Evaluation of a basic PBPK model in preclinical species for which tissue compositions are unknown
Bizzotto R (1), Nucci G (2), Petrone M (2), Poggesi I (2), Cobelli C (1), Gomeni R (2)
(1) Department of Information Engineering, University of Padova, Padova, Italy; (2) Clinical Pharmacokinetics/Modeling&Simulation, GlaxoSmithKline, Verona, Italy
Objectives: We are currently proposing a Bayesian approach for the predictions of human pharmacokinetics based on PBPK [1], in which the tissue partition process is parameterized using a tissue composition model [2]. Since the Bayesian procedure was used to fine tune a few model parameters based on in vivo preclinical pharmacokinetic data, the parameterization of the tissue composition model in all animal species is required. In dogs and monkeys no tissue composition data are reported in the literature; the present evaluation aimed therefore at evaluating the performance of the tissue composition model in these species resorting to the data available in rats, mice, rabbits and humans.
Methods: A set of 22 GSK compounds given IV to dogs was considered. Data on blood flows and tissue volumes were available from the literature. Tissue composition data were alternatively assumed equal to those reported for rats, mice, rabbits and humans. The predictive performance of the different tissue composition models was evaluated considering volumes of distribution and clearance values.
Results: The available tissue composition data provided adequate predictions of the volume of distribution at steady state in dogs, with average fold-errors of 2.2, 2.1, 2.2 and 2.3 based on the tissue data of mice, rats, humans and rabbits, respectively. The corresponding percentage of compounds predicted within 2-fold was 59, 50, 45 and 59%, respectively. This performance did not appear substantially degraded as compared to that obtained in rats with the appropriate tissue composition (average fold-error 1.8, 74% of compounds predicted within 2-fold). The performance for the systemic clearance predictions (dependent on the tissue composition model only for the blood to plasma partition) was good (fold-error 1.8-2.0) and essentially identical to that observed in rats.
Conclusions: The application of our Bayesian approach for the predictions of human pharmacokinetics based on PBPK modelling encouraged us to consider also the in vivo pharmacokinetic data obtained in dogs and monkeys. The models were reparameterized with the tissue composition data reported for other animal species. The predictive performance in dogs was comparable to that achieved with the appropriate tissue composition. The performance in monkeys is currently under evaluation and will be presented during the poster session.
References:
[1] Bizzotto R, Nucci G, Poggesi I, Gomeni R. Integration of preclinical information into a PBPK approach for predicting human pharmacokinetics. PAGE 2007
[2] Poulin P & Theil FP. Prediction of Pharmacokinetics prior to In Vivo Studies. II. Generic Physiologically Based Pharmacokinetic Models of Drug Disposition. J. Pharm. Sci. 2002, 91:1358-70.