tract cancer mortality.
8.3.6. Brenniman et al. (1978)
Brenniman et al. (1978) attempted to confirm the findings of Alavanja
et al. (1978) in a case-control study of gastrointestinal and urinary tract
cancer mortality among whites in 70 Illinois communities using both
chlorinated and nonchlorinated groundwater. The authors limited the study to
groundwater because of the possible introduction of confounding effects due
to agricultural runoff and industrial sewage in surface water. The 3208
cases and 43,666 controls used were those of Illinois deaths occurring
between 1973 and 1976. Controls were selected from a pool of noncancer
deaths after the elimination of certain special types of deaths, such as
perinatal deaths.
Chlorinated groundwater communities were matched with nonchlorinated
groundwater communities that were similar with respect to urbanicity and
Standard Metropolitan Statistical Area (SMSA) description. To ensure a
minimum follow-up period, water supplies were categorized as chlorinated or
nonchlorinated according to a "1963 inventory of municipal water facilities."
8-56
-------
Additionally, questionnaires were sent to water treatment plants in the
communities to verify the 1963 data. The beginning dates for chlorination
were obtained for many of the plants. Based on an EPA survey, it was found
that 14 chlorinated groundwater supply sources in Illinois had chloroform
concentrations ranging from less than 1 ^g/L to 50 pg/L, with a mean
concentration of 10.8 pg/L.
In females, statistically significant increased relative risks of cancer
of the large intestine and rectum (OR = 1.19, P < 0.05), as well as total
digestive tract cancer (excluding liver) (OR = 1.15, P < 0.05), were found
for chlorinated versus nonchlorinated Illinois groundwater supplies. With
respect to total gastrointestinal and urinary tract cancer, the risk was
significantly increased in females living within standard metropolitan
statistical areas (OR = 1.28, P < 0.025) and within urban areas (OR = 1.24,
P < 0.025) between chlorinated and nonchlorinated groundwater communities.
Where evidence was available concerning a history of chlorination, the
authors noted that the relative risk of total gastrointestinal and urinary
tract cancer tended to increase with time from initial chlorination, although
the change was small. The greatest increase occurred in urban nonstandard
metropolitan areas (OR = 1.14 if chlorinated since 1963 and nonsignificant,
but OR = 1.28 if chlorinated since 1953 and significant, P < 0.025).
Although several significant findings were observed in this study, the
authors dismissed the results of their own study on the basis that
confounding factors such as diet, smoking, and occupation were not
controlled. These authors felt that the findings were tenuous and did not
confirm the findings of Alavanja et al. (1978) either in strength or in
consistency. They state that "chlorination of groundwater does not seem to
8-57
-------
be a major factor in the etiology of site-specific gastrointestinal and
urinary tract cancers."
8.3.7. Struba (1979)
Struba (1979), as part of his Ph.D. thesis, completed a case-control
study of mortality in North Carolina on individuals who died at age 45 or
under during the period 1975-1978. The cancer sites studied were the rectum,
colon, and urinary bladder. Between 700 to 1500 cases per site were matched
with controls by age, race, sex, and geoeconomic region (coastal, piedmont,
or mountain). Noncancer deaths were excluded if cancer was listed as a
contributory or underlying cause of death. For colon and rectal cancer,
certain precancerous colonic disorders were excluded (ulcerative colitis,
familial polyposis, and adenomatous polyposis). Water data were classified
by source, treatment, and previous use. "Source" was defined as ground,
surface uncontaminated, or uncontaminated by upstream pollution. "Treatment"
was defined as none, prechlorinated, post-chlorinated, or both. "Previous
use" included the following 15 categories of upstream pollution for
contaminated water only:
(1) Tobacco manufacturing
(2) Textile manufacturing
(3) Textile bleaching and dyeing
(4) Furniture manufacturing
(5) Pulp and paper mills
(6) Chemical industries
(7) Petroleum refining
(8) Rubber and plastics manufacturing
(9) Leather tanning and finishing
(10) Abrassures, asbestos, minerals
(11) Primary metals industries
(12) Electroplating
(13) Electric power generation
(14) Urban areas > 50,000
(15) Out-of-state upstream discharge
8-58
-------
The author found small but significant odds ratios (1.3 to 2.0) for all
three sites (colon, rectal, and bladder) in rural areas, as well as
significant odds ratios for each of the water quality variables in many
stratified or combined analyses. Odds ratios for urban areas (population
over 10,000) were generally not significant. Urbanization was shown to be an
effect modifier for colon cancer and a likely confounder for rectal and
bladder cancers. The author considered socioeconomic status to be a likely
confounder for cancer of the rectum and bladder. Multivariate analyses
showed no evidence that occupation acted as a confounder for bladder cancer
in this study. To estimate migration effects, cases and controls were
stratified by place of birth and death (birth and death in the same county;
birth and death in North Carolina; and death in North Carolina, birth
unspecified) and substratified by region, age, race, sex, and urbanization.
Odds ratios for treatment (chlorinated and nonchlorinated) were computed for
all of the strata. For all three cancer sites, the group with least
migratory influence had the highest odds ratio, thus lending support to the
author's supposition that an increasing migratory effect is associated with a
decreasing risk of cancer of all three sites.
Additionally, Struba found an increasing gradient of risk from the
coastal regions of North Carolina to the mountains, a finding that he
maintains is consistent with a stronger contrast between surface water and
water from deep wells than between surface water and water from shallow
wells, which are known to be susceptible to contamination by surface water
seepage into groundwater aquifers. However, the author notes that this
difference could be due to differences in water treatment practices or
confounding by uncontrolled factors such as dietary habits or lifestyles.
8-59
-------
8.3.8. Discussion
These later ecological and case-control studies of chlorine exposure and
cancer risk from water supplies tend to support the finding of increased
risks of bladder, colon, and rectal cancer from exposure to chlorinated
water. It seems that this association is at best weak, although significant,
as evidenced by odds risk ratios that range as high as 3.6 in the Young
et al. (1981) study, but generally fall between 1.1 and 2.0 (see Table 8-16)
in the remaining case-control studies. The risk ratios derived in these
studies could be explained by the confounding effects of uncontrolled
influences such as smoking, diet, air pollution, occupation, and lifestyle,
but they appear to have some consistency across several independent and
diverse study groups. Of course, all of the case-control studies use
residence data and cause-of-death information from death certificates, and
thus are not strictly incidence studies. Bias can creep in from several
sources: differential survivorship rates due to proximity to better medical
care and treatment facilities, higher socioeconomic status, and the
possibility of migration of newly diagnosed cancer patients to major medical
care centers where chlorination is used to a greater extent. Underestimates
of risk can result from failure to control for migration before diagnosis,
misclassification of cause of death, and use of chlorination as a surrogate
variable in place of more direct measurements of chloroform, especially if
the chlorinated source contains few organic contaminants. Hence, the
association is weak but significant with regard to the three cancer types and
exposure to chlorinated drinking water. Since exposure to chlorine in water
is not the same as exposure to chloroform, the most that can be said is that
there is a suggestion of an increased risk of cancer of these three sites
from exposure to chloroform. If this risk truly exists, it may be due to an.
8-60
-------
TABLE 8-16.
CANCER RISK ODDS RATIOS AND 95% CONFIDENCE INTERVALS
(CHLORINATED VERSUS UNCHLORINATED)
Site
Rectum
Alavanja
et al . ,
19783
1.93
(1.31, 2.83)
Brenniman
et al.,
1978b
1.26 (crude)
(0.98, 1.61)
1.22 (adjusted)
Young
et al . ,
1981C
1.39 high
(0.67, 2.86)
1.16 medium
(0.58, 2.32)
1.13 low
0.61, 2.08)
Gottlieb
et al .
1981d,e
1.41
(1.07, 1.87)
Struba
1979d,e
1.53
(1.24, 1.89)
Colon 1.61
(1.28, 2.03)
1.08 (crude)
(0.96, 1.22)
1.1 (adjusted)
Bladder 1.69
(1.11, 2.56)
1.04 (crude)
(0.81, 1.33)
0.98 (adjusted)
1.51 high
(1.06, 2.14)
1.53 medium
(1.08, 2.00)
1.53 low
(1.11, 2.11)
1.04 high
(0.43, 2.50)
1.03 medium
(0.42, 2.54)
1.06 low
(0.60, 3.09)
1.05
(0.95, 1.18)
1.30
(1.13, 1.50)
1.07
(0.84 x 1.36)
1.54
(1.26, 12.88)
aCalculated for both sexes and all races combined. Confidence intervals were not
stated in Alavanja et al. (1978). Crump (1979) calculated them by applying the
method of Fleiss (1979) to data in Alavanja et al. (1978).
bCalculated for Caucasians of both sexes. Adjusted values were adjusted for age,
sex, urban/rural, and SMSA/nonSMSA. Confidence intervals were not stated in the
original report. Crump (1979) calculated them by applying the method of Fleiss
(1979) to data on total cases and total controls supplied by Dr. Brenniman in
personal communication.
cCalculated for white females and for high, medium, and low average daily
chlorine doses compared with no chlorination. Odds ratios and confidence
intervals computed by logistic regression, controlling for urbanization, marital
status, and site-specific occupation.
dCalculated for both sexes and all races combined.
estruba (1979) and Gottlieb et al. (1981) also computed odds ratios for surface
water versus groundwater as follows. Struba: rectum 1.55 (1.26, 1.91); colon
1.27 (1.10, 1.46); bladder 1.48 (1.22, 1.80). Gottlieb et al.: rectum 1.51
(1.21, 1.90); colon 0.95 (0.88, 1.03).
SOURCE: Crump and Guess, 1982.
8-61
-------
intermediate in the natural synthesis of chloroform (communication with Dr.
Kenneth P. Cantor, NCI).
In summary, it appears that there may be a suggestion of an increased
risk of certain forms of cancer (bladder, large intestine, and especially
rectum) due to exposure to chlorinated drinking water contaminated with
organic material. The significant associations of cancer at these three
sites from chlorinated drinking water (high in organic constituents) does not
constitute conclusive evidence of a definite risk of colon/rectal cancer with
chloroform, however it is suggestive. The evidence of a significant
association of kidney cancer with chloroform exposure in drinking water is
even more questionable, since it was based on the findings of only one study,
which was confined to males residing in counties where more than 85 percent
of the population was served by treated drinking water. A statistically
positive correlation was seen only in males residing in counties with over 85
percent treated drinking water. No association was observed in females in
these same counties, and the correlations were actually negative for both
males and females in counties with less than 85 percent treated drinking
water.
It appears that these case-control and ecological studies in humans
suggest a weak but significant association of certain forms of cancer (colon,
bladder, and especially rectal) with chlorinated drinking water contaminated
with organic material. Chloroform appears to be the single largest
constituent, but further epidemiologic research should be accomplished to
confirm these findings.
In view of what appears to be common problems with the many case-control
and ecological studies of chloroform, it seems appropriate to employ a study
8-62
-------
design such that these common problems could be eliminated or at least
reduced in scope. Such a suggested design is the retrospective prospective
cohort design. Ideally, the design calls for the identification of a large
group (say 1000) of individuals who were exposed to only chloroform sometime
in the past. The constructed cohort would then be followed through time from
date of initial employment in the job in which the exposure occurred to the
present and adverse health consequences occurring to the cohort during this
follow-up period would be compared with that of the general population.
Since hospital, laboratory, and university personnel used chloroform in
the past as an anesthetic, the feasibility of constructing a cohort from
personnel records of exposed individuals (i.e., lab technicians,
anesthesiologists, and nurses) maintained by these institutions should be
examined if they can be shown to have been exposed to chloroform. One
problem, however, that might remain is the possibility of the occurrence of
concomitant exposures to other substances that may be carcinogenic, though
presumably the profile of potential exposures would be different from that
found in all the drinking water studies. This possibility can be evaluated
at the time. The presumed route of exposure would be absorption and/or
inhalation in humans rather than ingestion, as it was in all the earlier
drinking water studies.
8.4. RISK ESTIMATES FROM ANIMAL DATA
Evidence for the carcinogenicity of chloroform consists of several
positive long-term studies in mice and rats (NCI, 1976; Roe et al., 1979;
Jorgenson et al., 1985). These positive animal studies report a dose-
dependent excess incidence of liver carcinomas in male and female mice, and
of malignant renal tumors in mice and rats. There is also indication (from
8-63
-------
binding studies and from mutagenicity tests that utilize endogenous or in
vivo metabolism) that chloroform may have the potential to be a weak mutagen.
Metabolism produces phosgene and other putative reactive metabolites that
covalently bind extensively to cellular lipids and proteins. Binding i_n vivo
to DNA, although at low levels, has been demonstrated. Organ localization
and binding intensity parallel acute cellular toxicity in liver and kidney of
experimental animals. There are no known qualitative differences among
species with regard to metabolic pathways or metabolite profiles.
It is important to note that the quantitative estimations of the impact
of chloroform as a carcinogen are made independently of the overall weight of
evidence that chloroform is carcinogenic in animals. The calculations are
made as if chloroform were a human carcinogen.
8.4.1 Possible Mechanisms Leading to a Carcinogenic Response for Chloroform
Possible mechanisms proposed for carcinogenesis include direct
interaction of a chemical or its metabolites with DNA, long-term tissue
injury, stimulation of cell proliferation, immunosuppression, hormonal
imbalances, or release of altered cells from growth control (Weisburger and
Williams, 1980, 1981). The current knowledge of chloroform metabolism and
the acute cellular toxicity of its reactive intermediate metabolites suggest
several different general cellular processes that possibly may lead to
carcinogenic activity (Davidson et al., 1982). One process may involve cell
death induced by the cellular toxicity of chloroform followed by the
consequent stimulation of DNA replication associated with cell multiplication
resulting in an indirect or promoting effect. The continuing process of DNA
replication results in increased proportions of single strand DNA, which is
more susceptible to irreversible binding by reactive intermediates than is
the double strand DNA. Or, it is possible that chloroform acts by stressing
8-64
-------
the fidelity of replication, increasing the possibility of introducing DNA
transcript errors. This general process is supported by the carcinogenicity
studies of Eschenbrenner and Miller (1945) and also Roe et al. (1979), as
well as by the sites of observed tumors. The location of chloroform-induced
primary tumors closely follows the sex and species specificity seen in the
acute organ toxicity as described by many investigators (see section on
toxicity). Eschenbrenner and Miller proposed that necrosis constituted a
necessary precursor for chloroform liver carcinogenicity. This process is
consistent with a threshold effect at the necrotizing doses at which
increased cell death and turnover occur. It is also consistent with the lack
of covalent binding of chloroform or its metabolites to nucleic acid as
observed by Uehleke and coworkers (1977) and by Diaz and Castro (1980), as
well as the negative results of the majority of bacterial mutagenicity assays
(see Chapter 7 on mutagenicity).
A second possible process may involve a suppression of certain
homeostatic mechanisms which maintain cellular integrity. For example,
depletion of a cellular detoxification compound like glutathione could be
expected to raise the background tumor incidence. If this type of process
were the only mechanism involved in chloroform carcinogenicity, a nonlinear
relationship between tumor incidence and exposure dose would be expected,
probably with a threshold dose below which no tumors would occur. This
possible carcinogenic process is supported by the work of Elkstrom and
Hogberg (1980) and others (Brown et al., 1974; Docks and Krishna, 1976), who
have observed that glutathione depletion occurs in chloroform-treated
hepatocytes i_n vitro as well as j_n vivo. Such depletion may reduce the
availability of glutathione for detoxification of certain reactive
intermediates of chloroform metabolism as well as other chemical species.
8-65
-------
Another general process that may lead to carcinogenic!ty involves the
metabolic production of active chloroform intermediate metabolites (phosgene,
carbenes, and free radicals) and their direct interaction with DNA. This
genotoxic process could be expected to be modulated by various homeostatic
mechanisms, such as DNA repair and immunological surveillance, but it is
generally regarded as lacking a threshold. This mechanism is supported by
the data of Agustin and Lim-Sylianco (1978), who found that chloroform
produced a positive result in the micronucleus test and host-mediated
mutagenicity assays in the mouse. Also, the positive mutagenicity results of
Callen et al. (1980) in yeast, the observations of abnormal sperm morphology
in chtoroform-treated mice (Land et al., 1981)* and sister chromatid exchange
in human lymphocytes and mouse marrow (Morimoto and Koizumi, 1983) provide
support for a genotoxic mechanism. Furthermore, the data available from
certain epidemiology studies (Linde and Mesnick, 1980; Cantor et al., 1978;
Hogan et al., 1979) can be considered to be in accord with this process.
In the absence of definitive evidence solely supporting any one of the
likely processes operative in the carcinogenic activity of chloroform, the
risk assessment performed by the Carcinogen Assessment Group considers the
process which is generally accepted to be associated with the greatest risk--
the genotoxic mechanism. The risk of chloroform carcinogenicity by this
process, with its lack of threshold, is appropriately estimated by a
mathematical model predicting zero incremental risk only at zero exposure.
In addition, at the present time, the lack of sufficient understanding about
mechanism of action involved in carcinogenesis does not allow for a
distinction between chemicals acting directly with DNA and those which do
not, nor can those which have not been shown to be genotoxic be considered to
have identifiable population thresholds or be safer than those which are
8-66
-------
considered to be genotoxic (Pereira, 1984). More research is required on the
mechanism of action of chloroform before a threshold model would be
appropriate for carcinogen risk assessment (IARC, 1983; Pereira, 1984).
8.4.2. Selection of Animal Data Sets
For chloroform, several studies in different animal species, strains,
and sexes, run at several doses and different routes of exposure, are
available. A choice must be made as to which data set(s) from these several
studies to use in the risk model. It may also be appropriate to correct for
differences in metabolism between species, and for different routes of
administration. The procedures used herein in evaluating such data are
consistent with the approach of making a maximum-likelihood risk estimate.
The following studies, selected as evidence of the carcinogenic activity
of chloroform from lifetime treatment studies in laboratory animals, have
been used in the mathematical extrapolation models.
8.4.2.1. NCI 1976 Bioassay (Mice): Liver Tumors (Table 8-17)—Male and
female mice were divided into two treatment groups of 50 animals per sex per
dose. A matched control group of 20 animals per sex, and pooled control
groups of 77 males and 80 females, were included in the study. Initially,
male mice received doses of 100 and 200 mg/kg in corn oil by gavage, while
females received 200 and 400 mg/kg beginning at 5 weeks of age, 5 days/week
for 78 weeks, with sacrifice at 92-93 weeks. However, after 18 weeks, these
doses were increased to 150 and 200 mg/kg for males and 250 and 500 mg/kg for
females. The time-weighted average doses were 138 and 277 mg/kg for males
and 237 and 477 mg/kg for females. Mean body weights at terminus were
approximately 35 g for males and 28 g for females. (See Section 8.1 for more
complete details.)
8-67
-------
TABLE 8-17. INCIDENCE OF TUMORS IN EXPERIMENTAL ANIMAL STUDIES
Subjects
B6C3F1 mice (female)a
B6C3F1 mice (male)a
Osborne-Mendel rats,
(male)b
ICI mice, (male)c
Osborne-Mendel rats,
(male)d
Dose,
mg/kg/day
0
238
477
0
138
277
0
90
180
0
60
0
19
38
81
160
Tumor Type
Hepatocel lular carcinoma
Hepatocellular carcinoma
Renal tubular eel 1
adenocarcinoma
Malignant kidney tumors
Renal tubular cell
adenomas and carcinomas
Incidence
rate(%)
0/20(0%)
36/45(80%)
39/41 (95%)
1/18 (6%)
18/50 (36%)
44/45 (98%)
0/19 (0%)
2/50 (4%)
10/50 (20%)
0/50 (0%)
9/48 (19%)
4/301 (1%)
4/313 (1%)
4/148 (3%)
3/48 (6%)
7/50 (14%)
aNCI, 1976.
bNCI, 1976.
CRoe et al., 1979
djorgenson et al., 1985
8.4.2.2. NCI 1976 Bioassay (Rats): Kidney Tumors (Table 8-17)—Chloroform
solutions in corn oil were given by gavage at dose levels of 90 and
180 mg/kg. Fifty male rats were treated at each dose, with colony control
groups (99) consisting of four vehicle-control groups, including matched
controls, given corn oil. The chronic study started at 52 days (7 weeks) of
age and ended with sacrifice of survivors at 111 weeks. Chloroform was
administered in corn oil 5 days/week for the initial 78 weeks. Initial body
weights were about 250 g; by 100 weeks, mean body weights were approximately
8-68
-------
500 g in controls and treated groups. (See pages 8-2 to 8-8 for fuller
details.)
8.4.2.3. Roe et al. 1979 Bioassay (Mice): Kidney Tumors (Table 8-17)--These
investigators conducted three experiments on the carcinogenicity of
chloroform in mice of four strains. Males of the ICI strain were the only
animals that showed an increased incidence of tumors. In the first two
experiments, chloroform in toothpaste vehicle was given by gavage. In the
third experiment, chloroform in arachis oil was tested in ICI mice by gavage
at a dose level of 60 mg/kg/day. Treated and control groups were each
composed of 52 male ICI mice. Chloroform was given by gavage 6 days/week for
80 weeks, beginning at 10 weeks of age and followed by an observation period
of 13 to 24 weeks. Controls were given arachis oil alone. (See Section
8.1.2.2 for fuller details.)
8.4.2.4. Jorqenson et al., 1985 Bioassay (Rats): Kidney Tumors (Table 8-17)
—Chloroform was administered in the drinking water of male rats and female
mice at concentrations of 200, 400, 900, and 1800 mg/L. Only the rats showed
an increased incidence of tumors. The rats were started on treatment at an
average age of 7 weeks and continued on study for 104 weeks. The doses
supplied to the rats in their drinking water correspond to time-weighted
average doses of 19, 38, 81, and 160 mg/kg body weight. The dose-related
increase of renal tumors in the male rats is consistent with the findings of
the earlier NCI (1976) study. However, the lack of response in the mice when
chloroform was administered in the drinking water suggests that earlier
reports of chloroform hepatocellular carcinoma in mice may be related in some
way to the dosing regimen, absorption patterns, or peak blood and tissue
levels of chloroform and its reactive metabolites. Chloroform administered
in corn oil provides lower peak'blood levels than the same dose of chloroform
8-69
-------
administered in aqueous solution (see Chapter 4). However, the daily single
bolus dose given by gavage would be expected to have different
pharmacokinetic characteristics than a similar amount of chloroform
administered in smaller intermittent doses over the twenty-four hour day.
The corn oil carrier used in the NCI gavage study has not been shown to
induce an increase in the incidence of liver tumors in mice, although this
has been postulated.
The studies discussed (see Table 8-17) indicate that the carcinogenic
response to chloroform may be strain-, species-, and sex-related, in addition
to being organ-specific primarily for liver and kidney, which are the target
organs of acute chloroform toxicity, metabolism, and covalent binding. All
long-term studies, with chloroform administered as a single daily oral dose,
have shown a dose-related increased incidence of neoplasms in two strains of
mice, both sexes of one strain and female mice of another strain, renal
neoplasms in male mice of a third strain, and hepatomas in mice of a fourth
strain (sex not reported). When tested in two rat strains, chloroform
produced renal neoplasms in male animals of one strain only (Osborne-Mendel).
Roe et al. (1979) noted that the acute oral LD50 in the B6C3F1 strain of the
NCI study was approximately double that of the four strains of their study,
and hence they suggested that the increased incidence of liver tumors in both
male and female mice of the NCI B6C3F1 strain may be related to the higher
doses used in the NCI mouse study (Table 8-17). Chloroform nephrotoxicity
was also found to be greater in the male ICI mouse strain, suggesting a
greater susceptibility of this sex and strain to cellular damage possibly
associated with a carcinogenic response. However, in the Eschenbrenner and
Miller (1945) strain A mouse study, only the males developed renal necrosis
8-70
-------
but without renal tumor formation upon repeated oral administration of
chloroform.
8.4.3. Interspecies Dose Conversion
8.4.3.1. General Considerations--Careinogenesis is a complex process, not
entirely understood, and no observational evidence exists to validate
satisfactorily the extapolated predictions of carcinogenesis from animal
models to humans. Thus, there is no scientific basis per se for choosing one
extrapolation method over another in extrapolation of the carcinogenic
response. However, in extrapolating the dose-carcinogenic response
relationships of laboratory animals to humans, the doses used in the
bioassays must be adjusted in some way to allow for such differences as size
and metabolic rates. Therefore, other biological information, particularly
interspecies data, is examined for chloroform, since physiologic,
biochemical, and toxicologic responses other than cancer may provide
information as to what factors might be considered and what method is
generally appropriate for the extrapolation from laboratory animals to humans
for this chemical. The major components requiring consideration in
determining an appropriate extrapolation base for scaling carcinogenicity
data in laboratory animals to humans are 1) toxicological data, 2) covalent
binding, and 3) metabolism and kinetics. The biological basis for
extrapolation of dose-carcinogenic response relationships has been outlined
and discussed previously (Davidson, 1984; Parker and Davidson, 1984).
8.4.3.1.1. Toxicologic data. There is a wealth of toxicologic data derived
from numerous studies on the mechanisms of chloroform toxicity (Chapter 5).
Chloroform, as an anesthetic, has a depressive effect on the central nervous
system (CNS) leading to respiratory depression and death. As CMS depression
8-71
-------
tends to occur at similar blood and tissue concentrations across species, the
acute LDcQ tends to be similar. Other toxicologic endpoints have been
measured, but these measures (for example, lowest effective dose producing
liver and kidney pathology) have been made in numerous separate studies and
are difficult to evaluate comparatively across animal species.
It has been well established, however, that chloroform produces liver
and kidney toxicity across species, i.e., man, rat, and mouse. The striking
experimental observation is that even within a species, strain and sex
differences determine the "susceptibility" for liver and kidney toxicity to a
fixed dose of chloroform (Eschenbrenner and Miller, 1945; Culliford and
Hewitt, 1957; Hill et al., 1975). These differences reflect differences of
magnitude of chloroform metabolism as well as other factors at these target
organs (see Chapter 4), and the extent of covalent binding parallels the LD50
and renal and liver cytotoxicity observed with strain and sex (Hill et al.,
1975).
The species and sex differences observed for the carcinogenic response
to chloroform (Table 8-17) are difficult to explain solely on the basis of
genotoxicity, unless it is assumed metabolism differs with species and sex.
On the other hand, the acceptance of a non-genetic mechanism of tumor
induction (from repeated cellular damage resulting in enhanced cellular
regeneration with increased frequency of gene miscoding or promotion of
background tumor cells, etc.) might suggest that the carcinogenic response to
chloroform in laboratory animals, dependent as it is on the dose, species,
strain, and sex (Table 8-17), cannot be readily extrapolated to man (apart
from problems of "threshold"). However, the human population encompasses a
full range of genetic variability, in contrast to the inbred strains of
laboratory animals, and, hence, a carcinogenic response observed only in
8-72
-------
certain species, strains, or sexes in laboratory animals may be expected to
be expressed in the human counterpart.
8.4.3.1.2. Covalent binding—The metabolism of chloroform in mammals gives
rise to reactive intermediate metabolites, including phosgene and, possibly,
carbene and chloride radicals. The irreversible binding to cellular
macromolecules of these reactive chemical species is generally believed to be
responsible not only for hepatic and renal damage from chloroform, but also
for the carcinogenic response, although the mechanism for the latter has not
been elucidated. The data of Figure 8-4 from Ilett et al. (1973) show that
in C57BL mice the amount of covalent binding increases with dose up to about
450 mg/kg, which seems to be a "saturation of metabolism" dose for these
mice. Furthermore, the amount of covalent binding from metabolism parallels
the dose-response curve for cytopathologic changes associated with acute
toxicity. Uehleke and Werner (1975) (Figure 8-5) found that, across species
(mouse, rat, rabbit, man), irreversible binding from chloroform metabolism,
as measured with hepatic microsomes from these species, was greater in humans
than in rabbits, and was greater in rabbits than in rodents. This indicates
that the capacity to metabolize chloroform/unit of microsomes is much greater
in man than rodent. That is, while liver weights between species are
proportional to body weight (W^O), the greater activity of microsome/unit
weight in the metabolism of chloroform indicates that man is certainly not
"limited" in metabolic capacity as compared with rodents, and indeed, may
have a capacity greater than expected from the usual extrapolations on body
weight or surface area.
8-73
-------
I 3
o>
"o
O
5
z
5
i-
z
UJ
O
O
I
O KIDNEY
• LIVER
6
CHLOROFORM DOSE, mmol/kg
Figure 8-4. Effect of increasing dosage of i.p.-injected ^c-chloroform
on extent of covalent binding of radioactivity in vivo to liver and kidney
proteins of male mice 6 hours after administration.
Source: llettetal. (1973).
8-74
-------
o
X
o
E
e
>
H
>
H
O
O
o
oe
I
PROTEIN
O HUMAN
D RABBIT
10
20
30 40
TIME. min.
50
60
70
Figure 8-5. Comparison of irreversible binding of radioactivity from 14C-CHCl3
to protein and lipid of microsomes from normal rabbit, rat, mouse, and human
liver incubated in vitro at 37°C in 62-
Source: Uehleke and Werner (1975).
8-75
-------
A basic question concerning chloroform is whether it has genotoxicity
potential, i.e., whether covalent binding to cellular DNA occurs. Most
investigators have found only minimal covalent binding to DNA (p. 4-58).
Reitz et al. (1980) measured DNA covalent binding in liver and kidneys of
mice after an oral dose of 240 mg/kg of ^C-chloroform. These investigators
found limited evidence that the metabolism of chloroform by the mouse in vivo
results in genotoxicity, i.e., detectable DNA alkylation. DNA alkylation
represented a binding index, CBI, of 1.5, a value well above the detection
limit after correction for background. The remaining question of whether the
degree of chloroform alkylation represents a highly significant genotoxicity
as opposed to "very low" DNA aklylation (as judged by Reitz et al.), depends
in large part on the validity of the experiments' methodology. Reitz et al.
include a direct comparison, in tabular form, of the alkylation of chloroform
with that of potent alkylating carcinogens. However, their chloroform DNA
alkylation experiment was not actually conducted with these "positive
controls." An overall conclusion can be made that while DNA alkylation did
occur, the comparative extent is open to question.
8.4.3.1.3. Metabolism and kinetics
8.4.3.1.3.1. Absorption—Chloroform is virtually completely absorbed
from the gastrointestinal tract when given by gavage in olive oil (60 mg/kg)
to mice, rats, and monkeys, and when given orally to man (0.1 to 1.0 g)
(Brown et al., 1974; Taylor et al., 1974; Fry et al., 1972). Withey et al.
(1982) observed that chloroform given in a corn oil vehicle was absorbed more
slowly than chloroform given in water. Nonetheless, the experimental data
support complete absorption of chloroform dosage for the conditions of the
NCI, 1976; Roe et al., 1979; and Jorgenson et al., 1985 carcinogenicity
studies and show no difference in absorption characteristics among species,
8-76
-------
including man.
8.4.3.1.3.2. First-pass effect—Chloroform, after gavage to animals or
oral ingestion by humans, is absorbed and transported by the portal blood
flow to the liver, where a portion is extracted and metabolized (depending on
dose) and a portion excreted through the lungs. In both animals and humans,
pulmonary excretion of unchanged chloroform is dose-dependent (Tables 8-18
and 8-19). Data are available (Table 8-18) from the studies of Brown et al.
(1974) for the pulmonary excretion in mice and rats after an oral dose of
60 mg/kg (6 percent and 20 percent of the dose, respectively). Pulmonary
excretion is the principal route of excretion of unmetabolized chloroform.
Therefore, the amounts of chloroform metabolized possibly contributing to the
carcinogenic response after a 60-mg/kg dose are 56.4 mg/kg for mice and
48.9 mg/kg for rats. For higher doses of the NCI studies, experimental data
are not available, and hence the body burdens must be calculated using
94 percent and 80 percent for the gavage dose to mice and rats, respectively,
although the resultant calculated body burdens will be higher than in
actuality because the portion (percent) of the dose excreted unchanged
through the lungs increases with the oral dose (see data for humans,
Table 8-19).
8.4.3.1.3.3. Saturation of metabolism—The mammalian capacity to
metabolize a chemical compound may nearly always be "saturated" if the dose
to the organism is high enough. Chloroform is no exception. While there are
not specific experimental data to illustrate chloroform saturation kinetics
in laboratory animals, the amount of covalent binding versus ^C-chloroform
dose from the data of Ilett et al. (1979) (Fig. 8-4) suggests in mice a
saturation of metabolism at about 75 mg/kg/hour. Fry et al. (1972) conducted
8-77
-------
TABLE 8-18. SPECIES DIFFERENCE IN THE METABOLISM OF 14C-CHLOROFORM
(ORAL DOSE OF 60 mg/kg)a
^C-radioactivity 43 hours after dose
Mean values as percent dose
Species No.
Mice 19
CF/LP,
CBA, C57
strains
Rats 6
S-D
Squirrel 6
Monkeys
Expired
14CHC13 or
metabolites
6.1
19.7
78.7
Expired
14C02
85.1
65.9
17.6
Urine +
feces Carcass
2.6 1.8
7.6 NR
2.0 NR
Total
95.6
93.2
98.3
Recalculated from the data of Brown et al., 1974.
NP = Nnt r-prnr-rloH
TABLE 8-19. PULMONARY EXCRETION OF CHLOROFORM FOLLOWING ORAL DOSE
Excretion of ^CHCIj: Percent of Dose3
Subjects
8 M and F
1
1
1
Excretion of J3CC>2 following 0.
Subjects
Male (1)
Female (62.7 kg) (1)
Dosea
(g)
0.5
1.0
0.25
0.10
5-g dose of 7^<
0.5
2.1
0.5
Mean for 8 hoursb
40.3
64.7
12.4
nil
17
Range
.8 to 66.6
NA
NA
NA
IHCIj: Cumulative percent ofdoseb
Time after dose
1.75 2.5
24.1 35.9
10.7 28.3
5.
49.
47.
5 7.
2 50.
5 48.
6
6
5
Recalculated from the data of Fry et al., 1972.
bWithin 4% of value calculated for infinite time.
NA = Not applicable.
8-78
-------
studies in man (Table 8-19) that indicate saturation of metabolism in man at
relatively low doses. These investigtors administered chloroform orally to
volunteers over a dose range of 0.10 to 1.0 g (about 1.5 to 15 mg/kg), and
found that the portion of the dose excreted unchanged via the lungs increased
from zero for the lowest dose to 65% with the highest dose. These results
suggest that chloroform metabolism is rate-limited in man. Since a
diminishing proportion of dose is metabolized with increased dose, the Vmax
for man appears to approximate 50 mg/hour or 0.7 mg/kg/hour as estimated from
the highest dose, 15 mg/kg.
8.4.3.1.4. Interspecies scaling of metabo1ism--From the data of Brown et al.
(1974) and of Fry et al. (1972), which provide information on the metabolism
of chloroform in mice (3 strains), rats, squirrel monkeys, and man, it is
possible to estimate the amount metabolized (mg) of an orally administered
common dose of 60 mg/kg body weight (probably a near saturating dose; see
above). These data, when plotted as the logarithm of the amount metabolized
versus the logarithm of the species body weight in accordance with the
allometric relationship Y = aWn ( Łn Y = Łn a + n^n W), gives a regression
line that closely fits the species data points (Fig. 8-6). The slope of the
line, n = 0.65, provides strong evidence that chloroform metabolism in these
four species is proportional to their surface areas (kr/^). Figure 8-6
provides justification for the extrapolation of the carcinogenic response
(metabolism-dependent) of mice and rats of the Jorgenson, Roe, and NCI
studies to man on the basis of body surface area (W2/3).
8.4.3.2. Calculation of Human Equivalent Doses—Available information on
metabolism and pharmacokinetics pertinent to the conditions of the Jorgenson
et al., 1985; Roe et al., 1979; and NCI, 1976 carcinogenicity assays may be
summarized as follows:
8-79
-------
8
•o
o
2* 6
"o
.o
a
o
E
co 4
5
z
o
o 2
E
468
Łn Body Weight
10
12
Figure 8-6. Allometric relationship (Y = aWn) between species body weight (in order:
mouse, rat, squirrel monkey, and man) and the amount metabolized of a common oral
dose of chloroform as calculated from the data of Fry et al. (1972) and Brown et al.
(1974). The species body weights assumed were: mouse, 30 g; rat, 300 g; squirrel
monkey, 850 g; and man, 70 kg. The slope of the regression line is 0.65, indicating
that metabolism of chloroform in these species is proportional to their surface area.
8-80
-------
1. After gavage or other oral administration to mice and rats or
oral administration to man, chloroform is rapidly and virtually
completely absorbed.
2. After gavage or oral administration, a portion of the dose
(increasing with the dose) is excreted via the lungs unchanged.
Metabolism of chloroform is "saturated" or near saturation at the
dose levels of the bioassays in which the chloroform was
administered as a single daily bolus. The effective dose (amount
metabolized to reactive metabolites) is the gavage dose minus the
percentage of dose excreted unchanged. For mice given 60 mg/kg
(conditions of Roe study), this is about 6 percent, and for rats
given the same dose, 20 percent. However, for the NCI study the
doses approximate 200 to 500 mg/kg/day for rats and mice, and hence
a 20 percent correction is very conservative and probably leads to
an overestimate of the amounts metabolized from these doses.
Virtually all of the chloroform administered in the drinking water
study is metabolized due to the dosing regimen.
3. The adipose tissue/blood partition coefficient for chloroform
is high (35), and the half-time of chloroform residence in adipose
tissue is relatively long in the rat and man (about 2 hr and 36 hr,
respectively); daily gavage doses may not be expected to be
completely cleared from the body within the 24-hr dosing interval,
and therefore chronic daily dosing, as occurs in the animal
bioassays, may result in the constant presence of chloroform (and
metabolites) in the body (except for weekends).
8-81
-------
4. There is an experimental basis for determining that the
metabolism of chloroform across species (including man) is
proportional to the surface area of the species (W*-/3).
5. There is no evidence to suggest any qualitative difference in
the metabolic pathways of the three species (mouse, rat, and man)
for chloroform.
6. In mice, covalent binding (from chloroform metabolism) has
been demonstrated in both liver and kidney (target organs) to be
proportional to the dose.
The total information available on the pharmacokinetics and metabolism
of chloroform in the mouse and rat is sparse and does not include information
for doses that bracket the high doses of the NCI carcinogenicity assay,
although such information is available for the doses used in the Roe et al .
assay. Nonetheless, the information on kinetics and metabolism justifies the
use of surface area (W^/3) as a basis for dose extrapolation from
experimental animals to man and the calculation of human equivalent doses
(see Figure 8-6) .
Using this pharmacokinetic and metabolic information, the lifetime
average human equivalent doses for the chloroform NCI mouse and rat
carcinogenicity bioassays (1976), for the chloroform-Roe et al . ICI mouse
carcinogenicity bioassay (1979), and for the chloroform in drinking water-
Jorgenson et al . rat carcinogenicity bioassay (1985), are calculated as
follows. The lifetime average daily exposure (LAE) of bioassay animals is
given by
duration exposure (wk) doses per wk
x effective dose,
duration experiment (wk) 7 nig /animal /day
8-82
-------
where the effective dose is the average assay dose per day in milligrams,
adjusted for pharmacokinetic and metabolic parameters (in this case decreased
by 6 percent for mice and by 20 percent for assay rats administered single
bolus daily doses). For humans, the LAE is given as follows:
LAE^| = LAE^ x scaling factor.
For chloroform, the appropriate basis for extrapolation from mouse or rat to
man has been determined from the experimental data to be that of body surface
area (W2/3). Therefore, the scaling factor is
body weight of man (70 kg)
^terminal body weight of assay animal (kg)y
Tables 8-20 through 8-23 show, respectively, the calculated lifetime
average human equivalent doses (LAEH) with corresponding tumor incidence for
the chloroform-NCI mouse bioassay, the chloroform-NCI rat bioassay, the
chloroform-Roe et al. mouse bioassay, and the chloroform-Jorgenson et al. rat
bioassay.
Figure 8-7 shows visually the relationship between the equivalent human
exposure dose (LAEH) and bioassay tumor incidence. Dose expression in the
form of human equivalent dose effectively "normalizes" the varying conditions
of the Individual bioassays, particularly species differences, to the single
human standard, hence facilitating a common dose-response comparison, as
shown in Figure 8-7. It should also be noted that, whereas in Figure 8-7 the
equivalent human dose in units of mg/m2/day is plotted, plots of equivalent
human dose in units of nig/day or (mg/kg)/day provide dose-response
relationships of an entirely similar form since all three dose units (nig/day,
mg/kg/day, and (mg/m2)/day) are directly proportional and interconvertible
8-83
-------
TABLE 8-20. CONTINUOUS HUMAN EQUIVALENT DOSES AND INCIDENCE OF
HEPATOCELLULAR CARCINOMAS IN MALE AND FEMALE B6C3F1 MICE
Gavage dose, Lifetime average
mouse daily dose
(mg/kg)b (mg/mouse)
0, female
238, female
477, female
0, male
138, male
277, male
3.74
7.61
2.75
5.52
Equivalent lifetime average
i human exposure dose^
(mg/d)
694.7
1394.5
436.2
876.3
(mg/m2)/d
375.5
753.8
235.8
473.7
(mg/kg)/d
9.9
19.9
6.2
12.5
Tumor
incidence
(*)
0/20 (0)
36/45 (80)
39/41 (95)
1/18 (6)
18/50 (36)
44/45 (98)
aWhere the terminal average weight of the male mice is 0.035 kg, female mice
is 0.028 kg,and standard man, 70 kg with 1.85 m2 surface area. Mouse-to-man
extrapolation factor is (70/0.035)2/3 for male mice and (70/0.028)2/3 for
female mice.
bSum of all doses divided by number of days dosed.
cLifetime average daily dose = mg/kg dose x body weight x 5/7 days x 78/91
weeks (corrected for 6% dose unmetabolized).
SOURCE: NCI, 1976.
TABLE 8-21. CONTINUOUS HUMAN EQUIVALENT DOSES AND INCIDENCE OF RENAL
TUBULAR-CELL ADENOCARCINOMAS IN MALE OSBORNE-MENDEL RATS
Gavage dose,
rat
(mg/kg)
0, male
90, male
180, male
Lifetime averag
daily dose
(mg/rat)
18.07
36.14
Equivalent lifetime average
e human exposure dosea
(mg/d)
487.0
973.9
(mg/m2)/d (mg/kg) /d
263.2
526.4
6.96
13.92
Tumor
incidence
(*)
0/99 (0)
4/50 (4)
12/50 (20)
aWhere the terminal average weight of the male rats is 0.5 kg and standard
man, 70 kg with 1.85 m2 surface area. Rat-to-man extrapolation factor is
(70/0.5)2/3.
bLifetime average daily dose = mg/kg dose x body weight x 5/7 days x 78/111
weeks (corrected for 20% dose unmetabolized).
SOURCE: NCI, 1976.
8-84
-------
TABLE 8-22. CONTINUOUS HUMAN EQUIVALENT DOSES AND INCIDENCE OF MALIGNANT
KIDNEY TUMORS IN MALE ICI MICE
Gavage dose,
mouse
(mg/kg)
Lifetime average
Equivalent lifetime average
human exposure dosea
daily doseb
(mg/mouse) (mg/d) (mg/m2)/d (mg/kg)/d
Tumor
incidence
0, male
60, male
1.45
229.4
124.0
3.38
0/50 (0)
9/48 (19)
aWhere the terminal average weight of the male mouse is assumed to be
0.035 kg and standard man, 70 kg with 1.85 m2 surface area. Mouse-to-man
extrapolation factor is (70/0.035)2/3.
bLifetime average daily dose = mg/kg dose x body weight x 6/7 days x 80/93
weeks (corrected for 6% dose unmetabolized).
SOURCE: Roe et al., 1979.
TABLE 8-23. CONTINUOUS HUMAN EQUIVALENT DOSES AND INCIDENCE OF RENAL
TUBULAR CELL ADENOMAS AND ADENOCARCINOMAS IN MALE OSBORNE-MENDEL RATS
Dosea
0
19
38
81
160
Lifetime averagi
daily dose
(mg/rat)
—
8.9
19.8
42.2
84.4
Equivalent lifetime average
2 human exposure dosea
(mg/d)
—
240
480
1038
2021
(mg/m2)/d
—
130
257
561
1094
(mg/kg)/d
—
3.43
6.86
14.83
28.87
Tumor
incidence
%
4/301 (1)
4/313 (1)
4/148 (3)
3/48 (6)
7/50 (14)
aWhere the terminal average weight of the male rat is 0.5 kg and the standard
man 70 kg with 1.85 m2 surface area. Mouse-to-man extrapolation factor is
(70/0.035)2/3.
bLifetime average daily dose = mg/kg dose x body weight x 7/7 days x 104/111
weeks (corrected for 0% dose unmetabolized).
SOURCE: Jorgenson et al., 1979.
8-85
-------
100
90
§ 80
O
cc
O
70
• MICE (FEMALE), LIVER (NCI)
°\ • MICE (MALE). LIVER (NCI)
LU 6Q _ V MICE (MALE). KIDNEY (ROE ET AL.)
CC O RATS (MALE), KIDNEY (NCI)
Ul
ffl A RATS (MALE), KIDNEY
^ t-f. (JORGENSON ET AL.)
*- 50
s
g
a
40
u
z
Ul
a
30
cc
O
10
ol L^ I ^ I I I I I I L
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
HUMAN EQUIVALENT DOSE, (mg/m2)/d metabolites
Figure 8-7. The relationship between the equivalent human exposure dose and bioassay
tumor incidence.
8-86
-------
one to another by means of the conversion factors of 70 kg and 1.85 m2
surface area for the "standard" man.
8.4.4. Choice of Risk Model
8.4.4.1. General Considerations—The data used by the CAG for quantitative
estimation are of two types: 1) lifetime animal studies and 2) human studies
where excess cancer risk has been associated with exposure to the agent. In
animal studies it is assumed, unless evidence exists to the contrary, that if
a carcinogenic response occurs at the dose levels used in the study, then
responses will also occur at all lower doses, with an incidence determined by
the extrapolation model.
There is no universally acceptable solid scientific basis for any
mathematical extrapolation model that relates exposure to cancer risk at the
extremely low concentrations that must be dealt with in evaluating
environmental hazards. For practical reasons, such low levels of risk cannot
be measured directly, either by animal experiments or by epidemiologic
studies. We must, therefore, depend on our current understanding of the
mechanisms of carcinogenesis for guidance as to which risk model to use. At
the present time, the dominant view of the carcinogenic process involves the
concept that most cancer-causing agents also cause irreversible damage to
DNA. This position is reflected by the fact that a very large proportion of
agents that cause cancer are also mutagenic. There is reason to expect that
the quantal type of biological response, which is characteristic of
mutagenesis, is associated with a low-dose linearity and linear nonthreshold
dose-response relationship. Indeed, there is substantial evidence from
mutagenicity studies with both ionizing radiation and a wide variety of
chemicals that this type of dose-response model is the appropriate one to
use. This is particularly true at the lower end of the dose-response curve;
8-87
-------
at higher doses, there can be an upward curvature, probably reflecting the
effects of multistage processes on the mutagenic response. The low-dose
linearity and nonthreshold dose-response relationship is also consistent with
the relatively few epidemiologic studies of cancer responses to specific
agents that contain enough information to make the evaluation possible (e.g.,
radiation-induced leukemia, breast and thyroid cancer, skin cancer induced by
arsenic in drinking water, liver cancer induced by aflatoxin in the diet).
There is also some evidence from animal experiments that is consistent with
the linear nonthreshold model (e.g., liver tumors induced in mice by
2-acetylaminofluorene in the large-scale EDg^ study at the National Center
for Toxicological Research and the initiation stage of the two-stage
carcinogenesis model in rat liver and mouse skin).
Because its scientific basis, although limited, is the best of any of
the current mathematical extrapolation models, the nonthreshold model, which
is linear at low doses, has been adopted as the primary basis for risk
extrapolation to low levels of the dose-response relationship. The risk
estimates made with such a model should be regarded as conservative,
representing the plausible upper limits for the risk; i.e., the true risk is
not likely to be higher than the estimate, but it could be lower.
The mathematical formulation chosen to describe the dose-risk
relationship at low doses is the linearized multistage model. This model
employs enough arbitrary constants to be able to fit almost any monotonically
increasing dose-response data, and it incorporates a procedure for estimating
the largest possible linear slope (in the 95% confidence limit sense) at low
extrapolated doses that is consistent with the data at all dose levels of the
experiment.
8-88
-------
The methods used by the Carcinogen Assessment Group (CAG) for
quantitative assessment are consistently conservative, i.e., tending toward
high estimates of risk. The most important part of the methodology
contributing to this conservatism is the linear nonthreshold extrapolation
model. There are a variety of other extrapolation models that could be used,
all of which would give lower risk estimates. These alternative models have
been used by the CAG for comparison purposes, and the results for chloroform
may be found in the Appendix of this document. The CAG feels that with the
limited data available from these animal bioassays, especially at the higher
dose levels required for testing, almost nothing is known about the true
shape of the dose-response curve at low environmental levels. The position
is taken by the CAG that the risk estimates obtained by use of the low-dose
linear nonthreshold model are plausible upper limits, and that the true risk
could be lower.
In terms of the choice of animal bioassay as the basis for
extrapolation, where more than one acceptable study is available, a general
approach is to use the most sensitive responder, on the assumption that
humans are as sensitive as the most sensitive animal species tested.
Extrapolations from animals to humans can be done on the basis of
relative body weights, surface areas, metabolic rates, or other measures.
The general approach is to use the extrapolation base (mg/kg, surface area,
etc.) that can be appropriately justified by the experimental data from
animals and humans. However, it is not usually clear which extrapolation
base is the most appropriate for the carcinogenic response per se. In cases
where there are insufficient experimental data to determine an appropriate
extrapolation base either directly or indirectly, the most generally employed
and conservative method is used, i.e., extrapolation from animal dose to a
8-89
-------
human equivalent dose on the basis of relative surface area (kr/^). For the
chloroform studies in rats and mice, the use of an extrapolation based on
surface area (W2/3) rather than body weight (Wl-0) would increase the unit
risk estimates by factors of approximately 6 to 13. However, for chloroform,
experimental data on metabolism and kinetics are used in dose extrapolation
from mouse to human.
8.4.4.2. Mathematical Description of Low-Dose Extrapolation Model--Let P(d)
represent the lifetime risk (probability) of cancer at dose d. The
multistage model has the form
P(d) = 1 - exp -(q0 + q^d + q2d2 + ... + q| 0, i = 0, 1, 2, ..., k
Equivalently,
Pt(d) = 1 - exp -(q^d + q2d2 + ...
where
Pt(d) = P(d) - P(0)
1 - P(0)
is the extra risk over background rate at dose d, or the effect of treatment.
The point estimate of the coefficients qn-, i =0, 1, 2, ..., k, and
consequently the extra risk function, P^(d), at any given dose d, is
calculated by maximizing the likelihood function of the data.
The point estimate and the 95-percent upper confidence limit of the
extra risk, P^(d), are calculated by using the computer program GLOBAL83,
developed by Howe (1983). At low doses, upper 95-percent confidence limits
on the extra risk and lower 95-percent confidence limits on the dose
producing a given risk are determined from a 95-percent upper confidence
limit, q^*, on parameter qj_. Whenever q^ > 0, at low doses the extra risk
8-90
-------
Pt(d) has approximately the form P^(d) = qj_* x d. Therefore, q^* x d is a
95-percent upper confidence limit on the extra risk, and R/qj* is a 95-
percent lower confidence limit on the dose, producing an extra risk of R.
Let LQ be the maximum value of the log-likelihood function. The upper-limit,
q^*, is calculated by increasing q^ to a value q^* such that when the log-
likelihood is remaximized subject to this fixed value, q^*, for the linear
coefficient, the resulting maximum value of the log-likelihood Lj_ satisfies
the equation
2 (L0 LI) = 2.70554
where 2.70554 is the cumulative 90-percent point of the chi-square
distribution with one degree of freedom, which corresponds to a 95-percent
upper limit (one-sided). This approach of computing the upper confidence
limit for the extra risk P^-(d) is an improvement on the Crump et al. (1977)
model. The upper confidence limit for the extra risk calculated at low doses
is always linear. This is conceptually consistent with the linear
nonthreshold concept discussed earlier. The slope, q^*, is taken as an upper
bound of the potency of the chemical in inducing cancer at low doses. (In
the section calculating the risk estimates, P^-(d) will be abbreviated as P.)
In fitting the dose-response model, the number of terms in the
polynomial is chosen equal to (h-1), where h is the number of dose groups in
the experiment, including the control group.
8.4.4.3. Adjustment for Less than Lifespan Duration of Experiment—If the
duration of experiment Le is less than the natural lifespan of the test
animal L, the slope q^*, or more generally the exponent g(d), is increased by
multiplying a factor (L/Le)3. We assume that if the average dose d is
continued, the age-specific rate of cancer will continue to increase as a
constant function of the background rate. The age-specific rates for humans
8-91
-------
Increase at least by the second power of the age and often by a considerably
higher power, as demonstrated by Doll (1971). Thus, we would expect the
cumulative tumor rate to increase by at least the third power of age. Using
this fact, we assume that the slope q^*, or more generally the exponent g(d),
would also increase by at least the third power of age. As a result, if the
slope q^* [or g(d)] is calculated at age Le, we would expect that if the
experiment had been continued for the full lifespan L, at the given average
exposure, the slope q^* [or g(d)] would have been increased by at least
(L/Le)3.
This adjustment is conceptually consistent with the proportional hazard
model proposed by Cox (1972) and the time-to-tumor model considered by Crump
(1979), where the probability of cancer by age t and at dose d is given by
P(d,t) = 1 - exp [-f(t) * g(d)]
8.4.4.4 Additional Low-Dose Extrapolation
In addition to the multistage model currently used by the CAG for low-
dose extrapolation, three more models, the probit, the Weibull, and the one-
hit models are used for comparison (Appendix A). These models cover almost
the entire spectrum of risk estimates that could be generated from existing
mathematical extrapolation models. Generally statistical in character, these
models are not derived from biological arguments, except for the multistage
model, which has been used to support the somatic mutation hypothesis of
carcinogenesis (Armitage and Doll, 1954; Whittemor^, 1978; Whittemore and
Keller, 1978). The main difference among these models is the rate at which
the response function P(d) approaches zero or P(0) as dose d decreases. For
instance, the probit model would usually predict a smaller risk at low doses
than the multistage model because of the difference of the decreasing rate in
the low-dose region. However, it should be noted that one could always
8-92
-------
artificially give the multistage model the same (or even greater) rate of
decrease as the probit model by making some dose transformation or by
assuming that some of the parameters in the multistage model are zero. This,
of course, would not be reasonable if the carcinogenic process for the agent
were not known a priori.
8.4.5. Unit Risk Estimates
This section deals with the unit risk for chloroform in air and water
and the potency of chloroform relative to other carcinogens that the CAG has
evaluated.
8.4.5.1. Definition of Unit Risk--The unit risk estimate for an air or water
pollutant is defined as the increased lifetime cancer risk occurring in a
hypothetical population in which all individuals are exposed continuously
from birth throughout their lifetimes to a concentration of 1 ng/m^ of the
agent in the air they breathe, or to 1 pg/L in the water they drink. This
calculation is done to estimate in quantitative terms the impact of the agent
as a carcinogen. Unit risk estimates are used for two purposes: 1) to
compare the carcinogenic potencies of several agents with each other, and 2)
to give a crude indication of the population risks that might be associated
with air or water exposure to these agents, if the actual exposures were
known.
8.4.5.2. Calculation of the Slope of the Dose-Risk Relationship for
Chloroform—Evidence of carcinogenic activity of chloroform from lifetime
treatment studies in laboratory animals includes significantly (P < 0.05)
increased incidences of hepatocellular carcinomas in female and male B6C3F1
mice (Table 8-20) and kidney tumors in male Osborne-Mendel rats (Tables 8-21
and 8-23); and kidney tumors in male ICI mice (Roe et al., 1979. Table 8-
8-93
-------
22). These data sets are used to estimate the carcinogenic risk of
chloroform.
To convert animal doses into equivalent human doses, the administered
dose is expressed as an average daily dose, reduced by the unmetabolized
portion (estimated in Section 8.4.2.1.3 to be 6 percent for mice and 20
percent for rats when administered by gavage as a bolus in corn oil, and
0 percent when in drinking water), and scaled to humans using a surface-area
correction. This is to account for the differences in metabolic rate as well
as the variations in absorption patterns of dissimilar dosing.
Using the incidence data in Tables 8-20 to 8-23 and the corresponding
human equivalent doses, the maximum likelihood estimates of the parameters
were calculated for each of the four models referred to above (see Table A-l
in Appendix A). These models can be used to calculate either point estimates
of risk at a given dose, or the virtually safe dose for a given level of
risk. The upper-bound estimates of the risk at 1 mg/kg/day, calculated from
each of these models on the basis of different data sets, are presented in
Table 8-24. From this table, it is observed that the multistage model
predicts a comparable risk on the basis of the different data sets, while the
probit and Weibull models are very unstable and predict a wide range of risk
depending on which data base is used for the risk calculation. The dose-risk
slope value, q^*, for chloroform is represented by the geometric mean of the
slopes obtained from the linearized multistage model, on the basis of liver
tumor data for female and male mice. Although the slope calculated from the
data for female mice is greater than that calculated from the data for male
mice, the estimates from both data sets are combined because the data for
males include an observation at a lower dose, and the response at this dose
does not appear to be inconsistent with the female data, if the linear dose-
8-94
-------
TABLE 8-24. UPPER-BOUND ESTIMATES OF CANCER RISK OF 1 mg/kg/day, CALCULATED
BY DIFFERENT MODELS ON THE BASIS OF DIFFERENT DATA SETSa
Data Base Multistage Probit Weibull One-Hit
Liver tumors in
female mice 1.8 x 10-lb 2.1 x io-l 4.8 x 10-1 1.3 x iQ-lb
(NCI, 1976)
Liver tumors in
male mice 3.3 x 1Q-2 5.7 x lo-ll 3.2 x io-3 1.5 x io-l
(NCI, 1976)
Kidney tumors in
male rats 2.4 x io-2 3.9 x 1Q-4 3. x 1Q-3 2.5 x 1Q-2
(NCI, 1976)
Kidney tumors in
male mice 1.0 x io-l NA NA 1.0 x io-l
(Roe et al., 1979)
Kidney tumors in
male rats 4.4 x io-3 g.O x io-5 4.3 xio-4 5.4 x io-3
(Jorgenson et al.,
1985)
^Upper-bound estimates are calculated by the one-sided 95% confidence limit.
t>At 1 mg/kg/day the dose-response curve diverges from a straight line. For
lower doses the dose-response slope is 2.0 x io-l per mg/kg/day.
NA = not applicable. Models are not applicable because there is only one
dosed group.
response relationship is assumed. This is also consistent with the various
data sets for mice if they are expressed in terms of human equivalent dose as
shown in Figure 8-7.
Thus, the slope is
(2.0 x 10-1 x 3.3 x 10-2)? = 8.1 x 10-2
This number differs little from the geometric mean of the q^* (upper-bound of
the linear parameter) calculated from all five data sets, and thus is used
herein as the slope for calculating risk at low doses.
The geometric mean estimate is consistent also with the risk calculated
by pooling both sexes of B6C3F1 mice and then estimating the dose-risk slope,
8-95
-------
which is 8.9 x 10-2. The sexes were pooled for comparison purposes only, and
different data sets cannot, in general, be pooled. In this case, however,
pooling may be justified because the dose ranges overlap, the responses are
identical, and as mentioned previously, the doses in terms of human
equivalent doses appear to be along the same dose-response curve (see Figure
8-7).
8.4.5.3. Risk Associated with 1 uq/m3 of Chloroform in Air—No studies exist
for directly estimating cancer risks from inhaled chloroform. In the absence
of such information, the risk from inhaled chloroform is considered the same
as the risk from orally ingested chloroform. This assumption is supported by
the presence of distal-site tumors in mice and rats ingesting chloroform.
The dose-response slope from Section 8.4.4.2 can be used to estimate the risk
from 1 jag/m3 of chloroform in air. It is assumed that low doses of
chloroform in air can be completely absorbed. For a person weighing 70 kg
and breathing 20 m3/day, 1 ng/m3 of chloroform in air is an effective dose of
d = (1 pg/m3)(10-3 mg/Vg)(20 m3/day)/(70 kg) = 2.9 x 1Q-4 mg/kg/day
The risk at this dose may be as high as
P = (8.1 x 10-2)(2.9 x 10-4) = 2.3 x 1Q-5
8.4.5.4. Risk Associated with 1 uq/Liter of Chloroform in Drinking Water--
For drinking water exposure, it is assumed that 100 percent of the chloroform
in drinking water can be absorbed, and that water intake is 2 L/day. Under
these assumptions, the daily dose from consumption of water containing 1 pg/L
(1 ppb) of chloroform is calculated as follows:
d = 1 pg/L x 2 L/day x 10-3 mg/^g x 1/70 kg = 2.9 x 10-5 mg/kg/day
Therefore, the risk associated with 1 pg/L of chloroform in water is
P = 8.1 x 10-2 x 2.9 x 10-5 = 2.3 x 10-6
8-96
-------
8.4.5.5. Interpretation of Unit Risk Estimates — For several reasons, the
unit risk estimate based on animal bioassays is only an approximate
indication of the absolute risk in populations exposed to known carcinogen
concentrations. There may be important differences in target site
susceptibility, immunological responses, hormone function, dietary factors,
and disease. In addition, human populations are variable with respect to
genetic constitution and diet, living environment, activity patterns, and
other cultural factors.
The unit risk estimate can give a rough indication of the relative
potency of a given agent as compared with other carcinogens. The comparative
potencies of different agents are more reliable when the comparisons are
based on studies in the same test species, strain, and sex, and by the same
route of exposure, although ordinarily the risk should be independent of
route of exposure except in special circumstances, for example, nasal or lung
carcinomas with inhalation exposure, or forestomach tumors with gavage
administration.
The quantitative aspect of carcinogen risk assessment is included here
because it may be of use in the regulatory decision-making process, e.g.,
setting regulatory priorities, evaluating the adequacy of technology-based
controls, etc. However, it should be recognized that the estimation of
cancer risks to humans at low levels of exposure is uncertain. At best, the
linear extrapolation model used here provides a rough estimate of the upper-
limit of risk; i.e., it is not likely that the true risk would be much more
than the estimated risk, but it could very well be considerably lower. Thus,
risk estimates for chloroform presented in this chapter should not be
regarded as immutable representations of the true cancer risks; however, the
estimates presented may be factored into regulatory decisions to the extent
8-97
-------
that the concept of upper limit risks is found to be acceptable. The slope
estimates can be used to compare the relative carcinogenic potency of
chloroform to that of other potential human carcinogens, in addition to being
used to calculate upper-bound incremental risks at low levels of exposure.
8.4.5.6. Reconciliation of Unit Risk Estimates with Epidemioloqical
Evidence—The unit risk estimates are consistent with available epidemiologic
data such as the odd ratios for bladder cancer, which were estimated to range
form 1.04 to 1.69 (Table 8-14). According to a survey of 76 water supply
systems in the United States, the chloroform measurements ranged from 1 ^g/L
to 112 pg/L. A rough estimate of the cancer risk on the basis of these
statistics ranges from
B = (1.04 - 1) x 7 x 10-4/112 = 3 x 10-7/(^g/L)
to
B = (1.69 - 1) x 7 x 10-4/1 = 5 x 10-4/(pg/L)
where 7 x 10-4 is the estimated background bladder cancer mortality rate in
the United States.
The unit risk estimate for chloroform in water (estimated in Section
8.4.5.4 to be 2.3 x 10-6) js well within this range.
8.4.5.7. Discussion—Since the carcinogenic activity of chloroform is
generally considered to reside in its reactive intermediate metabolites, the
amount of chloroform metabolized is considered to be the effective dose and
is used in calculating the dose-response relationship. The use of the amount
of chloroform undergoing biotransformation as the effective dose may not
eliminate all the uncertainties associated with the low-dose extrapolation,
however, because the dose at the receptor sites may not be linearly
proportional to the total amount metabolized. Thus, the true shape of the
dose-response relationship would still be unknown. However, it seems
8-98
-------
reasonable to expect that the uncertainty with regard to the low-dose
extrapolation would be somewhat reduced by considering the metabolized dose
as the effective dose, that the amount of chloroform metabolized better
reflects the dose-response relationship, since the toxicity and
carcinogenicity of chloroform is generally considered to be due to reactive
intermediate metabolites. To extrapolate from animals to humans, the amount
of chloroform metabolized per body surface area is assumed to be equivalent
(i.e., equally potent) among species. This assumption is by no means
supported by the empirical data. For chloroform however, there is some
evidence showing that, for a given dose in mg/kg by the oral route, the
amounts metabolized relative to body surface area are approximately equal
among species (see Figure 8-6). However, there are no actual observations
which support the proposition that the metabolized dose relative to body
surface area is equally effective in inducing tumors among different species.
An alternative approach for animal-to-human extrapolation would be to assume
that mg metabolized dose/kg/day is equivalent among species. If this
assumption is made, the potency factor qj_*, expressed in terms of
(mg/kg/day)~l, would be reduced. Brown et al. (1974) administered a 60 mg/kg
dose of chloroform orally to mice, rats, and squirrel monkeys and found the
corresponding percentages of dose expired unchanged to be respectively 6, 20,
and 78. On the basis of these data, Reitz et al. (1978) argued (implicitly)
that mg metabolites/kg should be used as an equivalent dose in the animal to
human extrapolation because mice are "more sensitive" to chloroform than rats
or humans since they metabolize a higher percentage of an administered dose
than than do larger species. It is true that the metabolism of chloroform on
a mg/kg body weight basis, has been found to be greater for the mouse than
the rat, greater for the rat than the monkey, and greater for these species
8-99
-------
than for humans. However, it is of note that the chloroform metabolized by
the rat, mouse, monkey, and human appear to be closely related to surface
area and metabolic rate. The mouse does not metabolize more chloroform than
the rat when the experimental data are expressed on a surface area basis.
The metabolism in humans is also proportional to body surface area. Thus,
only when the data are expressed on a mg/kg basis does the metabolism appear
to be mouse > rat > man.
Another aspect of uncertainty associated with the risk estimate is the
use of oral bioassay data to estimate the risk to humans by inhalation.
Although chloroform was a major general anesthetic agent for humans, in use
for a long period of time, the data that can be adequately used to compare
metabolism by oral and inhalation at low doses are not available. Therefore,
in this risk assessment, the uptake (total amount of metabolites)
corresponding to 1 iag/m3 of chloroform in air is calculated as the product of
inhaled concentration (1 ug/m3) times ventilation (20 m^/day).
8.5. RELATIVE CARCINOGENIC POTENCY
8.5.1. Derivation of Concept
One of the uses of the concept of unit risk is to compare the relative
potencies of carcinogens. To estimate relative potency on a per mole basis,
the slope of the maximum likelihood estimate at a fixed dose (mg/kg) is
multiplied by the molecular weight, and the resulting number is expressed in
terms of (mMol/kg/day)'1. This is called the relative potency index.
8.5.2. Potency Index
Figure 8-8 is a histogram of potency indices of 55 chemicals evaluated
by the CAG as suspect carcinogens. The actual data summarized by the
histogram are presented in Table 8-25. Where human data are available for a
compound, they have been used to calculate the potency index. Where no human
8-100
-------
20
18
16
14
4th QUARTILE . 3rd QUARTILE 2nd QUARTILE. 1st QUARTILE
1 X 10+1
4X 10+2
2 X 10+3
12
(J
10
o
LLJ
CC
u.
12
16
ill
-1
2345
LOG OF POTENCY INDEX
Figure 8-8. Histogram representing the frequency distribution of the potency indices of 55 suspect
carcinogens evaluated by the Carcinogen Assessment Group.
8-101
-------
TABLE 8-25. RELATIVE CARCINOGENIC POTENCIES AMONG 55 CHEMICALS EVALUATED BY THE CARCINOGEN ASSESSMENT GROUP
AS SUSPECT HUMAN CARCINOGENS
oo
i
Level
of evidence3
Compounds
Acryl onltMle
Aflatoxln Bj
AldMn
Ally! chloride
Arsenic
B[a]P
Benzene
Benzldene
Beryllium
1,3-Butadiene
Cadml urn
Carbon tetrachlorlde
Chlordane
CAS Number
107-13-1
1162-65-8
309-00-2
107-05-1
7440-38-2
50-32-8
71-43-2
92-87-5
7440-41-7
106-99-0
7440-43-9
56-23-5
57-74-9
Humans
L
L
I
S
I
S
S
L
I
L
I
I
Animals
S
S
L
I
S
S
S
S
S
S
S
L
Grouping
based on
IARC
criteria
2A
2A
3
1
2B
1
1
2A
2B
2A
2B
3
SI opeb
(mg/kg/day)-1
0.24(W)
2900
11.4
1.19x10-2
15(H)
11.5
2.9xlO-2(W)
234(W)
2.6(W)
1.0xlO-1(I)
6.1(W)
1.30x10-!
1.61
Molecul ar
weight
53.1
312.3
369.4
76.5
149.8
252.3
78
184.2
9
54.1
112.4
153.8
409.8
Potency
index0
lxlO+1
9X10"1"5
4xlO+3
9X10-1
2xlO+3
3xlO+3
2x10°
4xlO+4
2xlO+1
5x10°
7x10+2
2xlO+1
7x10+2
Order of
magnitude
(Iog10
index)
+1
+6
+4
0
+3
+3
0
+5
+1
+1
+3
+1
+3
-------
TABLE 8-25. (continued)
CO
o
CO
Compounds
Chlorinated ethanes
1,2-Dichloroethane
hexachloroethane
CAS Number
107-06-2
67-72-1
1,1,2,2-Tetrachloroethane 79-34-5
1,1,2-Trlchloroethane
Chloroform
Chromium VI
DDT
Dichlorobenzidine
1 ,1-D1 chl oroethyl ene
(Vinyl idene chloride)
Dichloromethane
(Methylene chloride)
Dieldrin
2,4-Dinitrotoluene
Diphenylhydrazine
Epichlorohydrin
B1s(2-chl oroethyl )ether
79-00-5
67-66-3
7440-47-3
50-29-3
91-94-1
75-35-4
75-09-2
60-57-1
121-14-2
122-66-7
106-89-8
111-44-4
Level
of evidence3
Humans Animals
I S
I L
I L
I L
I S
S S
I S
I S
I L
I S
I S
I S
I S
I S
I S
Grouping
based on
IARC
criteria
2B
3
3
3
2B
1
28
2B
3
28
28
2B
2B
2B
2B
Slope6
(mg/kg/day)-1
9.1xlO-2
1.42xlO-2
0.20
5.73x10-2
8.1x10-2
41(W)
0.34
1.69
1.16(1)
1 .4xlO"2
30.4
0.31
0.77
9.9xlO-3
1.14
Molecul ar
weight
98.9
236.7
167.9
133.4
119.4
100
354.5
253.1
97
84.9
380.9
182
180
92.5
143
Potency
index0
9x10°
3x10°
3xlO+1
8x10°
IxlO1
4xlO+3
1x10+2
4x10+2
1x10+2
1x10°
1x10+4
6X10+1
1x10+2
9x10-1
2x10+2
Order of
magnitude
(|og10
index)
+1
0
+ 1
+1
+1
+4
+2
+3
+2
0
+4
+2
+2
0
+2
(continued on the following page)
-------
TABLE 8-25. (continued)
Compounds
B1 s( chl oromethyl ) ether
Ethylene dlbromlde (EDB)
Ethylene oxide
Heptachlor
Hexachl orobenzene
^ Hexachl orobutadlene
1— •*
0
-P* Hexachl orocycl ohexane
technical grade
alpha Isomer
beta Isomer
gamma Isomer
Hexachl orod1benzod1ox1 n
Nickel refinery dust
Nickel subsulfide
Nltrosamlnes
Dimethyl nltrosamlne
Dlethylnltrosamine
D1 butyl nltrosamlne
N-nitrosopyrrol idlne
N-n1troso-N-ethylurea
CAS Number
542-88-1
106-93-4
75-21-8
76-44-8
118-74-1
87-68-3
319-84-6
319-85-7
58-89-9
34465-46-8
0120-35-722
62-75-9
55-18-5
924-16-3
930-55-2
759-73-9
Level
of evidence3
Humans
S
I
L
I
I
I
I
I
I
I
S
S
I
I
I
I
I
Animal s
S
S
S
S
S
L
S
L
L
S
S
S
S
S
S
S
s
Grouping
based on
I ARC
criteria
1
2B
2A
28
2B
3
2B
3
3
2B
1
1
2B
2B
2B
2B
2B
SI opeb
(mg/kg/day)-l
9300(1)
41
3.5x10-1(1)
3.37
1.67
7.75x10-2
4.75
11.12
1.84
1.33
6.2xlO+3
1.05(W)
2.1(W)
25.9(not by ql
43.5(not by ql
5.43
2.13
32.9
Molecul ar
weight
115
187.9
44.1
373.3
284.4
261
290.9
290.9
290.9
290:9
391
240.2
240.2
') 74.1
'') 102.1
' 158.2
100.2
117.1
Potency
1ndexc
1x10+6
8xlO+3
2X10+1
lxlO+3
5x10+2
2X10+1
lxlO+3
3xlO+3
5x10+2
4x10+2
2xlO+6
2.5x10+2
5.0x10+2
2x1 0+3
4xlO+3
9x10+2
_i_O
2x10+2
i 1
4xlO+3
Order of
magnitude
(]og10
index)
+6
+4
+1
+3
+3
+1
+3
+3
+3
+3
+6
+2
+3
+3
+4
+3
+2
+4
-------
TABLE 8-25. (continued)
CD
i
o
en
Compounds
N-n1troso-N-methyl urea
N-n1troso-diphenyl amine
PCBs
Phenol s
2,4,6-Trichlorophenol
CAS Number
684-93-5
86-30-6
1336-36-3
88-06-2
of
Level
evidence3
Humans Animals
I
I
I
I
S
S
S
s
Grouping
based on
IARC
criteria
2B
2B
2B
2B
Slopeb
(mg/kg/day)'1
302.6
4.92x10-3
4.34
1.99x10-2
Mol ecul ar
weight
103.1
198
324
197.4
Potency
index0
3xlO+4
1x10°
1x10+3
4x10°
Order of
magnitude
Oog10
index)
+4
0
+3
+1
Tetrachlorodibenzo-
p-dioxin (TCDD)
Tetrachl oroethyl ene
Toxaphene
Trichl oroethyl ene
Vinyl chloride
1746-01-6
127-18-4
8001-35-2
79-01-6
75-01-4
I
I
I
I
S
S
L
S
L/S
S
2B
3
2B
3/2B
1
1.56x10+5
5.1x10-2
1.13
1.1x10-2
1.75x10-2(1)
322
165.8
414
131.4
62.5
5x1 0+7
8x10°
5x10+2
1x10°
1x10°
+8
+1
+3
0
0
aS = Sufficient evidence; L = Limited evidence; I = Inadequate evidence.
^Animal slopes are 95% upper-bound slopes based on the linearized multistage model. They are calculated based on
animal oral studies, except for those indicated by I (animal inhalation), W (human occupational exposure), and H
(human drinking water exposure). Human slopes are point estimates based on the linear nonthreshold model. Not all
of the carcinogenic potencies presented in this table represent the same degree of certainty. All are subject to
change as new evidence becomes available. The slope value is an upper bound in the sense that the true value (which
is unknown) is not likely to exceed the upper bound and may be much lower, with a lower bound approaching zero.
Thus, the use of the slope estimate in risk evaluations requires an appreciation for the implication of the upper
bound concept as well as the "weight of evidence" for the likelihood that the substance is a human carcinogen.
cThe potency index is a rounded-off slope in (mmol/kg/day)"1 and is calculated by multiplying the slopes in
(mg/kg/day)-l by the molecular weight of the compound.
-------
data are available, animal oral studies are selected over animal inhalation
studies because most of the chemicals have been tested with animal oral
studies; this allows potency index comparisons by route.
The potency index for chloroform based on mouse hepatocellular
carcinomas in the NCI (1976) gavage study is 1 x 10.1. This is derived as
follows: the dose-response slope of 8.1 x 10-2 is multiplied by the
molecular weight of 119.38 to give a potency index of 1 x IQl. This places
chloroform among the least potent of the 55 suspect carcinogens, ranking in
the lowest quartile. The ranking of potency indices is subject to the
uncertainty of comparing potency estimates for different chemicals based on
different routes of exposure for a number of different species, using studies
whose quality varies widely. Furthermore, all potency indices are based on
estimates of low-dose risk using linear extrapolation from the observational
range. Thus, these indices are not valid for the comparison of potencies in
the experimental or observational range if linearity does not exist there.
8.6. SUMMARY
8.6.1. Qualitative
The carcinogenic potential of chloroform has been evaluated in several
animal species by experimental studies and in humans by epidemiologic survey.
Chronic animal studies have been conducted in eight strains of mice, two
strains of rats, and beagle dogs. In all of these studies, chloroform was
administered by the oral route and not by inhalation, an important route of
chloroform exposure for humans. However, a carcinogenic response from
chloroform exposure is not expected to be dependent upon the route of
assimilation into the body.
Chloroform in corn oil administered at an estimated maximally tolerated
dose (MTD) and one-half the MTD by gavage for 78 weeks produced a
8-106
-------
statistically significant increase in the incidence of hepatocellular
carcinomas in male and female B6C3F1 mice and renal epithelial tumors
(malignant and benign) in male Osborne-Mendel rats. A carcinogenic response
in female Osborne-Mendel rats treated with chloroform was not apparent in
this study. The use of more than two doses in these studies might have given
a more precise estimate of dose response.
Chloroform administered in the drinking water of male Osborne-Mendel
rats resulted in an increase in the incidence of renal tumors, thus
supporting the findings from an earlier study in which chloroform was
administered in corn oil by gavage. Chloroform administered in the drinking
water of female B6C3F1 mice, however, did not cause an increase in the
incidence of liver tumors in the mice as had been reported in previous
investigations. The lack of response in mice suggests that chloroform-
induced hepatocellular carcinomas in this strain of mice may be related to
the dosing regimen, absorption patterns, peak blood and target tissue levels.
The corn oil carrier has not been shown to induce an increase in the
incidence of liver tumors in mice.
A statistically significant increase in the incidence of renal tumors
(benign and malignant) was found in a study in male ICI mice treated with
chloroform in either toothpaste or arachis oil by gavage for 80 weeks.
Treatment with a gavage dose of chloroform in toothpaste for 80 weeks did not
produce a carcinogenic response in female ICI mice or in the male mice of the
CBA, C57BL, and CF/1 strains. Induction of malignant kidney tumors in male
ICI mice was greater when chloroform was administered, at the same dose, in
arachis oil rather than toothpaste.
A carcinogenic response was not observed in male and female Sprague-
Dawley rats given chloroform in toothpaste by gavage for 80 weeks, but early
8-107
-------
mortality was high in control and treatment groups. Gavage doses of
chloroform in toothpaste did not cause a carcinogenic response in male and
female beagle dogs treated for over seven years, although there was an
increased incidence of hepatic nodular hyperplasia. The results of
preliminary toxicity tests and the carcinogenicity studies suggest that doses
of chloroform in toothpaste given to mice, rats, and dogs in the
carcinogenicity studies approached those maximally tolerated by the test
species. However, daily chloroform doses given to mice and rats in
toothpaste or arachis oil were lower than those given in corn oil or drinking
water in other studies in which a positive carcinogenic response was
observed.
Hepatomas were found in NLC mice given chloroform in oil by gavage twice
weekly for an unspecified period of time, and in female strain A mice given
chloroform in olive oil by gavage once every 4 days for a total of 30 doses
at a level which produced liver necrosis. Small numbers of animals were
examined for pathology, the duration of these studies was either uncertain or
appeared to be less than the lifetime of the animals, and no control group of
NLC mice was discussed in the report. Although a carcinogenic effect of
chloroform was not evident in newborn (C57 * DBA2 Fl) mice given single or
multiple subcutaneous doses during the initial 8 days of life and observed
for their lifetimes, the dose levels used appeared well below a maximum
tolerated dose, and the period of treatment of the newborns was quite short
compared to lifetime treatment. Chloroform was ineffective at maximally
tolerated and lower doses in a pulmonary adenoma bioassay in strain A mice.
However, other chemicals that have shown carcinogenic activity in different
tests were also ineffective in this particular strain A mouse pulmonary
adenoma bioassay. Chloroform has been shown to promote growth and metastasis
8-108
-------
of marine tumors including the growth and spread of Lewis lung carcinoma,
Ehrlich ascites, and B16 melanoma cells in mice. The mechanisms by which
chloroform produced these effects are uncertain, and the relevance of these
endpoints to the evaluation of the carcinogenic potential of chloroform is
presently not clear, although in these studies chloroform promotes the growth
and spread of tumors at low exposure levels unlikely to cause observable
tissue damage. Chloroform in liquid solution did not induce transformation
of Syrian baby hamster kidney (BHK - 21/C1 13) cells in vitro at doses high
enough to produce toxicity. Additional testing of chloroform as a vapor
could have provided a comparison of cell transformation potential between
chloroform as a vapor and chloroform in a liquid solution.
There are no epidemiologic cancer studies dealing with chloroform per
se. However, chlorinated drinking water can contain significant amounts of
chloroform by virtue of chlorination of organic-laden raw water supplies.
There is a small, yet statistically significant increased risk of cancer of
the bladder, large intestine, and rectum associated with the presence of
chlorinated compounds in drinking water that is consistent across several
independent and diverse study populations. The risk estimates were
confounded by several factors: smoking, diet, air pollution, occupation, or
lifestyle. Bias can creep into these studies from differential survival
rates in various areas due to proximity to better medical care and treatment
facilities, higher socioeconomic status, and the possibility of migration of
cancer patients to medical care facilities in locales where chlorination of
water is used to a greater extent. Underestimates of risk may result from
failure to control for migration effects prior to diagnosis,
misclassification of cause of death, and use of chlorination as a surrogate
variable for chloroform, especially if few organic contaminants are present
8-109
-------
in the water. Some drinking water contaminants other than chloroform are
known to be carcinogenic, but they are generally found in much smaller
quantities, as compared with chloroform concentrations in organic-laden water
sources. The presence of these other carcinogenic substances and the
possibility of confounding makes it impossible to incriminate chloroform
solely as the cause of the excess cancer at the three tumor sites studied.
In summary, based upon several ecological and a few case control studies, a
small increased risk of cancer of the bladder, rectum, and large intestine
remains from water in which chloroform appears as a contaminant. However,
chloroform cannot be isolated as the sole cause of the excess cancer because
of the problems mentioned.
8.6.2 Quantitative
Five data sets are used to estimate the carcinogenic risk of chloroform.
The endpoints include liver tumors in female mice (NCI, 1976), liver tumors
in male mice (NCI, 1976), kidney tumors in male rats (NCI, 1976; Jorgenson et
al., 1985), and kidney tumors in male mice (Roe et a!., 1979). The unit risk
values at 1 mg/kg/day, calculated by the linearized multistage model on the
basis of these data sets, are comparable.
It is generally accepted that the carcinogenic activity of chloroform
resides in its highly reactive intermediate metabolites. Available data on
chloroform metabolism and pharmacokinetics pertinent to the conditions of the
carcinogenicity bioassays have been evaluated. Although this information is
relatively sparse, it is used in the extrapolation of the dose-carcinogenic
response relationships of laboratory animals to humans. After gavage or oral
administration to mice and rats, or oral administration to man, chloroform is
rapidly and virtually completely absorbed, with a portion of the dose
excreted via the lungs unchanged. In both animals and humans, pulmonary
8-110
-------
excretion of chloroform is dose dependent increasing with dose. The amount
of chloroform metabolized to reactive metabolites is the gavage or oral dose
minus the percentage of dose excreted unchanged. Metabolism of chloroform
approaches saturation at the dose levels of the bioassays in which the
chloroform was administered as a bolus. A greater proportional metabolism is
presumed for the smaller amounts of chloroform encountered in environmental
exposure. Different carriers, such as water and corn oil, used as vehicles
for chloroform in the bioassays, result in different absorption patterns.
Once absorbed into the body by any route, chloroform distributes throughout
body tissues, concentrates in lipid membranes, and accumulates in adipose
tissue. Chloroform has a relatively long half-time of residence in the body
when compared to similar chlorinated hydrocarbons. Since daily gavage doses
are not expected to be completely cleared from the body of the experimental
animal during a 24-hr dosing interval, the test animal may experience the
constant presence of chloroform and chloroform metabolites. In mice,
covalent binding of chloroform metabolites in both liver and kidney has been
demonstrated to be proportional to the dose, which is noteworthy since the
liver and kidney are the target organs for carcinogenicity. There is no
evidence to suggest any qualitative difference in the metabolic pathways or
metabolite profiles of mice, rats, and humans for chloroform. Also, there is
an experimental basis for determining the amount of chloroform metabolized in
different species, including man. The amount metabolized is proportional to
the surface area of each species.
The magnitude of the metabolic conversion of chloroform is judged to be
important to its carcinogenic potential. However, the present knowledge of
chloroform metabolism and related acute cellular toxicity suggests several
general cellular processes/mechanisms that may lead to a carcinogenic effect.
8-111
-------
Each of these mechanisms is supported by experimental data. In the absence
of definitive evidence solely supporting one of the likely processes, the
quantitative risk assessment is based on the assumption of a nonthreshold
mechanism, and consequently a mathematical model consistent with this
assumption is used. Results using other risk extrapolation models are
presented in the appendix.
Although the mathematical risk extrapolation model chosen is
conservative based upon a public-health point of view, the correction used in
the calculation of a human equivalent dose is scientifically conservative and
could lead to an overestimate of the amount of chloroform metabolized in the
test animals, thus giving a lower calculated risk value. In addition,
experimental data, which include those for covalent binding in human tissues,
suggest that humans may have a greater-than-expected capacity to metabolize
chloroform when compared to rodents, again indicating the possibility of a
higher risk for humans than estimated in the assessment.
The geometric mean, 8.1 * 10-2, Of the slope estimates calculated from
chloroform-induced liver tumors in male and female mice treated by gavage is
the value used to compare the relative potency of chloroform to other
carcinogens and to calculate the unit risks for drinking water and air. The
upper-bound estimate of the cancer risk based on gavage exposures to 1 pg/m3
of chloroform in air is 2.3 * 10-5. The upper-bound estimate of cancer risk
due to 1 pg/liter in water is 2.3 * 10-6. The latter estimate appears
consistent with the limited epidemiologic data available for humans.
8.7 Conclusions
Evidence that chloroform has carcinogenic activity is based on increased
incidences of hepatocellular carcinomas in male and female B6C3F1 mice, renal
epithelial tumors in male Osborne-Mendel rats, kidney tumors in male ICI
8-112
-------
mice, and hepatomas in NIC and female strain A mice. As stated elsewhere in
this document, no definite conclusions can be reached concerning the
mutagenicity of chloroform based on present evidence. lr\ vitro tests for
mutagenicity using bacterial and mammalian cells in culture (with and without
metabolic activation) have been uniformly negative; however, several j_n vivo
studies have been reported that show a positive mutagenic response.
The toxicity of chloroform in liver and kidney is considered to occur
through covalent binding of reactive intermediate metabolites, such as
phosgene, with cellular macromolecules. The evidence indicates that reactive
metabolites of chloroform can react extensively with proteins and lipids, and
minimally with nucleic acids. The intensity of metabolite binding and organ
localization parallels the acute liver and kidney cellular toxicity of
chloroform observed in experimental animals. Both the amount of binding and
the degree of toxicity appear to be dependent on animal species and genetic
strain, as well as on sex and age. Irreversible binding of chloroform
metabolites to cellular macromolecules supports several theoretical concepts
of the mechanism(s) for its carcinogenicity.
while no epidemiological studies have evaluated chloroform by itself,
several studies have been made of populations with chlorinated drinking
water, in which chloroform is the predominant chlorinated hydrocarbon
compound. Small increases in rectal, bladder, and colon cancer were
consistently observed by several case-control and ecological studies, several
of which are statistically significant. Because other possible carcinogens
were present along with chloroform, it is impossible to identify chloroform
as the sole carcinogenic agent. Therefore, the epidemiologic evidence for
chloroform's carcinogenicity must be termed inadequate.
8-113
-------
Chloroform gavage studies which show a statistically significant
increase of hepatocellular carcinomas in mice provide a basis to estimate
upper-bound incremental lifetime cancer risks due to chloroform exposure.
The risk value is useful for estimating the possible magnitude of the public-
health impact. The upper-bound incremental cancer risk is 8.1 x 10-2
per mg/kg/day. The CAG potency index for chloroform (defined as the slope
times the molecular weight) is 1 * 1C)1, ranking it in the lowest quartile of
55 chemicals that the CAG has evaluated as suspect carcinogens. The upper-
bound estimate of the incremental cancer risk due to ingesting 1 ^g/L of
chloroform in drinking water is 2.3 * 10-6. The upper-bound estimate of the
incremental cancer risk due to inhaling 1 pg/m3 of chloroform in air based
upon positive gavage carcinogenicity studies is 2.3 x 10-5. The upper-bound
nature of these estimates is such that the true risk is not likely to exceed
this value and may be lower.
Based on EPA's proposed Carcinogen Risk Assessment Guidelines,
chloroform is classified as having sufficient animal evidence and inadequate
epidemiologic evidence. The overall weight-of-evidence classification is
Group B2, meaning that chloroform is probably carcinogenic in humans.
Applying the International Agency for Research on Cancer (IARC) criteria, the
level of animal evidence for carcinogenicity is sufficient, and the overall
IARC classification is Group 2B, meaning that chloroform should be considered
to be a probable human carcinogen.
8-114
-------
8.8 REFERENCES FOR CHAPTER 8
Agustin, J.S.; Lim-Sylianco, C.Y. (1978) Mutagenic and clastogenic effects of
chloroform. Bull. Phil. Blochem. Soc. 1:17-23.
Alavanja, M.; Goldstein, I; Susser, M. (1978) A case-control study of
gastrointestinal and urinary tract cancer mortality and drinking water
chlorination. In: Water chlorination: environmental impact and health
effects, v. 2. Jolley, R.L.; Gorchev, H.; Hamilton, D.H., Jr., eds. Ann
Arbor, MI: Ann Arbor Science Publishers; pp. 395-409.
Ames, B.N.; McCann, J.; Yamasaki, E. (1975) Method for detecting
carcinogens and mutagens with the Salmonella/mammalian-microsome
mutagenicity test. Mutat. Res. 31:347-364.
Armitage, P.; Doll, R. (1954) The age distribution of cancer and a
multi-stage theory of carcinogenesis. Br. J. Cancer 8:1-12.
Brenniman, G.R.; Vasilomanolakis-Lagos, J.; Amsel, J.; Namekata, T.;
Wolff, A.M. (1978) Case-control study of cancer deaths in Illinois
communities served by chlorinated or nonchlorinated water. In: Water
chlorination: environmental impact and health effects. Jolley, R.J.;
Gorchen, H.; Hamilton, H., Sr., eds. Ann Arbor, MI: Ann Arbor Science
Publishers; pp. 1043-1057
Brown, D.M.; Langley, P.F.; Smith, D.; Taylor, D.C. (1974) Metabolism
of chloroform. I. The metabolism of 14C-chloroform by different species.
Xenobiotica 4:151-163.
Buncher, C.R. (1975) Cincinnati drinking water: an epidemiologic study of
cancer rates. Cincinnati, OH: University of Cincinnati Medical Center.
Callen, D.F.; Wolf, C.R.; Philpot, R.M. (1980) Cytochrome P-450-mediated
genetic activity and cytotoxicity of seven halogenated aliphatic
hydrocarbons in Saccharomyces cerevisiae. Mutat. Res. 77:55-63.
Cantor, K.P.; Hoover, R.; Mason, T.J.; McCabe, L.J. (1978) Association of
cancer mortality with halomethanes in drinking water. J. Natl. Cancer
Inst. 61(4): 979-985.
Capel, I.D.; Dorrell, H.M.; Jenner, M.; Pinnock, M.H.; and Williams D.C.
(1979) The effect of chloroform ingestion on the growth of some murine
tumors. Eur. J. Cancer 15:1485-1490.
Cox, C.B. (1972) Regression model and life tables. J. Roy. Stat. Soc. B 34:
187-220.
Crump, K.S. (1979) Dose-response problems in carcinogenesis. Biometrics 35:
157-167.
Crump, K.S.; Guess, H.A. (1982) Drinking water and cancer: review of
recent epidemiological findings and assessment of risks. Ann. Rev.
Public Health 3:339-357.
8-115
-------
Crump, K.S.; Guess, H.A.; Deal, L.L. (1977) Confidence intervals and test
of hypotheses concerning dose-response relations inferred from animal
carcinogenicity data. Biometrics 33:437-451.
Culliford, D.; Hewitt, H.B. (1957) The influence of sex hormone status on the
susceptibility of mice to chloroform-induced necrosis of renal tubules.
J. Endocrine!. 14:318-393.
Davidson, I.W.F. (1984) Biological basis for extrapolation across mammalian
species. EPA Issue Paper.
Davidson, I.W.F.; Sumner, D.D.; Parker, J.C. (1982) Chloroform: a review of
its metabolism, teratogenic, mutagenic, and carcinogenic potential.
Drug Chem. Toxicol. 5:1-87.
De Rouen, T.A.; Diem, J.E. (1975) The New Orleans drinking water controversy:
a statistical perspective. Am. J. Public Health 65:1060-1062.
Diaz Gomez, M.I.; Castro, J.A. (1980). Covalent binding of chloroform
metabolites to nuclear proteins-no evidence for binding to nucleic
acids. Cancer Lett. 9:213-218
Docks, E.L.; Krishna, G. (1976) The role of glutathione on chloroform-
induced hepatoxicity. Exp. Mol. Pathol. 24:13-22.
Doll, R. (1971) Weibull distribution of cancer: implications for models of
carcinogenesis. J. Roy Stat. Soc. A 13:133-166.
Ekstrom, T.; Hogberg, J. (1980) Chloroform induced glutathione depletion
and toxicity in freshly isolated hepatocytes. Biochem. Pharmacol.
29:3059-3064.
Eschenbrenner, A.B.; Miller, E. (1945) Induction of hepatomas in mice by
repeated oral administration of chloroform with observations on sex
differences. J. Natl. Cancer Inst. 5:251-255.
Fleiss, J.L. (1979) Confidence intervals for odds ratio in case-control
studies: the state of the art. J. Chronic Dis. 32:69-77.
Fry, J.; Taylor, T.; Hathaway, D.F. (1972) Pulmonary elimination of
chloroform and its metabolites in man. Arch. Int. Pharmacodyn.
196:98-111
Gottlieb, M.S.; Carr, J.K.; Morris, D.T. (1981) Cancer and drinking water in
Louisiana: colon and rectum. Int. Epidemiol. 10(2):117-125.
Harris, R.H. (1974) Implications of cancer-causing substances in Mississippi
River water. Washington, DC: Environmental Defense Fund,
Harris, R.H.; Page, T.; Reiches, N.A. (1977) Carcinogenic hazards of organic
chemicals in drinking water. In: Origins of Human Cancer; pp. 309-330.
8-116
-------
Heywood, R.; Sortwell, R.J.; Noel, P.R.B.; Street, A.E.; Prentice, D.E.;
Roe, F.J.C.; Wadsworth, P.P.; Worden, A.M.; Van Abbe, N.J. (1979) Safety
evaluation of toothpaste containing chloroform. III. Long-term study in
beagle dogs. Environ. Pathol. Toxicol. 2:835-851.
Hill, R.N.; Clemson, T.L.; Liu, O.K.; Vesell, E.S.; Johnson, W.D. (1975)
Genetic control of chloroform toxicity in mice. Science 190:159-160.
Hogan, M.D.; Chi, P.; Hoel, D.G.; Mitchell, T.J. (1979) Association between
chloroform levels in finished drinking water supplies and various site-
specific cancer mortality rates. J. Environ. Pathol. Toxicol. 2:873-887.
Hoover, R.N.; McKay, F.W.; Fraumeni, J.F. (1976) Fluoridated drinking water
and the occurrence of cancer. J. Natl. Cancer Inst. 57:757-768.
Howe, R.B. (1983) GLOBAL83: an experimental program developed for the U.S.
Environmental Protection Agency as an update to GLOBAL82: a computer
program to extrapolate quantal animal toxicity data to low doses (May
1982). Ruston, LA: K.S. Crump and Co., Inc. Unpublished.
Howe, R.B.; Crump, K.S. (1982) A computer program to extrapolate quantal
animal toxicity data to low dose. OSHA Contract No. 41USC252C3.
Ilett, K.F.; Reid, W.D.; Sipes, I.G.; Krishna, G. (1973) Chloroform toxicity
in mice: correlation of renal and hepatic necrosis with covalent binding
of metabolites to tissue macromolecules. Exp. Mol. Pathol. 19:215-229.
International Regulatory Liaison Group. (1979) Scientific bases for
identifying potential carcinogens and estimating their risks. Feb. 6,
1979.
Jorgenson, T.A.; Meierhenry. E.F.; Rushbrook, C.J.; Bull, R.J.; Robinson, M.
(1985) Carcinogenicity of chloroform in drinking water to male Osborne
Mendel rats and female B6C3F1 mice. Fund. Appl. Toxicol. 5:760-769.
Kovar, J.; Krewski, D. (1981) RiskSl: a computer program for low-dose
extrapolation of quantal response toxicity data. Canada: Department of
Health and Welfare.
Kruse, C.W. (1977) Chlorination of public water supplies and cancer -
Washington County, Maryland experience. A final report, Johns Hopkins
University, School of Hygiene and Public Health, to the Office of
Research and Development, Health Effects Research Laboratory,
Cincinnati, OH. Unpublished draft.
Kuzma, R.J.; Kuzma, C.M.; Buncher, C.R. (1977) Ohio drinking water source and
cancer rates. Am. J. Public Health 67:725-729.
Lehman, K.B.; Hasegawa, 0. (1910) Studies of the asorption of chlorinated
hydrocarbons in animals and humans. Arch. Hyg. 72:327-342 (German).
Linde, H.W.; Mesnick, P.S. (1980) Cause of death of anesthesiologists from
the chloroform era. Govt. Reports Announcements and Index. 80:1148
8-117
-------
Mah, R.A.; Spivey, G.H.; Sloss, E. (1977) Cancer and chlorinated drinking
water. Final report on EPA Contract No. CA-6-99-3349-J to Health Effects
Research Laboratory, Office of Research and Development, U.S.
Environmental Protection Agency, Cincinnati, OH.
Mantel, N.; Schneiderman, M.A. (1975) Estimating safe levels, a hazardous
undertaking. Cancer Res. 35:1379-1386.
Mason, T.J.; McKay, F.W. (1974) U.S. cancer mortality by county: 1950-1969.
Washington, DC: U.S. Government Printing Office.
McCabe, L.J. (1975) Association between trihalomethanes in drinking water
(NORS data) and mortality. Cincinnati, OH. Health Effects Research
Laboratory, U.S. Environmental Protection Agency.
Moore, D.H.; Chasseaud, L.F.; Majeed, S.K.; Prentice, D.E.; Roe, F.J.C.
(1982) The effect of dose and vehicle on early tissue damage and
regenerative activity after chloroform administation to mice. Fd. Chem.
Toxicol. 20:951-954.
Morimoto, K.; Koizumi, A. (1983). Trihalomethanes induce sister chromatid
exchanges in human lymphocytes in vitro and mouse bone marrow cells i_n
vivo. Environ. Res. 32:72-79.
National Academy of Sciences. (1978) Epidemiological studies of cancer
frequency and certain organic constituents of drinking water. A review
of recent literature published and unpublished. Washington, DC.
National Cancer Institute (NCI). (1976) Report on carcinogenesis bioassay of
chloroform. PB-264018. Springfield, VA: National Technical Information
Service.
National Institute for Occupational Safety and Health (NIOSH). (1974)
Criteria for a recommended standard: Occupational exposure to
chloroform. DHEW Publication No. 75-114.
Page, T.; Harris, R.H.; Epstein, S.S. (1976) Drinking water and cancer
mortality in Louisiana. Science 193:55-57.
Palmer, A.K.; Street, A.E.; Roe, F.J.C.; Worden, A.N.; Van Abbe, N.J. (1979)
Safety evaluation of toothpaste containing chloroform. II. Long-term
studies in rats. J. Environ. Pathol. Toxicol. 2:821-833.
Parker, J.C.; Davidson, I.W.'F. (1984) Extrapolation of the carcinogenic
response. Abstract of paper presented at the annual meeting of the
Society of Risk Analysis: Uncertainty in Riskj Assessment, Risk
Management, and Decision Making. Sept. 30-Oct. 4, 1984. Knoxville, TN.
Pereira, M. (1984) The genotoxic/epigenetic distinction: Relevance to cancer
policy. Env.Res. 34:175-191.
8-118
-------
Purchase, I.F.H.; Longstaff, E.; Ashby, J.A.;Styles, D.; Lefevre, 0.;
Lefevre, P.A.; Westwood, F.R. (1978) An evaluation of 6 short-term tests
for detecting organic chemical carcinogens and recommendations for their
use. Br. J. Cancer. 37:873-959.
Rafferty, P. (1979) Drinking water quality and cancer mortality patterns.
Master's Thesis. Chapel Hill, NC: University of North Carolina;
unpublished.
Reddy, B.S.; Cohen, L.A.; McCoy, G.D.; Hill, P.; Weisburger, J.H.;
Wynder, E.L. (1980) Nutrition and its relationship to cancer. (Advances
in Cancer Research, v. 32) Valhalla, NY: Naylor Dana Institute for
Disease Prevention, American Health Foundation.
Reitz, R.H.; Fox, T.R.; Quast, J.F. (1982) Mechanistic considerations for
carcinogenic risk estimation: chloroform. Environ. Health Persp.
46:163-168.
Reitz, R.H.; Gehring, P.J.; Park, C.N. (1978) Carcinogenic risk estimation
for chloroform: an alternative to EPA's procedure. Food Cosmet.
Toxicol. 16:511-514.
Reitz, R.H; Quast, J.F.; Scott, W.T.; Watanabe, P.G.; Gehring, P.J. (1980)
Pharmacokinetics and macromolecular effects of chloroform in rats and
mice: implications for carcinogenic risk estimation. Water Chlorinat.
Environ. Impact Health Eff. 3:893-993.
Reuber, M.D. (1979) Carcinogenicity of chloroform. Environ. Health Perspect.
31:171-182.
Roe, F.J.C.; Carter, R.L.; Mitchley, B.C.V. (1968) Test of chloroform and
8-hydroxyquinoline for carcinogenicity using newborn mice. Br. Emp.
Cancer Campgn. 46:13.
Roe, F.J.C.; Palmer, A.K.; Worden, A.M.; Van Abbe, N.J. (1979) Safety
evaluation of toothpaste containing chloroform. I. Long-term studies in
mice. J. Environ. Pathol. Toxicol. 2:799-819.
Rudali, G. (1967) Oncogenic activity of some halogenated hydrocarbons used in
therapeutics. UICC Monogr. Ser. 7:138-143.
Salg, J. (1977) Cancer mortality and drinking water in 346 counties of the
Ohio River Valley Basin. Ph.D. Thesis. Department of Epidemiology,
University of North Carolina. Also final report, EPA Contract No. 5-03-
4528-J to Health Effects Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Cincinnati, OH.
Struba, R.J. (1979) Cancer and drinking water quality. Ph.D. Thesis.
University of North Carolina, Chapel Hill, NC, 156 pp. Available from:
Ann Arbor, MI: University Microfilms International, Publication
No.8022514.
Styles, J.A. (1977) A method for detecting carcinogenic organic chemicals
using mammalian cells in culture. Br. J. Cancer. 36:558-563.
8-119
-------
Styles, J.A. (1979) Cell transformation assays. In: Pagett, 6.E. ed. Topics
in Toxicology: Mutagens in sub-mammalian systems: Status and
significance. Baltimore, MD: University Park Press, pp. 147-164.
Tarone, R.E.; Gart, J.J. (1975) Review of "The implications of cancer-
causing substances in Mississippi River water" by R.H. Harris. Bethesda,
MD: Mathematical Statistics and Applied Mathematics Section, National
Cancer Institute.
Taylor, D.C.;Brown, D.M.; Kuble, R.; Langley, P.F. (1974) Metabolism of
chloroform. II. A sex difference in the metabolism of 14C-chloroform in
mice. Xenobiotica. 4:165-174.
Theiss, J.C.; Stoner, G.D.; Shimkin, N.B.; Weisburger, E.K. (1977) Test for
carcinogenicity of organic contaminants of United States drinking waters
by pulmonary tumor response in strain A mice. Cancer Res. 37:2717-2720.
Tuthill, R.W.; Moore, G.S.; Polakoff, D.W. (1979) A statistical model for
predicting chloroform levels in chlorinated surface water supplies.
J. Am. Water Works Assoc. 71:37-39.
Uehleke, H; Werner, T; Greim, H.; Kramer,M. (1977). Metabolic activation of
haloalkanes and tests in vitro for mutgagenicity. Xenobiotica 7:393-400.
Uehleke, H.; Werner, T. (1975, December) A comparative study on the
irreversible binding of labeled halothane, trichlorofluoromethane,
chloroform, and carbon tetrachloride to hepatic protein and lipids i_n
vitro and i_n vivo. Arch. Toxicol. 34:289-303.
U.S. Environmental Protection Agency. (1975) Preliminary assessment of
suspected carcinogens in drinking water: Report to Congress.
U.S. Environmental Protection Agency. (1979) National interim primary
drinking water regulations; control of trihalomethanes in drinking
water; final rule.
U.S. Environmental Protection Agency (1984, November) Proposed Guidelines for
Carcinogen Risk Assessment. Federal Register 49(227):46294-46301.
Vasilenko, P.; Magno, L. (1975) Factors relating to the incidence of cancer
mortality in New Jersey. Princeton, NJ: Princeton University;
unpublished.
Weisburger, E.K. (1977) Carcinogenicity studies on halogenated hydrocarbons.
Environ. Health Perspect. 21: 7-16.
Weisburger, J.H.; Williams, G.M. (1980). Chemical Carcinogens. In
"Toxicology: The Basic Science of Poisons." J. Doull, C.D. Klassen and
M.O. Amdur, Eds. 2nd Ed. pp84-138. MacMillan, New York.
Weisburger, J.H.; Williams, G.M. (1981). Carciniogen testing. Current
problems and new approaches. Science. 214-401-407.
8-120
-------
Whittemore, A. (1978) Quantitative theories of carcinogenesis. In: Advances
in cancer research, v. 27. New York NY: Academic Press, Inc., pp. 55-58.
Whittemore, A.; Keller, B. (1978) Quantitative theory of carcinogenesis.
Soc. Ind. Appl. Math. Rev. 20:1-30.
Wilkins, J.R. (1978) Chlorination of public water supplies and cancer in
Washington County, Maryland. Ph.D. Thesis. Department of Environmental
Health Sciences, Johns Hopkins University School of Hygiene and Public
Health, 1979. Also final report with C.W. Kruse on EPA Contract No.
R805198-01-0 to Health Effects Research Laboratory, Office of Research
and Development, U.S. Environmental Protection Agency, Cincinnati, OH.
Wilkins, J.R.; Reiches, N.A.; Kruse, C.W. (1979) Organic chemical
contaminants in drinking water and cancer. Am. J. Epidemiol. 110:420-
448.
Withey, J.R.; Collins, B.T.; Collins, P.G. (1982) Effect of vehicle on the
pharmacokinetics and uptake of four halogenated hydrocarbons from the
gastrointestinal tract of the rat. Pre-print paper submitted to J. Appl,
Toxicol.
Young, T.B.; Kanarek, M.S.; Tslatis, A.A. (1981) Epidemiologic study of
drinking water chlorination and Wisconsin female cancer mortality.
J. Natl. Cancer Inst. 67(6): 1191-1198.
8-121
-------
APPENDIX 8A
COMPARISON AMONG DIFFERENT EXTRAPOLATION MODELS
Four models used for low-dose extrapolation, assuming the independent
background, are:
Multistage: P(d) = 1 - exp [-(q^d + ... + qkdk)]
where qi are non-negative parameters
Probit:
n/j\ r A + Bln.(d) c/ , ,
r(d)= j _ ^ f(x)dx
where f(.) is the standard normal probability density function
Weibull: P(d) - 1 - exp [-bdk]
where b and k are non-negative parameters
One-hit: P(d) = 1 - exp [-bd]
where b is a non-negative parameter.
The maximum likelihood estimates (MLE) of the parameters in the multistage
and one-hit models are calculated by means of the program GLOBAL83, which was
developed by Howe (1983). The MLE estimates of the parameters in the probit
and Weibull models are calculated by means of the program RISK81, which was
developed by Kovar and Krewski (1981).
Table A-l presents the MLE of parameters in each of the four models that
are applicable to a data set.
8A-1
-------
TABLE A-l. MAXIMUM LIKEHOOD ESTIMATE OF THE PARAMETERS FOR EACH OF THE FOUR EXTRAPOLATION MODELS,
BASED ON DIFFERENT DATA BASES
CD
I
no
Data base
Liver tumors in female
mice (NCI, 1976)
Liver tumors in male
mice (NCI, 1976)
Kidney tumors in male
rats (NCI, 1976)
Kidney tumors in male
mice (Roe et al., 1979)
Kidney tumors in male
rats (Jorgenson et al . ,
1985)
Multistage
qi = 1.58 x
q2 = 0
(q^* = 2.0 x
qi = 0
q^ = 1.57 x
(q^* = 3.3 x
qj_ = 4.24 x
qp = 1.11 x
(qj* = 2.35 x
qi = 6.14 x
(q:* = 1.0 x
qi = 6.05 x
qj = 1.57 x
q3 = 0
q4 = o
qj* = 4.41 x
10-1
10-l)a
10-2
10-2)
10-3
10-3
10-2)
10-2
10-1)
10-4
10-4
10-3
Probit
A = -1.84
B = 1.17
A = -6.83
B = 3.49
A = -3.36
B = 1.01
NA
A - -3.74
B = 0.78
Weibull
b = 2.04 x 10-1
k = 0.90
b = 1.07 x1 10-3
k = 3.23
b = 2.97 x 10-3
k - 1.72
NA
b = 4.84 x 10-4
k = 1.70
One-hit
b = 1.58 x 10-1
b = 1.31 x 10-1
b = 1.70 x 10-2
b = 6.14 X 10-2
b = 3.42 x 10-3
aqj* is the 95% upper-bound confidence limit of the linear parameter in the multistage model.
NA = not applicable. The models are not applicable since there is only one dose group.
------- |