Yingying Tian, Sebastian Frechen, Ahmed Abbas Suleiman, Uwe Fuhr
Department of Pharmacology, Clinical Pharmacology Unit, University Hospital of Cologne, Cologne, Germany
Background: Paroxetine exposure increases more than linear with increasing doses and with decreasing number of active CYP2D6 alleles, which precludes individualized dosing by simple methods.
Objectives: To develop a physiological-based pharmacokinetic (PBPK) model for paroxetine able to describe its nonlinear kinetics among different genotypes.
Methods: PK data were obtained from an open, randomized, single-dose, 2-period, crossover pharmacokinetic study with two different formulations of paroxetine involving 40 healthy volunteers. CYP2D6 genotyping was carried out for all subjects. Data were analyzed using nonlinear mixed-effects modeling implemented in NONMEM V7.3.0. First step was to develop a single compartment model incorporating both inter-occasional variability and CYP2D6-mediated clearance. The structural model was then refined and transformed to a PBPK model by the introduction of physiological compartments and parameters and by incorporation of the mechanism-based inhibition process.
Results: A conventional one-compartment PK model with additional compartments for liver and enzyme and combined error model was developed. Among the covariate relationships tested, weight was identified as a significant covariate on volume of distribution of the central compartment. The model produces estimates pooled clearances according to the expected phenotype of CYP2D6*1, *2, *35 and CYP2D6 *9, *41.The model showed some underprediction for high concentrations. When temporarily removing subjects carrying the less frequent CYP2D6 alleles *9, *35 or *41 present in this study, the model was essentially unchanged.
Conclusions: The current model may serve as a promising base for further refinements. In particular, splitting of the pooled clearances into the respective clearances per underlying allele as well as a better description of the mechanism-based inhibition process need to be addressed.
Reference: PAGE 24 (2015) Abstr 3477 [www.page-meeting.org/?abstract=3477]
Poster: Drug/Disease modeling - Absorption & PBPK