(693a) Pharmacokinetic Parameter Estimation of Drug Distribution in An Entire Organism

Before clinical trials, routine drug dose response experiments in animals provide invaluable information about the pharmacokinetics and pharmacodynamics of new therapeutic agents. The drug response curves in various organs such as the liver, heart, brain or lungs hold quantitative information to ascertain the laws of drug fate in the living organism. Yet, rigorous analyses of drug response curves are not widely practiced in the pharmaceutical industry.

Therefore, this paper proposes a computer-based methodology for the quantitative assessment of physiologically-based pharmacokinetics (PBPK) in a global model of an organism such as a rat. The application concerns preclinical stages of new drug design when subtherapeutic, therapeutic and supertherapeutic doses need to be determined. Currently, dose estimation is typically performed on animal species during preclinical trials; interspecies scaling is then done by extapolation with large uncertainties about the global transport mechanisms, the clearances and mass balance errors. Our proposed methods use a detailed vasculature network to compute steady state blood perfusion through each organ based on experimental flow resistances and mass balances. Detailed information about tissue vascular volume and interstitial space can be incorporated into the model from medical images using image reconstruction tools. Allometric scaling laws allow to account for differences among individuals of a species based on weight. In addition, changes related to age or pathological conditions such as obesity or diabetes can be accounted for. Quantitative interspecies scaling can also be easily modeled as the system is based on first principles equations and is independent of the PBPK model choice. The modeler can adjust the system complexity by either selecting a global model or a set of organ-specific models. For parameter estimation, a set of experimentally measured concentration profiles is used as input into each subsystem, such as blood, plasma and tissue. The solution for the estimation problem using mathematical programming techniques yields the desired set of kinetic parameters together with the best-fit drug-time concentration profiles. This method also allows for automatic evaluation of model quality based on the least squares minimization to the experimental data set. Different PBPK models can be compared, thus helping in selection of proper kinetics.

In conclusion, this article presents a method for semi-automated physiologically-based pharmacokinetics model selection and parameter estimation based on experimental datasets of drug delivery. The results include time-dependent drug concentrations in each organ and provide valuable insights into the mechanics of drug transport in a study organism. The successful application of the proposed methods will lead to better design of preclinical trials, more knowledge gain from preclinical animal experimentation and eventually lead to shorter drug development times.