This chapter reviews information presented in the Environmental Protection Agency (EPA) draft Integrated Risk Information System (IRIS) assessment of the effects of tetrachloroethylene on the hematopoietic system, especially the development of mononuclear-cell leukemia (MCL) in rats and lymphomas in humans. The information is considered in the context of the other evidence on carcinogenicity in Chapter 11, where EPA’s assessment of carcinogenic risks of tetrachloroethylene is evaluated.
The draft IRIS assessment proposes to use the finding of MCL in male F344 rats as the most sensitive tumor response, supporting its weight-of-evidence classification of tetrachloroethylene as “likely to be carcinogenic to humans” by all routes of exposure. The use of MCL to support that conclusion is based primarily on two studies: those of the National Toxicology Program (NTP 1986) and the Japan Industrial Safety Association (JISA 1993). Both studies reported that chronic inhalation administration of tetrachloroethylene to male and female F344 rats caused “positive trends” in MCL with increasing dose. As the draft IRIS document correctly points out, the scientific reliability of those studies has been questioned in part because of “high spontaneous background incidences, use of special supplemental analysis to aid in data interpretation, and the relevance of MCL in F344/N rats to human hazard” (p 4-159, lines 21-23). The committee similarly questions the use of the tetrachloroethylene exposure bioassays in the F344 rat for cancer risk assessment for those reasons and others discussed below.
In the NTP (1986) study, F344/N male and female rats were exposed chronically to tetrachloroethylene at 200 and 400 ppm. The incidence of MCL in males was 77% in the 200-ppm group and 74% in the 400-ppm group, and in females 60% and 58%, respectively. The background incidences of MCL in the controls were high (56% in males and 36% in females). Such high backgrounds
make it difficult to interpret the biological significance of the increase in the incidence of MCL observed in the treatment groups. Indeed, NTP has decided to stop using its F344/N rat colony in its bioassays for reasons that include the high background rate of MCL (King-Herbert and Thayer 2006). A supplemental analysis performed by NTP considered disease progression, latency, and various statistical treatments. The analysis suggested an increase in tumor incidence over controls at both test concentrations despite the high spontaneous tumor incidence in the controls.
The significance of MCL findings in multiple NTP bioassays that used the F344 rat was the subject of a recent reanalysis by Thomas et al. (2007), which EPA should reference in the draft IRIS assessment. They examined the incidence of leukemia in 2-year bioassays that included untreated male and female F344 rats from 1971 to 1998. They found that background tumor incidence increased substantially, from 7.9% to 52.5% in males and from 2.1% to 24.2% in females, over that period. The analysis also found that MCL responses are highly variable and subject to substantial modulation by dietary factors.
Thomas et al. (2007) also evaluated MCL incidence in male and female rats exposed to 500 chemicals. On the basis of 34 NTP studies that yielded evidence of a chemically related increase in the incidence of leukemia, which included the 1986 NTP study of tetrachloroethylene, the authors conducted a reanalysis of dose-response data by comparing results with four statistical methods: Fisher’s test for pair-wise comparison of leukemia incidence between a dose group and a control group, the Cochran-Armitage test for incidence trend, logistic regression for incidence, and life tables for survival-adjusted incidence. Tetrachloroethylene was one of five chemicals shown by the authors to produce leukemia in both sexes of rats. They used the rigid Food and Drug Administration (FDA) statistical criteria for testing dose-related cancer incidences (p < 0.01 for pairwise comparison; p <0.005 for trend test). The results in male rats in the 1986 NTP study revealed a significant dose-response trend when analyzed with a life table (p = 0.004) assuming that MCL is lethal but a nonsignificant trend with logistic regression (p = 0.097) assuming the MCL is nonlethal. Pairwise comparisons revealed dose-related incidences (p = 0.046) for both dose groups, and the trend test yielded a p value of 0.034; neither met the FDA criteria for statistical significance. The borderline significance of the trend test and nonsignificance of logistic regression for the latter two comparisons could be explained in part by the fact that the incidences did not follow an incrementally increasing relationship with dose. In female rats in the NTP study, use of a life table (p = 0.053), logistic regression (p = 0.012), a trend test (p = 0.018), and Fisher’s test (p = 0.014 and 0.022, respectively, for two doses) all revealed a borderline significant dose-related incidence. However, there is inconsistency in statistical significance between the sexes and uncertainty about the shape of the dose-response curve, especially in the lower range of the study. The authors recommended the use of life-table analysis for survival-adjusted leukemia incidence, noting that it is “closer to reflecting the true statistical significance of the carcinogenic effect” than logistic-regression treating dose as linear. Life-table
analysis (log-rank test) accounts for time-to-event information, is capable of testing nonlinear dose-response relationships of arbitrary shapes, and is therefore more flexible than the Cochran-Armitage trend test. Survival analysis also is more relevant than logistic regression for more lethal tumors such as MCL. Overall, Thomas et al. showed a moderately significant dose-response relationship for tetrachloroethylene, but this finding should be evaluated by EPA with a weight-of-evidence approach suggested in its 2005 Guidelines for Carcinogen Risk before conclusions are drawn.
In the 1993 JISA study, F344/DuCrj rats were exposed to tetrachloroethylene at 50, 200, and 600 ppm. The draft IRIS document focuses on the JISA report for cancer dose-response assessment because the study included a 50-ppm exposure concentration, which is one-fourth the lowest exposure concentration in the 1986 NTP study. As in the NTP study, there was a high incidence of MCL in the controls (22% in males and 20% in females). Against that high spontaneous incidence of MCL, the incidence of MCL in male and female rats exposed to tetrachloroethylene at 50, 200, and 600 ppm was 28%, 44%, and 54% and 34%, 32%, and 38%, respectively. Moreover, the historical rate of MCL for the Japanese laboratory is very high. There was no incremental increase in MCL incidence in female rats with increasing dose. In contrast, EPA concluded that male rats displayed a dose-dependent increase in MCL although in the analysis background values were subtracted from the incidences in animals treated with tetrachloroethylene (Figure 5-6 in the draft IRIS assessment), and this may lead to a false impression. Such manipulation of data is not widely accepted in statistical practice, because it artificially reduces the uncertainty caused by the variation in the background rate. As noted in reviews by Caldwell (1999) and Ishmael and Dugard (2006), the unusually high background rate of MCL in control (untreated) rats weakens the ability to separate the background response from possible chemically induced responses, particularly when the chemically induced response above background is low. The committee recommends that the statistical approaches applied by Thomas et al. (2007) to the NTP study be applied also to the JISA study.
It is unclear whether MCL is a relevant predictor of human leukemias or other adverse health effects. Thomas et al. (2007) argue that MCL is a large granular lymphocytic leukemia (LGLL) of natural-killer (NK) cell origin that shares “some characteristics” with a rare human NK-LGLL. However, they also note that in contrast with F344 rats, human NK-LGL leukemia is rare, occurs primarily in the young, and may be associated with Epstein Barr virus (EBV) although no such virus-leukemia association is known to contribute to the etiology of rat LGLL/MCL. EPA contends that MCL is “similar” to human lymphoid cancers (T-cell and NK-LGL leukemias) and relies on a study (Stromberg 1985) that compared morphologic characteristics between rat MCL and human T-cell lymphoma. EPA considers that to be supportive evidence, despite the fact that these cancers arise in different tissues and that the cell origin in both cases is unknown. EPA states (EPA 2008, p. 4-161) that “discounting a rodent neoplasm simply because it has no human counterpart is not a scientifically defensi-
ble position. Strict site concordance is not a requirement for relevance in extrapolation of hazard potential.” The committee agrees with those statements, but notes that the available data should be used to provide a more convincing argument. Similarly, EPA argues that humans are heterogeneous and so could have the same inherited susceptibility as F344 rats, but provides no scientific basis for that argument.
Few human data are available for assessing the relationship between tetrachloroethylene exposure and the risk of specific cell types of lymphohematopoietic cancers. Several studies have assessed the risk of chronic lymphocytic leukemia in humans (Morton and Marjanovic 1984; Travier et al. 2002; Ji and Hemminki 2005, 2006), but otherwise the finest classification of outcomes used was “leukemia,” “lymphoma,” “non-Hodgkin lymphoma” (NHL), and “Hodgkin disease” (HD). The EPA draft IRIS assessment concludes (p. 4-184) that the epidemiologic data “suggested an association between lymphoma and tetrachloroethylene.” The committee concurs with that conclusion but would add that the data are relatively weak and inconsistent. Associations between those cancers and exposure to tetrachloroethylene are based on very small numbers and thus are statistically unstable. The positive associations with tetrachloroethylene are sometimes observed only for lymphomas in women: NHL reported by Spirtas et al. (1991) and Anttila et al. (1995) and HD reported by Blair et al. (2003) and Miligi et al. (2006). It is not clear why those differences in sex-specific results appear; they may be due to residual confounding, in that it is unlikely that men would have appreciably lower exposures than women in the same jobs. It is also possible that sex-specific susceptibility issues are contributing to this observation. Other large cohort studies (Boice et al. 1999; Lynge et al. 2006) found no association in either women or men, and no dose-response effects have been observed. Epidemiologic studies of the association vary with study design, validity, specificity of exposure assessment, type of population studied, and sample size, all of which contribute to the inconsistency of results and reduce the committee’s confidence in the conclusions that can be drawn from the data. The committee also noted a number of factual errors in this section of the IRIS draft that should be corrected; such errors detract from overall confidence in the draft’s conclusions.
MODE OF ACTION
Given the high background rate of MCL in F344 rats, it is important to question whether tetrachloroethylene induces MCL or promotes an increase over the background rate. However, few data are available for addressing the question. According to EPA, a link to a mode of action (MOA) for tetrachloroethylene-induced MCL implicates a circulating genotoxic metabolite that is formed in
the kidney by cleavage of a cysteine conjugate, S-(1,2,2-trichlorovinyl)-L-cysteine (TCVC) and may cause DNA damage in bone marrow. The EPA draft discusses studies that showed that a related (trichloroethylene-derived) cysteine conjugate, S-(1,2-dichlorovinyl)-L-cysteine, caused DNA alterations and toxicity in the bone marrow, lymph nodes, and thymus of calves (Bhattacharya and Schultze 1971, 1972; Lock et al. 1996). The finding that TCVC did not induce those responses in the same study does not appear to have factored into EPA’s support of the hypothesis of a genotoxic MOA. The committee judges that a genotoxic MOA of tetrachloroethylene induction of MCL involving the cysteine conjugate β-lyase pathway is highly speculative and not supported by data.
The committee found some additional data on tetrachloroethylene that might be relevant for MOA analyses. They include studies by Marth et al. (Marth et al. 1985, 1989; Marth, 1987) and a study by Seidel et al. (1992) on tetrachloroethylene toxicity in mice. In the Marth et al. studies, NMRI mice were orally exposed to tetrachloroethylene at 0.05 or 0.1 mg/kg per day for 7 weeks. The mice exhibited a reversible hemolytic anemia and had microscopic evidence of splenic involvement (Marth et al. 1985), and tetrachloroethylene was found to accumulate in the spleen (as shown in Figure 2 of Marth et al. 1989), where MCL is thought to originate. Nevertheless, hemolytic anemia arises as a result of a defect in the mature red-cell membrane, as opposed to the various forms of leukemia which are thought to arise as a result of mutational changes early in bone-marrow-cell differentiation. Thus, hemolysis would not be expected to play a role in leukemogenesis. The observations reported by Marth et al. have not been reproduced or reported by any other laboratory.
Seidel et al. (1992) exposed hybrid mice (C57/BL/6 × DBA/2) to tetrachloroethylene at 270 ppm (11.5 weeks) and 135 ppm (7.5 weeks) 6 hours/day 5 days/week. Reductions in the numbers of lymphocytes/monocytes and neutrophils were observed, but they returned to control values over the next 3 weeks. There were no effects on spleen colony-forming units (CFU-Ss), but evidence of a reduction in red cells was supported by decreases in erythroid colony-forming units and erythroid burst-forming units and evidence of reticulocytosis. The data suggest a reversible bone marrow depression.
Inhibited production of both red cells and various forms of white cells have been reported after exposure to a variety of leukemogens (such as anticancer alkylating agents or benzene). The leukemogens usually decrease CFU-Ss, an effect not observed with tetrachloroethylene exposure (Seidel et al. 1992). They also usually decrease the bone marrow myeloid progenitors, CFU-GEMM, CFU-GM, and CFU-E/BFU-E, the latter of which was also decreased by tetrachloroethylene (Seidel et al. 1992). EPA should consider reviewing the evidence from models of leukemia induced in humans by chemicals (such as benzene and chemotherapeutic agents) to determine whether there are similarities with tetrachloroethylene-induced MCL.
The Marth et al. studies and the Seidel et al. study provide indirect evidence that tetrachloroethylene exposure induces effects associated with MCL and known leukemogens, respectively, but are insufficient to support the argu-
ment that tetrachloroethylene induces MCL or a related form of leukemia. In addition, those studies investigated tetrachloroethylene exposure in mice, a species in which MCL has never been observed. The only evidence that tetrachloroethylene induced MCL comes from exposure studies with F344 rats. Nevertheless, the effects of tetrachloroethylene on hemolysis in mice and on bone marrow function provide the basis of a hypothesis that could be explored to demonstrate the mechanism by which tetrachloroethylene could, within some dose range, affect the spleen.
The majority of the committee finds that EPA has not adequately justified the use of MCL data over the evidence for liver or kidney cancer in its cancer risk assessment. Evidence of tetrachloroethylene-induced leukemia from epidemiologic studies is limited and inconsistent. The NTP (1986) and JISA (1993) study results of increased MCL incidences in F344 rats given tetrachloroethylene by inhalation are also questionable because of the high background rates of MCL in control animals. More thorough statistical evaluation of the data, such as the life-table analysis proposed by Thomas et al. (2007), could provide a stronger basis for drawing conclusions. However, MCL resulting from tetrachloroethylene exposure has not been observed in other strains of rats or other animal species, and no definitive evidence is available to support a hypothesized MOA by which tetrachloroethylene increases MCL in F344 rats. Those are all sources of uncertainty surrounding the relevance of MCL to human cancer risk. The information is considered in the context of the other evidence on carcinogenicity in Chapter 11, where EPA’s assessment of carcinogenic risks of tetrachloroethylene is evaluated.