NAS Report on Cancer Risks from Trichloroethylene (TCE)
27 Jul 2006 in Regulatory Science, Regulatory Policy, Information Quality, Peer Review
On July 27, the National Research Council of the National Academies released a report concluding that while additional research should be performed, enough information exists for the Environmental Protection Agency to complete its risk assessment of trichloroethylene. EPA's 2001 draft risk assessment was highly controversial. Similar to what has happened after EPA published other controversial draft risk assessments, several other federal agencies joined with EPA to sponsor the NRC review (in this case, the Departments of Defense and Energy and NASA).
The table below summarizes the committee's conclusions regarding several controversial toxicological issues. We rely on text in the committee's Report Brief:
| Issue | NAS Committee Findings |
| Kidney Toxicity and Cancer | "Trichloroethylene and some of its metabolites in the glutathione-conjugation pathway have been shown to be both toxic and carcinogenic to the kidneys. There is concordance between animal and human studies, which supports the conclusion that trichloroethylene is a potential kidney carcinogen. Studies with experimental animals and human tissues indicate a genotoxic mode of action. The metabolite S-dichlorovinyl-L-cysteine has been linked with the development of kidney cancer, but there are no studies of the carcinogenic potential of this metabolite. The magnitude of exposure needed to produce kidney damage is not clear. Thus, it is not possible to predict whether humans are more or less susceptible than other animals to trichloroethylene induced kidney cancer." |
| Liver Toxicity and Cancer | "The epidemiologic evidence is mixed; some studies show an excess of liver cancer in trichloroethylene exposed populations while other studies do not. Animal data on trichloroethylene indicate that relatively high doses are needed to induce liver toxicity and cancer, even in susceptible strains of mice. Three major oxidative metabolites (trichloroacetic acid, dichloroacetic acid, and chloral hydrate) can contribute to liver toxicity and cancer in rodents. The mode of action of trichloroacetic acid as a rodent liver carcinogen is not a likely mode of action in the human liver. For the metabolite chloral hydrate, differing rates of oxidation and conjugation in rats and humans make it unlikely that the mode of action in mice is relevant to humans. The mode of action for the metabolite dichloroacetic acid in rodents is understood, but whether this metabolite is formed in humans has not been established and differences between mice and human suggest that humans would be much less susceptible to liver carcinogenesis. Thus, exposure to trichloroethylene at concentrations relevant to the general public is not likely to induce liver cancer in humans. However, it is possible that much higher exposure to trichloroethylene, such as in certain high-risk occupations or in heavily contaminated locales, could result in increased risks of liver toxicity and cancer." |
| Reproductive and Developmental Toxicity | "Evidence from animal and epidemiologic studies suggest that exposure to trichloroethylene and one or more of its metabolites might be associated with congenital heart defects. Although there are inconsistencies in the animal data, plausibility for trichloroethylene-induced cardiac teratogenesis is increased by the fact that the most frequently observed cardiac defects in human studies are consistent with those found in animal studies. Research in animals and humans also indicates that trichloroethylene impairs intrauterine growth. However, the specific metabolites involved and the mode of action responsible for cardiac teratogenesis and poor intrauterine growth remain to be elucidated. Rodent studies also show that trichloroethylene can affect fertility in males (reduced spermatogenesis) and females (decreased fertility of oocytes), but the relevance of these findings to humans is not clear." |
| Neurotoxicity | "Studies show that inhalation of trichloroethylene causes neurotoxic effects in laboratory animals and humans that are similar in nature (e.g., massiter reflex latency, motor incoordination, changes in heart rate) and occur at comparable concentrations of exposure. It has been suggested that exposure to trichloroethylene during early development could enhance its effects on the nervous system, but the available data are insufficient to draw firm conclusions. Some studies suggest a contribution of trichloroethylene to Parkinson’s disease. Multiple mechanisms appear to contribute to the neurotoxic action of trichloroethylene, and further study is needed to elucidate them more precisely." |
| Respiratory Toxicity and Cancer | "Trichloroethylene has been shown to induce lung tumors in rodents. The mode of action for this effect is localization of trichloroethylene metabolites in the Clara cells of the lungs. The collective evidence indicates that rodents and humans are significantly different in their capacity to metabolize trichloroethylene in the lungs, with humans having less capacity. Results of most epidemiologic studies of occupational exposure to trichloroethylene do not show a strong association between trichloroethylene exposure and increased incidence of lung tumors. Thus, pulmonary cancer does not appear to be a critical end point in assessing human health risks to trichloroethylene." |
| Immunotoxicity | "Studies in genetically susceptible rodents have shown that trichloroethylene exacerbates underlying autoimmune disease, and supporting information comes from multiple human studies of scleroderma and exposures to organic solvents. Some individuals might be genetically susceptible to developing autoimmune disease. The metabolites and the mode of action involved have not been elucidated, but a role for chloral has been implicated in mouse models." |
The human cancer risk estimates in the EPA's 2001 draft risk assessment depend heavily on a meta-analysis of the available epidemiologic studies. EPA relied on the study by Wartenburg et al. (concluding that kidney cancer is associated with TCE exposure). A competing meta-analysis not used by EPA was performed by Kelsh et al. (concluding that no such association exists).
The committee appears to have dismissed both:
There are two available meta-analyses, one developed by Wartenberg et al., whose analysis EPA used in its draft health risk assessment, and another by Kelsh et al. The committee found several weaknesses in the techniques used in both analyses. Problems included the use of subjective, tiered systems to classify and weigh studies, separate analyses of case-control and cohort studies, and the fact that these analyses did not consider identifying amounts of exposure in the studies. The report recommends that a new meta-analysis be developed to support a human health risk assessment.
NRC was asked to "examine issues critical for developing an objective, realistic, scientifically based health risk assessment" (Full Report at 1) These descriptors originate in a 2001 policy directive issued by the Office of Management and Budget::
[R]isk assessment should be an objective, realistic, and scientifically balanced analysis.
In at least one very important respect, however, the committee failed to fulfill this charge. The committee analyzed scientific information through the filter of EPA cancer risk assessment policies that are specifically designed and intended not to be objective, but be "public-health protective." EPA's 2004 Staff Paper on Risk Assessment Principles and Practices declares:
[S]since EPA is a health and environmental protective agency, EPA’s policy is that risk assessments should not knowingly underestimate or grossly overestimate risks. This policy position prompts risk assessments to take a more “protective” stance given the underlying uncertainty with the risk estimates generated. (p. 13)
The committee "was not asked to perform a risk assessment or to address risk management issues" (p. 2). However, it did so implicitly by forcing its recommendations to adhere to EPA's risk assessment policies, which contain important (and sometimes crucial) risk management judgments.
