JPR Monteleone and NHG Holford
Department of Pharmacology & Clinical Pharmacology, School of Medicine, University of Auckland, Private bag 92019, Auckland, New Zealand.
ABSTRACT:
The presence of interoccasion variability (IOV) in toxicokinetic parameters was investigated using the approach of Karlsson and Sheiner (1).
Data gathered from an oral, 4 week, toxicological study of RO 40-8757 in dogs was analyzed using MKMODEL and NONMEM. Analysis of individual subject data, using MKMODEL, revealed large variations in the concentration profile from treatment occasion to occasion suggesting IOV. A two-compartment, first order input model was selected for subsequent population analysis.
The presence of IOV was established by examining the change in objective function value (OBJ) when subject data was treated in two ways: each occasion for each subject treated as if from a different individual (SEP), or each occasion for each subject treated as if from the same individual (POOL).
An improved OBJ in the SEP vs POOL data treatment served as a diagnostic for the presence of IOV. An OBJ change of 70 together with observed large concentration profile variations indicated the presence of IOV.
IOV for relative bioavailability, clearance or central compartment volume using the POOL data was modeled with NONMEM. A model without IOV was used as a reference for comparing the changes attributed to IOV. All three IOV models resulted in improved OBJs compared to the reference model.
Another study of RO 40-8757 has suggested bioavailability as a primary source of IOV attributed to feeding times in relation to dosing. However, the model of IOV in relative bioavailability did not give the best fit; instead, incorporating IOV in central compartment volume was a better model. These findings suggest a more complex origin for IOV.
The presence of IOV raises difficulties for predicting delayed toxicodynamic effects with models that need identical toxicokinetic parameters on each occasion.
REFERENCES:
1. M.O. Karlsson and L.B. Sheiner. Estimating bioavailability when clearance varies with time. Clin. Pharmacol. Ther. 55:623-637 (1994).