Exposures to trichloroethylene (TCE) induce several types of cancer in expe
rimental animals. TCE is metabolized in the body to multiple compounds, sev
eral of which are carcinogenic at relatively high doses in rodents. Risk as
sessments can be pursued with any one of the animal cancer endpoints after
consideration of their relevance for humans exposed at low environmental co
ncentrations. Among the TCE metabolites, dichloroacetic acid (DCA), trichlo
roacetic acid (TCA), chloral hydrate (CHL), and 1,2-dichlorovinylcysteine (
DCVC) are each considered carcinogenic in rodents. Two of these, DCA and CH
L, are used therapeutically in humans, while a third, TCA, has metabolic ef
fects similar to compounds used therapeutically in humans to reduce lipid c
oncentrations in blood. Each of these three compounds produces biological r
esponses that are expected in humans. However, these biological responses,
although they serve as precursors for tumor formation in rodents, are not e
xpected to lead to tumors in humans at any environmentally relevant exposur
e situations. 1,2-DCVC, formed by a minor pathway of TCE metabolism, is fur
ther metabolized in kidney to a reactive thioketene. Of the observed animal
carcinogenic responses, only the kidney tumors from the DCVC pathway are c
onsidered relevant as predictors of human cancer. Factors that increase TCE
metabolism to glutathione conjugates or predispose humans to kidney damage
should increase risks posed to workers from high-concentration exposures t
o TCE. Based on our knowledge of the cytotoxic and mutagenic modes of actio
n expected for these active metabolite(s), a nonlinear cancer risk assessme
nt approach is recommended to be used with the kidney tumor data to assess
the human risks of TCE.