This chapter reviews information about the effects of trichloroethylene on the immune system, particularly information generated since the U.S. Environmental Protection Agency released its draft health risk assessment (EPA 2001b). Consideration is given to how the new information factors into previous assessments of the immunosuppressive and autoimmune effects of trichloroethylene, species differences, dose-response relationships, and mode of action information.
Immunotoxicity can be divided into two areas depending on whether the immune system is activated (such as in allergies or in chemical-induced autoimmune diseases) or suppressed by xenobiotics (foreign or nonendogenous chemicals, including drugs and environmental chemicals). Mammalian immune systems have innate and adaptive components that play important roles in resistance to infections and cancer. The immune systems of mammals are formed by primary lymphoid organs, including yolk sac, fetal liver, bone marrow, and thymus. Secondary lymphoid organs (e.g., lymph nodes, spleen, mucosa-associated lymphoid tissues) store differentiated cells that await activation by environmental antigens or undergo endogenous selection processes to discriminate self from nonself. T and B cells are activated in clonally restricted (antigen-specific) ways, and they demonstrate a memory response. One feature of innate immunity is that the responding cells (macrophages, natural killer cells, granulocytes) do not demonstrate clonal specificity. However, families of receptors have been identified (such
as toll-like receptors) that allow innate cells to respond to certain families of environmental molecules or toxins (e.g., endotoxin). Xenobiotics may interfere with normal immune system homeostasis by affecting the formation of immune cells; modifying cell-to-cell interactions; modifying cell activation, proliferation, or differentiation; altering cell selection; and enhancing or suppressing the release of immune products such as cytokines, chemokines, antibodies, and complement factors.
The immunotoxicity of chemicals is evaluated in animal models, in in vitro studies, and occasionally in humans after occupational or environmental exposures. Environmental epidemiology studies are often conducted to determine whether xenobiotic exposures are associated with disease. Because of the complexity of the innate and adaptive immune systems, no single assay can be used to study the potential toxicity of xenobiotics. Instead, a tiered approach has been developed and validated by several laboratories for studies in animals (Luster et al. 1988, 1992). Although there is no single immune assay or parameter that can be used to determine whether a xenobiotic exerts a toxic effect on the immune system, certain combinations of markers and functional assays can predict immunotoxicity (Luster et al. 1992). Additionally, the aforementioned assays are useful only for evaluating immunosuppressive chemicals. Few established assays exist for assessing hypersensitivity reactions of xenobiotics, and experimental models of autoimmunity are limited in their application and extrapolation to human autoimmune diseases.
The potential immunosuppressive and immunomodulating properties of trichloroethylene in acute, subchronic, and chronic exposures in animals have not been fully evaluated. Sanders et al. (1982) found that trichloroethylene at concentrations of 2.5-5 mg/mL (in drinking water for 4 or 6 months) resulted in suppression of humoral and cell-mediated immunity in female CD1 mice. Bone marrow stem cell activity was depressed at drinking water concentrations of 0.1-1 mg/mL. Male mice were less affected. Wright et al. (1991) found a depression in natural killer cell activity in the liver, decreased lipopolysaccharide lymphocyte mitogenesis, and decreased spleen weights after intraperitoneal exposures of Sprague-Dawley rats to trichloroethylene at 5 mmol/kg/day for 3 days. Natural killer cell activity in the liver was also depressed at 0.5 mmol/kg/day for 3 days. B6C3F1 mice receiving the high-dose regimen also demonstrated spleen cell toxicity, and they were more sensitive than rats to the natural killer cell suppression in the liver, with effects observed at 0.05 mmol/kg/day for 3 days. Aranyi et al.
(1986) found that acute exposures to various solvents decreased the host resistance responses to Klebsiella pneumoniae. In these studies, trichloroethylene was not evaluated, but a related solvent, perchlorethylene, had a small effect. Kauffmann et al. (1982) found that mice exposed to chloral hydrate at 1/10th (144 mg/kg over 14 days) and 1/100th of a median lethal dose (14.4 mg/kg over 14 days) had no changes in immune parameters. However a 90-day exposure to trichloroethylene at 0.07 and 0.7 mg/mL in drinking water produced a significant decrease in humoral immunity in female, but not male, mice. Park et al. (1993) found that trichloroethylene at 50-200 parts per million (ppm) increased infection in bacteria-challenged (Streptococcus zooepidimicus) mice.
Kaneko et al. (2000) determined that inhalation of high concentrations of trichloroethylene (500-2,000 ppm) for 8 weeks depressed the serum IgG in mrl/lpr mice and increased the formation of lymphoblastoid cells. Changes in T-cell subsets (helper to suppressor ratio) were detected at 2,000 ppm after 8 weeks of exposure. The significance of these findings is difficult to assess because the investigators used an autoimmune-prone mouse strain (mrl/lpr), which is not commonly used in studies of immunosuppression.
In summary, various studies indicate that exposures to moderate or high concentrations of trichloroethylene over long periods have the potential to produce immunosuppression in animal models. There are important differences in the amounts and types of immunosuppression depending on species and gender.
Epidemiology and case studies revealed that solvents, including trichloroethylene, might be associated with certain human autoimmune diseases. These reports triggered investigators to evaluate the effect of trichloroethylene in animal models susceptible to the induction of autoimmune disease, most notably the MRL mouse. Autoimmune disease has not been reported in normal mice treated with trichloroethylene, although a small increase in autoantibodies has been noted in some studies (see below).
Findings in Rodents
Several laboratories have reported that trichloroethylene causes or exacerbates underlying autoimmune diseases in genetically susceptible MRL mice. Effects have been observed at doses as small as 0.1 mg/kg/day in drinking water for 4 weeks (Griffin et al. 2000a), which the authors calculated may be below the current threshold limit value of 50 ppm set by the American Conference of Industrial Hygienists. Recently, extensive mechanistic work has been performed and several biologically plausible hypotheses have
been advanced (Khan et al. 1995, 2001; Gilbert et al. 1999, 2004; Griffin et al. 2000a,b,c; Blossom et al. 2004). Several studies focused on the need for metabolism of trichloroethylene to chloral or dichloroacetic acid to produce autoimmune-induced hepatitis in genetically susceptible mice (Griffin et al. 2000a,b) (see Chapter 4), a syndrome that bears potential mechanistic similarities to halothane-induced hepatotoxicity in rodents and humans. Evidence has been presented that trichloroethylene or its metabolites may activate T cells (Gilbert et al. 1999, 2004; Griffin et al. 2000a) and/or alter T cell regulation and survival (Blossom et al. 2004) associated with polyclonal disease, as detected by circulating anti-DNA and other antibodies in genetically susceptible mice. Theoretically, trichloroethylene metabolites may be increased with enhancers of the CYP2E1 gene, and autoimmunity in MRL mice induced by trichloroethylene has been shown to be inhibited with CYP2E1 inhibitors (Griffin et al. 2000b). Despite these hypotheses, the mechanism(s) by which trichloroethylene exacerbates autoimmune disease in MRL mice has not been elucidated and the relevance to human exposures and disease has not been established.
Gilkeson et al. (2004) administered trichloroethylene at 1,000-10,000 parts per billion (ppb) in drinking water to NZB/NZW mice for 26 weeks. They found an increase in anti-DNA antibodies with trichloroethylene at 1,000 ppb and an increase in kidney disease at 10,000 ppb. In the same study, B6C3F1 mice developed a small increase in autoantibody production, but no kidney disease was detected.
White et al. (2000) found a lack of evidence for trichloroethyleneinduced autoantibody production and systemic-lupus-erythematosus-like disease in Brown Norway rats when trichloroethylene was given by oral gavage 5 days a week for 6 weeks at 100-400 mg/kg.
Changes in Hematologic Parameters in Dogs
Hobara et al. (1984) found that acute inhalation exposure to trichloroethylene (200 and 500 ppm) or intravenous injection (50 mg/kg) in beagles produced a transient decrease in circulating leukocytes, most notably neutrophils that rebounded to near control concentrations after several hours.
Following is a brief qualitative review of some human studies that have investigated trichloroethylene in relation to immunologic end points. It is provided to give a perspective on some of the important areas to be pursued as part of the risk assessment for trichloroethylene. A more thorough review of the epidemiologic studies in terms of methods, exposures, and results are
necessary to fully characterize the immunologic hazards posed by trichloroethylene (see Chapter 2 for guidance on how this should be done).
Immunosuppression and Immunomodulation
Byers et al. (1988) reported on a human leukemia cluster putatively exposed to high concentrations of trichloroethylene and other solvents via contaminated drinking water. There were long-term alterations in peripheral blood T-cell subsets in the family members of those with leukemia, which suggests of an immunologic abnormality. There were increases in infections as well as an increased number of autoantibodies in this cohort. Lehman et al. (2002) found that children exposed in utero to volatile organic compounds, including trichloroethylene, had a shift toward TH1 γ-interferonproducing T cells analyzed 6 hours after birth. These two studies suggest that immunologic changes may be seen in solvent- or trichloroethylene-exposed humans, although it has been difficult to quantify the exposures.
Numerous investigators have found an association between exposure to organic solvents, including trichloroethylene, and the human autoimmune diseases scleroderma (Saihan et al. 1978; Lockey et al. 1987; Waller, et al. 1994; Bovenzi et al. 1995, 2004; Nietert et al. 1998, 1999; Pandey and Takeuchi 1999; Pandey et al. 2001; Czirjak and Kumanovic 2002; Garabrandt et al. 2003; Pandey 2004) and Stevens-Johnson Syndrome (Pantucharoensri et al. 2004). The risk of scleroderma may be correlated with particular CYP2E1 and CYP2C19 polymorphisms (Povey et al. 2001), suggesting that trichloroethylene metabolism could be important in this disease. However, more research is needed to elucidate this possibility.
ISSUES FOR IMMUNOTOXICITY RISK ASSESSMENT
Extrapolation of Animal Data to Humans
A biologically plausible mechanism has been hypothesized for trichloroethylene-induced systemic autoimmunity and autoimmune hepatitis that involves the bioactivation of trichloroethylene to chloral in genetically susceptible mice. This mechanism might explain clonally restricted diseases, such as autoimmune-induced hepatitis, but does not explain polyclonal diseases. Chloral has been shown to bind to circulating proteins leading to an alteration in self that resulted in autoantibody formation and a chemically induced autoimmune syndrome. It is unclear whether this mechanism exists in humans, although it is notable that people with polymorphisms
in CYP2E1 appear to have a higher risk of solvent-induced autoimmune disease. More studies are needed to incorporate aspects of innate and adaptive immune responses to study this and other proposed mechanisms of trichloroethylene-induced autoimmune diseases in humans.
Animal studies indicate that chronic exposure to trichloroethylene at moderate to high concentrations might have to potential to produce immunosuppression in animal models. However, there is no evidence to suggest that trichloroethylene is immunosuppressive in humans.
A potential biomarker for trichloroethylene exposure has been identified in mouse studies—namely, a chloral-protein adduct has been detected in tissues of trichloroethylene-treated mice (Griffin et al. 2000c). The usefulness of this marker in serum has not been demonstrated in humans.
No general statements can be made about the susceptibility of rodent and nonrodent species to trichloroethylene compared with that in humans. It is important to evaluate potential mechanisms of immunotoxicity to determine whether those mechanisms are operative in humans. Pharmacokinetic modeling of specific metabolites of trichloroethylene is important to consider and apply to specific mechanisms that may be responsible for immunotoxicity.
Little is known about the genes that determine susceptibility to autoimmune and other immune diseases in humans. Although it is likely that environmental xenobiotics can act as triggers or exacerbants of autoimmune disease, there have not been adequate studies to make strong correlations. In addition, it is likely that multiple genes control susceptibility, some of which may play more important roles than others. There may be significant differences between rodent species and humans. It is important to explore genetic susceptibilities in human and animal models. Genes controlling metabolism and pharmacokinetic behavior of trichloroethylene likely are polymorphic and may influence effects on the immune system.
FINDINGS AND RECOMMENDATIONS
Laboratory results consistently show that some strains of mice are sensitive to autoimmune disease induction or exacerbation after exposure to trichloroethylene. The dose of trichloroethylene required to produce effects depends on the route and length of exposure. It is difficult to extrapolate the results from genetically prone mice to humans, but these results suggest that humans with genetic susceptibilities may be at increased risk for autoimmune disease after exposure to by trichloroethylene. Animal data support the concept that trichloroethylene-mediated exacerbation of autoimmunity
may be due to metabolism of trichloroethylene to chloral, which is a biologically plausible mechanism for humans.
More animal research is needed to clarify the metabolites and modes of action responsible for trichloroethylene-induced autoimmunity and immunosuppression. Epidemiology studies should further examine connective tissue diseases and other autoimmune diseases (including Stevens-Johnson Syndrome) or immunologic alterations (e.g., changes in T cell subsets, incidence of autoantibodies) in populations exposed to trichloroethylene.
Results from mouse studies suggest that choral forms protein adducts that lead to an alteration of self proteins and the production of autoantibodies. CYP2E1, a known human polymorphic enzyme, may play a role in the formation of these protein adducts (Griffin 2000c). Inhibition of CYP2E1 was found to decrease the incidence and severity of autoimmune diseases in mice. Therefore, genetic polymorphisms in CYP2E1 may play a role in exacerbating autoimmune disease.
Genetic polymorphisms that may play a role in the metabolism of trichloroethylene should be further examined to determine sensitivity factors and to characterize potentially sensitive populations.