Applications of computational toxicology methods at the Agency for Toxic Substances and Disease Registry.

In its efforts to provide consultations to state and local health departments, other federal agencies, health professionals, and the public on the health effects of environmental pollutants, the Agency for Toxic Substances and Disease Registry relies on the latest advances in computational toxicology. The computational toxicology laboratory at the agency is continually engaged in developing and applying models for decision-support tools such as physiologically based pharmacokinetic (PBPK) models, benchmark dose (BMD) models, and quantitative structure-activity relationship (QSAR) models. PBPK models are suitable for connecting exposure scenarios to biological indicators such as tissue dose or end point response. The models are used by the agency to identify the significance of exposure routes in producing tissue levels of possible contaminants for people living near hazardous waste sites. Additionally, PBPK models provide a credible scientific methodology for route-to-route extrapolations of health guidance values, which are usually determined from a very specific set of experiments. Also, scientists at the computational toxicology laboratory are using PBPK models for advancing toxicology research in such areas as joint toxicity assessment and child-based toxicity assessments. With BMD modeling, all the information embedded in an experimentally determined dose-response relationship is used to estimate, with minimum extrapolations, human health guidance values for environmental substances. Scientists in the laboratory also rely on QSAR models in the many cases where consultations from the agency are reported for chemicals that lack adequate experimental documentation.