Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Toxicity of Selected Contaminants 79 TRICHLOROETHYLENE ethene, tricolor CAS No. 79~1-6 ClHC = CC12 Trichloroethylene was evaluated in the first and third volumes of Drinking Water arid Health (NationalResearch Council, 1977,pp.777-781;1980, pp. 155-166~. The following material, which became available after the 1980 report was prepared, updates and, in some instances, reevaluates the information contained in the previous reviews. Also included are some ref- erences that were not assessed in the earlier reports. Trichloroethylene has also been the subject of a monograph by the International Agency for Re- search on Cancer (1979a), in which toxicity data on animals and humans are evaluated. METAB O LI SM The continuing high level of interest in the metabolic disposition of trichlo- roethylene has undoubtedly been stimulated by the increasing accumula- tion of evidence that its metabolic products are the cause of functional im- pairment and tissue damage in various organs. Henschler and Bonse (1978) postulated that the first step in the metabolism of this compound and other chlorinated ethylenes is the formation of an epoxide, a highly reactive intermediate with alkylating properties. Furthermore, their stud- ies of synthetic epoxides indicate that the ethylenes with unsymmetrical chlorine substitution (a series to which Trichloroethylene belongs) are more highly electrophilic than the analogs with symmetrically positioned chlo- rine atoms. In the rat, Trichloroethylene induces its own metabolism by hepatic microsomes, an effect that increases the toxicity of the agent and, paradoxically, results in the destruction of the enzymes that catalyze the metabolism (Pessayre et al., 1979b, 1980~. Trichloroethylene appears to be activated by a phenobarbital-induced form of cytochrome P450 to an active species that can bind and inactivate the cytochrome (Costa et al., 1980~. In both rats and mice, the induction of hepatic microsomal enzymes led to a greater degree of Trichloroethylene bioactivation and increasing quantities of trichloroethylene-related mate- rials covalently bound not only to hemes and cytochromes but also to vari- ous macromolecules, proteins, lipids, and DNA (Pessayre et al., 1980; Sipes and Gandolfi, 1980; Stott et al., 1982~. These studies extend and confirm previously postulated mechanisms of Trichloroethylene toxicity.
80 DRINKING WATER AND HEALTH For a more complete review of the literature and discussion of the data, the reader should consult the report of Stott et al. (1982~. In rodents there is a good correlation between the rate and extent of trichloroethylene metabolism and the degree of observed hepatic toxicity. The importance of dose in the interpretation of the biochemical effects of trichloroethylene is underscored by the work of Stott et al. (1982), which showed that little alkylation of DNA by trichloroethylene metabolites oc- curs in the absence of cytotoxicity. In the same study, the investigators demonstrated that B6C3F~ mice are capable of metabolizing more trichlo- roethylene to a toxic intermediate than are Osborne-Mendel rats. Extend- ing this finding one step further, the authors cited the kinetic work of Monster et al. (1976) and Filser and Bolt (1979) as a basis for concluding that humans metabolize approximately 20 times less trichloroethylene than rats, on a weight basis. Since the metabolism of trichloroethylene is an important factor in its toxicity, one must view toxicity data cautiously when making interspecies comparisons. Dichloroacetic acid has been shown to be a urinary metabolite of trichlo- roethylene in the mouse (Hathway, 19801. If confirmed, this finding would help to explain the relatively high toxicity of trichloroethylene in this spe- cies. A highly reactive intermediate almost certainly precedes the forma- tion of the dichloroacetic acid. Studies by Nomiyama and Nomiyama (1979a,b) emphasize the differences in the rate and disposition of trichlo- roethylene in humans, rats, and rabbits. Although the disparity among the doses given introduces some uncertainty in the interpretation of the data, it is clear that the metabolism of trichloroethylene proceeds at a much slower rate in humans than in the rat or rabbit. In addition, a much greater pro- portion of trichloroethylene is excreted as trichloroacetic acid in the urine of humans than in the urine of rats or rabbits. Trichloroethanol is a major metabolite in rats, rabbits, and humans. Using an in vitro rat liver micro- somal system, Leighty and Fentiman (1981) demonstrated that trichloro- ethanol is conjugated to palmitic acid. Whether this conjugation occurs in viva and whether it occurs to an extent that would be important in the disposition of trichloroethylene remains to be established. HEALTH ASPECTS Observations in Humans Konietzko and Reill (1980) monitored 10 serum enzymes in 20 male volun- teers for signs of toxicity following 4-hour exposures to 95 ppm (511 ma/ m3) doses of trichloroethylene. The dose was calculated to be equivalent to an 8-hour maximum allowable concentration (MAC) 50-ppm (2S8 mg/m3)
Toxicity of Selected Contaminants 81 exposure. No effects on serum enzyme activities were observed at the end of the exposure period or at either of two points during the first 24 hours. At the end of the 4-hour exposure, the following mean blood concentra- tions were measured: trichloroethylene, 5.8 ~g/ml; trichloroethanol, 3.4 g/ml; and trichloroacetic acid, 7.2 ~g/ml. At 24 hours, trichloroacetic acid concentrations had risen to 15.3 ~g/ml, whereas the concentrations of unchanged trichloroethylene had dropped to 0.18 ~g/ml. The results of this study lend further support to the conclusion that trichloroethylene is not toxic in humans at exposure levels > 100 ppm (535 mg/m3~. Observations in Other Species Acute Effects The committee found no new reports that would amplify or extend the data cited in the previous reviews. Chronic Effects Studies in which mice were given trichloroethylene by gavage 5 days/week for 3 weeks indicated that 250 mg/kg caused minimal hepatotoxic effects (Stott et al., 1982~. The effects of this treatment were evident only upon histopathological examination of the liver. Daily doses of 500 mg/kg or greater caused an increase in liver weight and centrilobu- lar hypertrophy of the liver. In contrast, a 3-week exposure of rats to tri- chloroethylene doses of 1,000 mg/kg/day resulted in no liver damage. Mutagenicity Trichloroethylene was nonmutagenic in the Ames Sal- monella assay when tested with TA100 in a 10-liter desiccator. Exposure levels were as high as 20~o in air (v/v) for up to 16 hours. The assay was performed in the presence and absence of a phenobarbital-induced liver S9 fraction from male mice (Bartsch et al., 1979~. Chloral hydrate a metab- olite of trichloroethylene was found to be mutagenic in strain TA100 in the Salmonella standard plate incorporation assay in doses ranging from 0.5 to 10.0 mg/plate. The mutagenic activity was enhanced in the presence of rat liver S9 mix. Negative results were obtained with the frameshift mu- tant TA98 (Waskell, 19781. When tested in Saccharomyces cerevisiae D7, trichloroethylene induced both point mutations and mitotic gene conver- sion. These effects were observed only in the presence of mouse liver S10 at doses ranging from 10 to 40 mM concentrations. A mutagenic response was also obtained in the host-mediated assay with mice; both point muta- tions and gene conversion were observed in S. cerevisiae D7, and gene con- version in S. cerevisiae D4, when recovered from the li~rer and kidneys after both acute and subacute dosing. S. cerevisiae recovered from lung tissue showed little, if any, genetic effects (Bronzetti et al., 1980~. In another study with S. cerevisiae D7, no exogenous mammalian meta-
82 DRINKING WATER AND HEALTH boric activation was necessary to elicit recombinations, mitotic gene con- versions, and gene reversions after incubation for 1 hour at 37°C at a con- centration of 15 mM. The highest dose used, a 22 mM concentration. was toxic to 99.7% of the cells. The investigators found that the yeast cyto- chrome P450 system mediated the metabolic conversion (Caller et al., 1980~. No mutagenic effects were observed in a dominant lethal assay with NMRI Han/BOA male mice exposed for 24 hours to 50-, 202-, and 450- ppm (268, 1,086, and 2,420 mg/m3) concentrations of trichloroethylene. The following parameters were examined: fertilization rate, postimplanta- tion loss, preimplantation loss, and dominant lethal mutations (Slacik-Er- ben et al., 19801. In summary, trichloroethylene was mutagenic in one microbial muta- genicity assay and in a host-mediated assay in mice. Negative results were obtained in a dominant lethal assay in mice. Carcinogenicity A draft technical report of a study by the National Toxicology Program (1982a) described a bioassay of trichloroethylene for carcinogenicity in both sexes of B6C3F~ mice and Fischer 344 rats. The results for Marchall, ACI, and Osborne-Mendel rats are also contained in that report. Positive findings of carcinogenicity in B6C3F~ mice in a pre- vious study by the National Cancer Institute (1976) were discussed in the first and third volumes of Drinking Water and Health (National Research Council, 1977, pp. 778-779; 1980, pp. 163-164~. In the more recent study, the B6C3F~ mouse was used primarily as a laboratory positive control. The B6C3F~ mouse and the Osborne-Mendel rat were also used to clarify ques- tions raised by the earlier study concerning the housing of the test animals in the same room with animals receiving a variety of other volatile chemi- cals and by the presence of epichlorohydrin as a stabilizer in the test com- pound. A single 1,000 mg/kg bw dose of trichloroethylene in corn oil was administered by gavage to 50 mice of each sex, and doses of 1,000 and 500 mg/kg bw were given in the same manner to 50 rats of each sex 5 days/ week for 2 years. For each group of test animals there were corresponding groups of controls composed of 50 animals of each sex. The trichloroethyl- ene was stabilized with an amine antioxidant (diisopropylamine) and con- tained no detectable traces of 1,2-epoxybutane or epichlorohydrin. The sensitivity of the analytical method used to detect epichlorohydrin was 0.001 No. Each group of animals was housed in a separate room. The results observed in the mice support the previous National Cancer Institute (1976) findings that trichloroethylene significantly increased the incidence of hepatocellular carcinomas in B6C3F~ mice of both sexes. In general, the primary findings indicate that the Marshall, ACI, Os- borne-Mendel, and Fischer 344 rats have a much lower sensitivity to the
Toxicity of Selected Contaminants 83 carcinogenicity of trichJoroethylene than was observed for the B6C3F' mouse. The tumor rates and sites are detailed in the following section on risk estimation. In a chronic inhalation study of trichloroethylene stabilized with an amine antioxidant, Henschler et al. (1980) found no evidence of carcino- genicity in mice, rats, or Syrian hamsters. In an epidemiological study of cancer deaths among a small study group of 518 men occupationally exposed to trichloroethylene, there was no sta- tistically significant excess of cancer deaths (Axelson et al., 1978~. The au- thors noted that the study could not conclusively rule out an increased can- cer risk to humans, especially with respect to types of neoplasms that are rare in humans, including liver tumors. Carcinogenic Risk Estimate In the study by the National Toxicology Program (1982a), there was an increased incidence of hepatocellular carci- nomas in the exposed mice. The tumor incidences are summarized in Table II-8. Each set of data showing a statistically significant increase was used to estimate lifetime risk and an upper 95~o confidence estimate of lifetime risk in humans after a daily consumption of 1 liter of water containing the compound in concentrations of 1 ~g/liter. The estimates of risk are based on the multistage model for carcinogenesis described earlier in this chapter for chlorobenzene. The conversion of animal to human doses is again based on body surface area, assuming the following weights: humans, 70 kg; rats, 400 g; and mice, 33 g. The conversion formula is: animal consumption = human con- sumption X (human weight/animal weight)''3. The human dose estimates were also reduced by a factor of 5/7 to take into account the fact that the test animals were only Savaged 5 days per week. Using the data from the study by the National Toxicology Program (1982a), the committee estimated the lifetime risk and upper 95~o confi- dence estimate of lifetime risk to humans after a daily consumption of TABLE Il-8 Tumor Incidence in Trichloroethylene-Exposed Micea Tumor Animals Sex Site Dose, Tumor mg/lcg/day Rates B6C3F' mouse Male Liver B6C3F' mouse Female Liver O. 1,000 8/48, 30/s0 O. 1,000 2/48, 13/49 a Based on data from the National Toxicology Program, 1982a.
84 DRINKING WATER AND HEAl TH 1 liter of water containing the compound in a concentration of 1 ~g/liter (Table II-9). The National Toxicology Program's Technical Reports Review Subcom- mittee reviewed the data from the bioassay of the B6C3~ mice exposed to trichloroethylene. This group questioned certain aspects of the entire bio- assay, i.e., that an excess number of deaths were attributed to Savage er- rors and that the maximum tolerated dose (MTD) was exceeded in both doses given to the rats. This matter is now undergoing further review. In addition, that subcommittee has not as yet approved the trichloroethylene studies involving Marshall, Osborne-Mendel, and ACI rats. In previous volumes of Drinking Water and Health, the risk estimates from male and female rats and mice were averaged to yield one composite number. If one averages the data in Table II-9, the estimated upper 95~O confidence estimate of lifetime risk per ~g/liter is 3.3 X 10-7. This is sim~- lar to the 1.1 X 10-7 upper 95370 confidence estimate of lifetime risk per ~g/liter, which was calculated in the first volume of Drinking Water and Health (National Research Council, 1977, p. 794~. Using the criteria for interpreting animal carcinogenicity data as out- lined in Chapter I, the committee based the above calculations on limited evidence. Teratogenicity Schwetz et al. (1975) exposed pregnant mice and rats to 300 ppm (1,630 mg/m3) for 7 hours/day on days 6 through 15 of gestation. No fetal toxicity or teratogenicity was found. Dorfmueller et al. (1979) ex- posed female rats to 1~800 ppm (9,800 mg/m3) for 6 hours daily for 2 weeks before pregnancy and for the first 20 days of gestation. They found anomalies of skeletal and soft tissues, which were considered to be indica- tive of developmental delay. No behavioral effects were observed in the off- spring of the treated animals. Examination of sperm from mice exposed by inhalation to 0.3~o for 4 hours daily for 5 days revealed increased abnor- malities after 28 days (Land et al., 1981~. TABLE II-9 Carcinogenic Risk Estimates for Trichloroethylenea Upper 95~o Confidence Esti- Estimated Human mate of Lifetime Cancer Animal Sex Lifetime Riskb Risk per ~g/liter B6C3FI mouse Male 3.77 x 10-7 5.48 X 10-7 B6C3F' mouse Female 6.84 X 10-8 1.12 x 10-7 a Based on data from thc National To~cicology Program, 1982a. bAssuming a daily consumption of I liter of watcr containing the compound in a concentration of 1 ~B/ liter.
Toxicity of Selected Contaminants 85 Rather limited data indicate that trichloroethylene does not have signifi- cant teratogenic potential in mice and rats. CONCLUSIONS AND RECOMIdENDATIONS Information that became available since the last review of trichloroethyl- ene in Drinking Water and Health (National Research Council, 1980) is not sufficient to permit any more realistic assessment of suggested no- adverse-response levels than those made in Volume 3. Moreover, the addi- tional studies recommended in the last review identify needs that are still present. A greater understanding of the differences among various species will be necessary before data obtained in studies in animals can be extrapo- lated to humans with more confidence. The current risk estimation for exposure to trichloroethylene is very sim- ilar to that calculated in Volume 1 of Drinking Water and Health. Fur- thermore, this reconfirmation is important when and if limits for trichlo- roethylene in drinking water are established. It is also important that the uncertainties surrounding the carcinogenicity bioassay be resolved prior to reaching any decision on the potential for adverse health effects in humans following exposure to trichloroethylene. A chronic SNARL for trichlo- roethylene is not calculated because of its carcinogenicity in animals, but acute SNARL's were presented in Volume 3. VINYL CHLORIDE (MONOCHLOROETHYLENE) ethene, chIor~ CAS No. 75 014 H2C = CHCl Vinyl chloride was evaluated in the first volume of Dunking Water and Health (National Research Council, 1977, pp. 783-787~. The following ma- terial, which became available after the 1977 report was prepared, updates and, in some instances, reevaluates the information contained in the pre- vious review. Also included are some references that were not assessed in the earlier report. METAB OLISM After oral and inhalation exposures, vinyl chloride is rapidly absorbed, and its uptake can be saturated by either route (Bolt, 1978; Watanabe et al., 1976a,b). Phannacokinetics data indicate a dose-dependent metabo-