Influence of Enzyme Modulation on Drug Metabolism and Pharmacokinetics: Implications for Therapeutic Safety and Personalized Treatment
TL;DR
This virtual-population PK simulation shows that the enzyme inhibition/induction fenomenological model and/or the genetic-type defined phenotype are a unique and important contributor for the simulated drug exposure and can be incorporated in personalised prescribing by using enzyme-modulation-aware dosing algorithms, or by using genotype-informed dosing algorithms.
Abstract
Background: Variability in drug clearance mediated by drug-enzyme interaction can happen through three pathways, resulting in variable drug exposures; pharmacological inhibition, pharmacological induction and metaboliser phenotype, which takes a genetic approach, and these pathways are not often analysed in a single reproducible modelling approach. The goal of this study is to perform a population pharmacokinetic simulation study to quantify the impact of phenotyping of metaboliser (genotype of 2D6) on overall systemic drug exposure along with the effects of CYP3A4 enzyme inhibition and induction on the overall systemic exposure in a virtual patient population.Methods: A single-compartment population pharmacokinetic model was moved over following the allometric principles and then 2 independent virtual populations of 900 virtual adult subjects (comprising 300 patients on each treatment arm) were simulated with 1000 patients, each with a fixed dose of the simulated model CYP2D6 substrate, and randomly distributed in silico metaboliser phenotype according to the published in vivo CYP2D6 metaboliser phenotype population frequencies. Independent t-tests, geometric mean ratios (GMR) and perceived 90% CI, Kruskal-Wallis test were used to evaluate group differences in log transformed AUC0–∞.Results: The administration of the simulated strong CYP3A4 inhibitor raised the AUC0–∞ by a GMR of 4.98 (90% CI 4.78–5.19; p<0.0001); the strong inducer reduced the AUC0–∞ to a GMR of 0.31 (90% CI 0.29–0.32; p<0.0001). The geometric mean AUC0–∞ was significantly higher in the Pm phenotype compared with the other phenotypes (519 ng·h/mL for ultrarapid metabolisers vs 8,873 ng·h/mL for poor metabolisers [H=420.9, df=3, p<0.0001]); and all PMs were above a pre-specified supratherapeutic threshold of exposure, whereas none UMs were.Conclusions: This virtual-population PK simulation shows that the enzyme inhibition/induction fenomenological model and/or the genetic-type defined phenotype are a unique and important contributor for the simulated drug exposure and can be incorporated in personalised prescribing by using enzyme-modulation-aware dosing algorithms, or by using genotype-informed dosing algorithms.