You are here:
Local and global sensitivity analysis of a physiologically based pharmacokinetic model for oral uptake of carbon tetrachloride in rats
Citation:
Williams, D., M. Evans, J. Bruckner, AND J. Simmons. Local and global sensitivity analysis of a physiologically based pharmacokinetic model for oral uptake of carbon tetrachloride in rats. Tutorial Workshop on Parameter Estimation for Biological Models, Raleigh, NC, July 25 - 28, 2018.
Impact/Purpose:
The overall goal of this work is to reduce uncertainty for the oral route of carbon tetrachloride modeling using a calibrated PBPK model for rat. This work is aimed at reducing uncertainty for chronic reference dose calculations using PBPK models calibrated with rat data.
Description:
Carbon tetrachloride (CCl4) is a persistent volatile compound that exhibits liver toxicity after exposure to high doses, and it has been prioritized among the initial 10 chemicals for evaluation under the Lautenberg Chemical Safety Act. Due to its persistence in the environment, the chemical has been detected in drinking water. Quantitative predictions for oral exposure have large uncertainty, particularly for chronic reference dose calculations. To reduce uncertainty, physiologically-based pharmacokinetic (PBPK) models are used to predict internal from external dose. PBPK models have also been applied to reduce uncertainty in rat to human extrapolations. Our rodent PBPK model uses a two-phase GI absorption model for two exposure scenarios: instantaneous bolus administration, and a delayed intragastric infusion over two hours. This model introduces four unknown parameters, each of which were either fit or optimized. Simulation results from this model matched well with experimental data, but because the model absorption parameters were unknown, we also established parameter identifiability and uncertainty. Using local sensitivity analysis, each of the four parameters were shown to be identifiable, and preliminary global sensitivity analysis results showed first and higher order effect contributions. Further parameter calibration would lead to greater model confidence, thus reducing quantitative uncertainty in risk assessment. (This abstract does not reflect US EPA policy).