L. Görlitz, K. Coboeken
Computational Systems Biology, Bayer Technology Services GmbH, Leverkusen, Germany
Objectives: The assessment of inter-individual variability is a key aspect in PK modeling which is usually performed using compartmental models combined with covariate mixed effect models. They are not capable of identifying physiological sources of variation. Physiology-based pharmacokinetic (PBPK) models enable a comprehensive simulation of drug pharmacokinetics at the whole-body scale based on drug distribution models and extensive collections of physiological parameters. By integration of specific experimental data, models are used to analyze and represent the key processes leading to the observed behavior and to investigate the expected pharmacokinetics in groups of healthy volunteers or patients by processing population simulations. A combination of nonlinear mixed-effects modeling and PBPK would allow identification of main sources of inter-individual variability.
Methods: We apply a PBPK-model of Theophylline, an anti-asthmatic drug, built in PK-Sim/MoBi [1] combined with a nonlinear-mixed effects model [2] written in R [3] to the data described in [4]. The data contain urine and plasma samples of 8 healthy male volunteers after PO-administration of 185 mg and after IV-administration of 208 mg Theophylline. Each individuals’ PBPK-model contains BMI- and age-adjusted organ volumes and blood flows derived from PK-Sims built in database. Starting with a no-random-effects model the NLME-PBPK is iteratively enriched by adding random effects found after visual inspection of the residual plots. The final model is identified if no further random effect can be added to the model (based on p-values).
Results: The approach is capable of identifying a model nicely representing the Theophylline plasma and urine concentrations. Inter-individual variation can be attributed to different glomerular filtration rates, different degrees of enterohepatic cycling, metabolization in the liver and absorption of the PO-administered drug in the GI-tract.
Conclusions:We showed that combining PBPK-modeling with nonlinear mixed effects approaches allows the identification of sources of inter-individual variation. Although generating comparable results on the population variation level, only this approach allows the explanation of this variation which is a highly valuable information for study planning as it e.g. provides hypotheses for population stratification.
References:
[1] Willmann S, Lippert J, Sevestre M, Solodenko J, Fois F, Schmitt W. (2003) PK-Sim: A physiologically based pharmacokinetic ‘whole-body’ model. Biosilico. 1(4):121-24.
[2] Pinhero J, Bates D. (2000) Mixed Effects Models in S and S-Plus, Springer.
[3] R Development Core Team (2011). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/.
[4] Kaumeier HS et al. Arzneimittelforschung. 1984;34(1):92-5
Reference: PAGE 21 (2012) Abstr 2554 [www.page-meeting.org/?abstract=2554]
Poster: New Modelling Approaches