You are here:
AN EXAMPLE OF MODEL STRUCTURE DIFFERENCES USING SENSITIVITY ANALYSES IN PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS OF TRICHLOROETHYLENE IN HUMANS
Citation:
YOKLEY, K. AND M. V. EVANS. AN EXAMPLE OF MODEL STRUCTURE DIFFERENCES USING SENSITIVITY ANALYSES IN PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS OF TRICHLOROETHYLENE IN HUMANS. BULLETIN OF MATHEMATICAL BIOLOGY. Elsevier Science Ltd, New York, NY, 69(8):2591-625, (2007).
Impact/Purpose:
The idea for this study came about from the comparison of two different PBPK models for TCE that are available in the published literature. One model simplifies the concept of metabolite distribution, while the other makes use of full PBPK models describing metabolite distribution. Sensitivity analysis was then used to illustrate the importance of making appropriate decisions about model structure based on how much information is available around the model parameters.
Description:
Abstract Trichloroethylene (TCE) is an industrial chemical and an environmental contaminant. TCE and its metabolites may be carcinogenic and affect human health. Physiologically based pharmacokinetic (PBPK) models that differ in compartmentalization are developed for TCE metabolism in humans, and the focus of this investigation is to evaluate alternative models. The two models fomulated differ in the compartmentalization of metabolites; more specifically, one model has compartments for all chemicals and the other model has only a generalized body compartment for each the metabolites and contains multiple compartments for the parent,TCE. The models are compared through sensitivity analyses in order to selectively discriminate with regards to model structure. Sensitivities to a parameter of cardiac output (Qcc) are calculated, and the more compartmentalized model predictions for excretion show lower sensitivity to changes in this parameter. Since information about cardiac output across a population is not often incorporated into a PBPK model, the more compartmentalized (“unlumped”) model is probably a more ap- propriate mathematical description of TCE metabolism, but further study may be necessary to decide which model is a more reasonable option if distributional information about Qcc is used. The study is intended to illustrate how sensitivity analysis can be used in order to make appropriate decisions about model development when considering physiological parameters than vary across the population.
