You are here:
USE OF SENSITIVITY ANALYSIS ON A PHYSIOLOGICALLY-BASED PHARMACOKINETIC (PBPK) MODEL FOR CHLOROFORM IN RATS TO DETERMINE AGE-RELATED TOXICITY
Eklund, C R., M V. Evans, AND J E. Simmons. USE OF SENSITIVITY ANALYSIS ON A PHYSIOLOGICALLY-BASED PHARMACOKINETIC (PBPK) MODEL FOR CHLOROFORM IN RATS TO DETERMINE AGE-RELATED TOXICITY. Presented at Society of Toxicology 42nd Annual Meeting, Salt Lake City, Utah, March 9-13, 2003.
USE OF SENSITIVITY ANALYSIS ON A PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODEL FOR CHLOROFORM IN RATS TO DETERMINE AGE-RELATED TOXICITY.
CR Eklund, MV Evans, and JE Simmons. US EPA, ORD, NHEERL, ETD,PKB, Research Triangle Park, NC.
Chloroform (CHCl3) is a disinfection byproduct for which several PBPK models are available. Since PBPK models account for physiological differences, they are well-suited to predict differences due to aging. We used a five compartment PBPK constant inhalation model for CHCl3 (Evans et al., 2002) to examine age differences and variation in input parameters. Model simulations were performed in MatLab? running on Red Hat? Linux. We calculated and compared sensitivity coefficients (SCs) for three model input parameters and four model outputs over three age groups (~ 2, 3, and 24 months) and three chamber concentrations, 10, 100 and 500 ppm for 6 hours. We examined the effect of changing Vmax on rate of amount of CHCl3 metabolized (RAML) and amount of CHCl3 in liver (AL), cardiac output (QCC) on venous CHCl3 blood concentration (CV), and fat:blood partition coefficient (PCfat) on amount of CHCl3 in fat (AF). At 10 ppm CHCl3, changes in the 4 output parameters examined (RAML, AL, CV, and AF) were consistently much less than the changes applied to the input parameters. In contrast, at 500 ppm CHCl3, SCs for CV were greater than 1 for all 3 ages. At 500 ppm, the SCs for AF were greater than 1 for adult and old rats, but not young rats. At 500 ppm, the Scs for RAML approached 1, and for AL were much less than 1. Changes in Vmax produced opposite changes in RAML and AL. Increasing Vmax increased RAML and decreased AL. As the absolute magnitude of a SC serves as a measure of the impact of a change in a given input parameter on a model output of interest, SC results can be used to gauge the usfefulness of further refinement of selected input parameters. Where SCs are large, increased parameter accuracy ma be needed; uniformly small SCs may indicate further refinement is not needed. This is an example of how modeling can suggest a reduction in the number of animals used. (This abstract does not necessarily reflect EPA policy.)