Simulation Studies Examining Possible Mechanisms of Trichloroethylene Carcinogenicity. R. C. Lee, E. G. Luebeck, S. M. Bartell, and E. M. Faustman, Univ. of Washington Dept. Environmental Health, Fred Hutchinson Cancer Research Center, Seattle, WA
There has been much recent interest regarding methods of evaluating potential human cancer risk associated with trichloroethylene (TCE). The mechanisms of TCE toxicity that may be important to human risk are highly uncertain. Stochastic biologically based dose-response (BBDR) models offer possible means of incorporating cellular-level mechanistic considerations into TCE risk assessment; thus reducing uncertainty. This paper examines the sensitivity of the Moolgavkar-Venzon-Knudson (MVK) 2-stage model to variations in model parameters, and the ability of the model to distinguish between initiating and promoting activity of TCE. Maximum-likelihood estimation is used to fit parameters to simulated data sets assuming different carcinogenic mechanisms and low-dose response curves for TCE. Monte Carlo simulations are used to simulate experimental variability under several mechanistic assumptions. Simulations indicate that a largely promotional mechanism provides the most plausible model parameter estimates. Assuming that net cell proliferation rate follows a quadratic dose response, modeling excess risks as linear results in overestimation of risk. Changes in experimental design are evaluated that allow carcinogenic mechanisms to be more clearly distinguished within the model. Preliminary work examines the impact of toxicokinetic assumptions on model results. These analyses provide information regarding uncertainties associated with carcinogenicity of TCE, and provide possible guidance for laboratory-based toxicological evaluations.