I. Nestorov, L. Aarons, M. Rowland
Centre for Applied Pharmacokinetic Research, School of Pharmacy & Pharmaceutical Sciences, Manchester University,Manchester M13 9PL, UK
Based on a set of experimental data, including the temporal profiles of a homologous series of 9 barbituric acids in 14 rat tissues and arterial plasma after i.v. administration, we developed a global model for the barbiturate series, consisting of: (i) an 18- compartment whole body physiologically based pharmacokinetic (PBPK) model for the tissue distribution of the compounds; and (ii) a mechanistically based model, describing the relationship between the drug-dependent parameters of the PBPK model (tissue-to-unbound plasma distribution ratios and clearances) and the LogP (octanol-to-water partition coefficient) values of the congeners. As, across this homologous series, the other physicochemical properties of the drugs are relatively constant, this model views lipophilicity, expressed in terms of LogP, as the covariate for drug distribution. The parameter values of the global model were optimised using NONMEM from the pooled experimental data for all tissues and all homologues. We encountered a number of practical problems, connected with both the large dimensionality of the model and variability of the experimental data. Prior to the NONMEM analysis, the original 18 compartment model had to be reduced to a 6 compartment one by an appropriate lumping of the tissue compartments, displaying similar kinetic properties. The nature of the problems and the solutions, proposed by us, are discussed.
Reference: PAGE 7 (1998) Abstr 289 [www.page-meeting.org/?abstract=289]
Poster: poster