Prediction models for voriconazole pharmacokinetics based on pharmacogenetics: AN exploratory study in a Spanish population

Abstract

Individualisation of the therapeutic strategy for the oral antifungal agent voriconazole (VCZ) is extremely important for treatment optimisation. To date, regulatory agencies include CYP2C19 as the only major pharmacogenetic (PGx) biomarker in their dosing guidelines; however, the effect of other genes might be important for VCZ dosing prediction. We developed an exploratory PGx study to identify new biomarkers related to VCZ pharmacokinetics. We first designed a 'clinical practice VCZ-AUC prediction model' based on CYP2C19 to be used as a reference model in this study. We then designed a multifactorial polygenic prediction model and found that genetic variability in FMO3, NR1I2, POR, CYP2C9 and CYP3A4 partially contributes to VCZ total area under the concentration-time curve (AUC_0-∞) interindividual variability, and its inclusion in VCZ AUC_0-∞ prediction algorithms improves model precision. To our knowledge, there are no PGx studies specifically relating POR, FMO3 and NR1I2 polymorphisms to VCZ pharmacokinetic variability. Further research is needed in order to test the model proposed here.

Keywords: AUC prediction model; CYP2C19; Pharmacogenetics; Pharmacokinetics; Voriconazole.