Renal Toxicity and Cancer
This chapter reviews information presented in the Environmental Protection Agency (EPA) draft Integrated Risk Information System (IRIS) assessment of the toxic and carcinogenic effects of tetrachloroethylene on the kidney. The metabolism of tetrachloroethylene by the kidney is critical for its toxicity and carcinogenicity in that organ. The major metabolites of tetrachloroethylene responsible for renal effects are formed by the glutathione metabolic pathway (see Chapter 2 for an overview of toxicokinetics). The following sections address renal toxicity and carcinogenicity separately, but they are not necessarily independent end points. This information is considered in the context of the other evidence on carcinogenicity in Chapter 11, where EPA’s assessment of carcinogenic risks posed by tetrachloroethylene is evaluated.
The draft IRIS assessment notes that published information on renal toxicity in humans is not well developed. That is because typical screening tests that use plasma are insensitive. For instance, blood urea nitrogen and creatinine, which accumulate in plasma when glomerular filtration is diminished, do not increase until renal function is about half of normal, and urinalysis is not typically performed. Epidemiologic studies of the effects of tetrachloroethylene exposure on renal function have been reported, and EPA summarizes the findings in a table. The discussion focuses on urinary proteins that are indicative of tubular damage, because β-lyase is found in the proximal tubule. The strengths and weaknesses of the various studies are noted by EPA, and consistencies and inconsistencies are discussed. In general, different reports examined different urinary proteins, which have different sensitivity and selectivity as markers of tubular function. Estimated exposure differed among the reports, as did the number of subjects. Effects on glomerular function, as assessed on the basis of
albuminuria, are discussed briefly. The draft IRIS assessment notes that the results are contradictory. It should also note that some albumin is normally filtered, so small increases in the amount of albumin in the urine can be indicative of tubular damage (the result of failure to reabsorb the small amount filtered). EPA’s table should also include the negative findings on albumin in studies by Verplanke et al. (1999) and Lauwerys et al. (1983) and on total protein by Vyskocil et al. (1990). EPA concluded that the epidemiologic studies provided evidence suggestive of subtle damage in renal tubules. The committee agrees with that assessment.
Several types of epidemiologic studies have been used to explore a possible association between jobs in which workers are exposed to tetrachloroethylene and renal-cell carcinoma (RCC), including cohort mortality studies, case-control studies, and nested case-control studies. Ultimately, the methodologic challenges of studying such a rare cancer as RCC, assessing tetrachloroethylene exposure accurately, and evaluating inconsistencies in results among studies limit the conclusions that can be drawn from the epidemiologic data. Most of the studies either did not have explicit information about exposures or had considerable methodologic limitations.
Pesch et al. (2000) conducted a population-based case-control study in Germany that estimated tetrachloroethylene exposure with a job-exposure matrix (JEM) and a job-task exposure matrix (JTEM). The latter is usually superior for estimating specific exposures. The data were acquired in in-person interviews, so information on occupational history was obtained and confounding covariates (such as smoking) were well measured. An increased odds ratio (OR) for tetrachloroethylene exposure was observed in men who had a medium exposure index (OR, 1.4; 95% confidence interval [CI], 1.1-1.7) or a substantial exposure index (OR, 1.4; 95% CI, 1.1-2.0) on the basis of the JEM. However, the results based on the JTEM were less convincing (OR, 1.2; 95% CI, 0.9-1.7 and OR, 1.3; 95% CI, 0.7-2.3 for medium and substantial exposure, respectively). In contrast, no association was observed in women on the basis of the JEM, but a positive albeit imprecise association was observed on the basis of the JTEM for medium and substantial exposure (OR, 2.2; 95% CI, 0.9-5.2 and OR, 2.0; 95% CI, 0.5-7.8, respectively). Those variable results are representative of inconsistencies among studies. Lynge et al. (2006) (listed in Table 4B-4 of the EPA draft but not discussed in the renal-cancer section) conducted a nested case-control study in four Scandinavian countries in a cohort of about 47,000 persons employed in the laundry and dry-cleaning industry as of 1970 and followed through 1997-2001 to identify incident cancers. Multiple cancers were assessed, including 56 RCC cases in men and 154 in women. The cohort was divided into those who were not exposed to the dry-cleaning process, dry-cleaners and other exposed workers, and others working in dry-cleaning. Risk was also estimated by
duration of employment in dry-cleaning occupations. Tetrachloroethylene was the most commonly used solvent in dry-cleaning during the study interval; the mean concentration over the interval of the study was estimated as 24 ppm. The adjusted relative risk of RCC for dry-cleaners compared with unexposed workers was 0.67 (95% CI, 0.43-1.05) on the basis of 29 cases in the exposed. There was no evidence of increasing risk with increasing duration of employment as dry-cleaners. Mandel et al. (1995) pooled data from a multicenter international case-control study of RCC; the study was conducted in six centers in five countries (Australia, Denmark, Germany, Sweden, and the United States) and included 1,732 cases and 2,309 controls. Occupational histories, collected in inperson interviews, were used to estimate exposures to specific chemicals or tasks. The study reported an increased OR of 1.4 (95% CI, 1.1-1.7) associated with exposure to dry-cleaning solvents, but no exposure response was observed on the basis of duration of exposure.
Several other studies, although methodologically sound, were too small or did not have sufficient information about exposure to be informative (Aschengrau et al. 1993; Mellemgaard, et al. 1994; Schlehofer et al. 1995; Dosemeci et al. 1999).
There are inconsistencies in the draft IRIS assessment. Nine studies are listed as larger case-control studies. Of them, EPA judged the case-control studies of Aschengrau et al. (1993), Partanen et al. (1991), and Pesch et al. (2000) to be of high quality, citing exposure information, adequate control of confounding, and histologic confirmation. It is then noted that “these two case-control studies carry greater weight than observations in the other case-control studies identified in Table 4B-4.” The Aschengrau et al. study is not listed in Table 4B4; and this suggests that the Partanen et al. and Pesch et al. studies are those considered to be the studies given greater weight. The point should be clarified. The Lynge et al. (2006) study is not discussed in the “Kidney Cancer in Humans” section of the draft IRIS assessment.
Overall, the epidemiologic literature provides little support for a causal association between tetrachloroethylene exposure and cancer of the kidney. Study results are inconsistent. In addition, those studies that tried to assess dose-response by using the imperfect surrogate of “duration of exposure,” found no association between duration and risk. EPA’s assessment of the data appropriately labels the evidence supporting an association between tetrachloroethylene and renal cancer as “limited,” and the epidemiologic evidence does not appear to weigh heavily toward classifying tetrachloroethylene as a likely carcinogen.
The draft IRIS assessment summarizes the studies of tetrachloroethylene toxicity across species, sexes, and routes and durations of exposure. Significant renal toxicity has been observed in lifetime bioassays in rats and mice of both
sexes (NCI 1977; NTP 1986; JISA 1993). Degenerative changes in the proximal tubule are reported as cloudy swelling, fatty degeneration, and necrosis of the epithelium. Some tubules were filled with hyaline casts; inflammatory cells, fibrosis, and focal mineralization were also reported. Effects of shorter exposures depended on route, duration, and dose. In short term (28-42 days) gavage studies, male rats showed signs of renal damage (Green et al. 1990; Philip et al. 2007). Inhalation exposure of male and female rats and mice to tetrachloroethylene for 28 days caused no effects at 400 ppm, and exposure of male rats for 10 days at 1,000 ppm resulted in an increase in hyaline droplets (Green et al. 1990). Inhalation exposure to tetrachloroethylene for 13 weeks resulted in karyomegaly in male and female mice but not in rats (NTP 1986); the response was minimal at 200 ppm and increased in severity with exposure concentration.
Renal-tubular adenoma and carcinoma were observed in male rats in the NTP (1986) bioassay and to a lesser extent in the Japan Industrial Safety Association (JISA 1993) studies. Tetrachloroethylene caused a low rate of induction of renal tumors in rats; although the yield at the high dose was not statistically significant. In the NTP bioassay, induction of renal tumors was dose-dependent. The incidence was 1 of 49 in the control group, 3 of 49 in the 200-ppm group, and 4 of 50 in the 400-ppm group. There are wide confidence limits on the data, and some of the error bars approach zero. There is a very low spontaneous incidence of renal tumors in Fischer 344 rats (Haseman et al. 1998). Induction of renal tumors in rats by tetrachloroethylene is therefore easily observed against a low background. In addition, the controls had only benign tumors, not malignant tumors, whereas the high-dose group had two malignant tumors. In the draft IRIS assessment, EPA calculates the chance that two animals will have a rare tumor to be less than 0.001, giving biological relevance to the finding. Maltoni and Cotti (1986) observed no increase in kidney tumors following tetrachloroethylene administration by gavage to male Sprague-Dawley rats. Overall, the dose-dependent induction of renal tumors in one experiment against the low background incidence of renal tumors in rats exposed to tetrachloroethylene indicates that tetrachloroethylene can induce renal tumors in rats. After integrating the results of the studies, the committee concluded that tetrachloroethylene induces renal tumors in rats. EPA considers the renal tumors to be suggestive of an effect and notes that it is similar to the effects of other chlorinated ethanes and ethylenes. The committee agrees with EPA’s assessment.
Mode of Action
EPA considered key events and potential modes of action for renal-tumor formation following tetrachloroethylene exposure and concluded that the mechanisms are not understood.
The draft IRIS assessment discusses an α2μ-globulin nephropathy mode of action of tetrachloroethylene-induced renal carcinogenesis in detail. Renal tumors that arise solely by α2μ-globulin nephropathy are not considered relevant to human risk assessment, because α2μ-globulin nephropathy is specific to the male rat. Although hyaline droplets that contain α2μ-globulin have been reported after exposure to high concentrations of tetrachloroethylene, the histopathologic findings reported in the inhalation bioassays were not consistent with the α2μ-globulin-mediated mode of action (NTP 1986; JISA 1993). Gavage bioassay (NCI 1977) showed that histopathologic characteristics were more consistent with α2μ-globulin nephropathy. However, in all these bioassays, similar histopathologic findings in the kidney were reported in female rats and male and female mice. These positive responses are not consistent with the male rat specificity of the α2μ-globulin nephropathy mode of action and therefore contradict a role of α2μ-globulin nephropathy in renal tetrachloroethylene tumorigenesis. The committee agrees with EPA’s assessment that α2μ-globulin nephropathy is not supported as a mode of action in tetrachloroethylene-induced renal carcinogenesis.
Tetrachloroethylene can stimulate the peroxisome proliferation response, as indicated by cyanide-insensitive palmitoyl CoA oxidation activity, in the kidneys of mice but not rats (Goldsworthy and Popp 1987). Odum et al. (1988) reported similar findings; mouse kidney samples were pooled for assays, so statistical analysis was not conducted on mouse kidneys. The peroxisome proliferation response does not correlate with tumor response and therefore is not consistent with a role of peroxisome proliferation as a mode of action in renal tumorigenesis. EPA notes that activation of peroxisome proliferator-activated receptors has not been established as a mode of renal tumorigenesis. The committee agrees that the data do not support peroxisome proliferation as a mode of action.
The draft IRIS assessment also considers immunotoxicity and immunosuppression as a mode of action of tetrachloroethylene tumorigenesis. In humans, immune-mediated renal damage is most often seen as damage to the glomeruli. The reports of renal damage in humans are based on abnormal protein in the urine; the pattern of proteinuria is indicative of tubular, not glomerular, damage. Thus, the type of renal damage seen is not consistent with an immunotoxic mode of action. The draft IRIS assessment notes that immunesystem-mediated effects of organic solvents and the formation of protein adducts are related to autoimmune diseases, not to immunosuppression and therefore inconsistent with immunosuppression as a mode of action.
Tetrachloroethylene causes toxic nephropathy in high doses, and this was observed in the cancer bioassay studies (NCI 1977; NTP 1986; JISA 1993). EPA considered a mode of action in which renal cytotoxicity and subsequent proliferation—as part of the repair process, not associated with α2μ-globulin—result in renal-tubular neoplasia. Renal toxicity has been observed with various metabolites of tetrachloroethylene (Lash et al. 2007; Elfarra and Krause 2007). Each of the three major metabolic pathways of tetrachloroethylene yields metabolites that are cytotoxic (Dekant et al. 1986c, 1988; Vamvakas et al. 1989a,c;
DeMarini et al. 1994; Werner et al. 1996; Volkel and Dekant 1998; Muller et al. 1998a; Dreessen et al. 2003). Chronic nephrotoxicity has been reported in male rats at the termination of all long-term bioassays but also has been observed in chronic bioassays at 2 years in female rats and both sexes of mice, none of which develop tumors. Despite this inconsistency, it is not possible to rule out a role of chronic toxicity in tumor formation.
The draft IRIS assessment concludes that a mutagenic mode of action cannot be ruled out. The committee agrees with this assessment. A mutagenic mode of action is supported by the findings after exposure to the structurally similar trichloroethylene. Some metabolites derived from S-(1,2,2-trichlorovinyl) glutathione (TCVG), the glutathione conjugate of tetrachloroethylene, have been shown to be mutagenic in bacterial systems (Vamvakas et al. 1989a,d) or to cause unscheduled DNA synthesis (Vamvakas et al. 1989c). Others react with DNA in vitro (Muller et al. 1998a,b). S-(1,2,2-Trichlorovinyl)-L-cysteine (TCVC) causes a greater response than dichlorovinyl cysteine in mutagenicity tests using Salmonella (Dekant et al. 1986c) and in renal toxicity (Birner et al. 1997). Tetrachloroethylene has not been shown to be mutagenic with or without activation by S9 in Salmonella or in mammalian cells. However, when tetrachloroethylene was activated with purified glutathione S-transferase, glutathione, and rat kidney fractions, TCVG was formed, and consequent mutagenic activity in Salmonella was clearly demonstrated, as described by EPA. S9 activation of tetrachloroethylene did not induce mutation in cultured mouse lymphoma L5178Y cells.
SUMMARY AND RECOMMENDATIONS
EPA concluded there is limited evidence that tetrachloroethylene causes cancer in humans, and the committee agrees with this assessment. EPA evaluated bioassay studies to provide evidence suggestive of an effect. The committee considers this and the similarity to trichloroethylene to support the conclusion that tetrachloroethylene induces kidney tumors in rodents. While the mode of action of tetrachloroethylene tumorigenesis is not understood, the α2μ-globulin nephropathy and peroxisome proliferator modes of action are not consistent with experimental results. A mutagenic mode of action cannot be ruled out.
Further studies are needed to determine whether tetrachloroethylene and its metabolites formed from TCVG (TCVC, chlorothioketene, and sulfoxide metabolites) are mutagenic in other mammalian cell assays (mutation to 6thioguanine resistance in cultured V79 Chinese hamster lung fibroblasts or in Chinese hamster ovary cells). It is possible that any of the metabolites of TCVG contribute to the carcinogenicity of tetrachloroethylene in rat kidney, but this needs to be studied. Further data on the sequencing of DNA from tetrachloroethylene-induced renal tumors for mutations of the von Hippel Landau tumor-suppressor gene, other tumor-suppressor genes and oncogenes, and their downstream effectors (for example, p27 that controls cell-cycle progression) are needed to determine whether TCVG and similar tetrachloroethylene metabolites