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.

-------