Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability
Citation:
Ring, C., R. Pearce, Woodrow Setzer, B. Wetmore, AND J. Wambaugh. Identifying populations sensitive to environmental chemicals by simulating toxicokinetic variability. ENVIRONMENT INTERNATIONAL. Elsevier Science Ltd, New York, NY, 106:105-118, (2017).
Impact/Purpose:
• Agency Research Drivers - The Frank R. Lautenberg Chemical Safety for the 21st Century Act directs the Agency to consider exposure when prioritizing chemicals for risk evaluations. To meet this Congressional mandate, the EPA needs methods to estimate levels of environmental chemicals and predict internal concentrations in humans following exposure. • Science Challenge – Use of high-throughput assays for Agency regulatory decisions is limited by uncertainties in estimating exposure dose. To address this limitation, data and tools are necessary to compare exposure model predictions with bioactive concentrations from high-throughput bioassays. • Research Approach – This research refined previous methods to extrapolate from in vitro data to human exposures, with a focus on more realistic human exposure pathways. The research developed new high throughput PBPK models, a framework for population-based AOP simulation, and methods for characterizing uncertainty of these models for risk estimation. • Results – In vitro toxicokinetic (TK) data allow the prediction of risk from high throughput chemical toxicity screening. The CDC NHANES provides biometric and chemical exposure data for the modern U.S. population. We examined how demographic differences in TK and exposure change risk, including for children and women of reproductive age. We find that the elderly in particular have slightly increased risk due to lower clearance per kilogram bodyweight. • Anticipated Impact/Expected use – The tools and methods developed allow extrapolation of data on the toxicokinetics for environmental chemicals. These approaches enable exposure estimations for thousands of chemicals, addressing a critical Agency need for pre-prioritization under Frank R. Lautenberg Chemical Safety for the 21st Century Act.
Description:
We incorporate inter-individual variability, including variability across demographic subgroups, into an open-source high-throughput (HT) toxicokinetics (TK) modeling framework for use in a next-generation risk prioritization approach. Risk prioritization involves rapid triage of thousands of environmental chemicals, most which have little or no existing TK data. Chemicals are prioritized based on model estimates of hazard and exposure, to decide which chemicals should be first in line for further study. Hazard may be estimated with in vitro HT screening assays, e.g., U.S. EPA’s ToxCast program. Bioactive ToxCast concentrations can be extrapolated to doses that produce equivalent concentrations in body tissues using a reverse dosimetry approach in which generic TK models are parameterized with 1) chemical-specific parameters derived from in vitro measurements and predicted from chemical structure; and 2) with physiological parameters for a virtual population. We have developed HTTK-Pop, a software package to simulate population physiological parameters based on the most recent CDC NHANES data on distributions of demographic and anthropometric quantities in the modern U.S. population. HTTK-Pop implements a Monte Carlo approach, accounting for the correlation structure in physiological parameters, which is used to estimate ToxCast oral equivalent doses for the most sensitive portion of the population. For risk prioritization, oral equivalent doses are compared to estimates of exposure rates based on NHANES urinary analyte biomonitoring data. The inclusion of inter-individual variability in the TK modeling framework allows targeted risk prioritization for demographic groups of interest, including potentially sensitive life stages and subpopulations.
