Leiden University Scholarly Publications

Introduction: Pharmacogenetics-informed drug prescribing is increasingly applied in clinical practice. Typically, drug metabolizing phenotypes are determined based on genetic test results, whereupon dosage or drugs are adjusted. Drug-drug-interactions (DDIs) caused by concomitant medication can however cause mismatches between predicted and observed phenotypes (phenoconversion). Here we investigated the impact of CYP2C19 genotype on the outcome of CYP2C19-dependent DDIs in human liver microsomes.

Methods: Liver samples from 40 patients were included, and genotyped for CYP2C19*2, *3 and *17 variants. S-mephenytoin metabolism in microsomal fractions was used as proxy for CYP2C19 activity, and concordance between genotype-predicted and observed CYP2C19 phenotype was examined. Individual microsomes were subsequently co-exposed to fluvoxamine, voriconazole, omeprazole or pantoprazole to simulate DDIs.

Results:<...

Introduction: Pharmacogenetics-informed drug prescribing is increasingly applied in clinical practice. Typically, drug metabolizing phenotypes are determined based on genetic test results, whereupon dosage or drugs are adjusted. Drug-drug-interactions (DDIs) caused by concomitant medication can however cause mismatches between predicted and observed phenotypes (phenoconversion). Here we investigated the impact of CYP2C19 genotype on the outcome of CYP2C19-dependent DDIs in human liver microsomes.

Methods: Liver samples from 40 patients were included, and genotyped for CYP2C19*2, *3 and *17 variants. S-mephenytoin metabolism in microsomal fractions was used as proxy for CYP2C19 activity, and concordance between genotype-predicted and observed CYP2C19 phenotype was examined. Individual microsomes were subsequently co-exposed to fluvoxamine, voriconazole, omeprazole or pantoprazole to simulate DDIs.

Results: Maximal CYP2C19 activity (V_max) in genotype-predicted intermediate metabolizers (IMs; *1/*2 or *2/*17), rapid metabolizers (RMs; *1/*17) and ultrarapid metabolizers (UMs; *17/*17) was not different from V_max of predicted normal metabolizers (NMs; *1/*1). Conversely, CYP2C19*2/*2 genotyped-donors exhibited V_max rates ∼9% of NMs, confirming the genotype-predicted poor metabolizer (PM) phenotype. Categorizing CYP2C19 activity, we found a 40% concordance between genetically-predicted CYP2C19 phenotypes and measured phenotypes, indicating substantial phenoconversion. Eight patients (20%) exhibited CYP2C19 IM/PM phenotypes that were not predicted by their CYP2C19 genotype, of which six could be linked to the presence of diabetes or liver disease. In subsequent DDI experiments, CYP2C19 activity was inhibited by omeprazole (−37% ± 8%), voriconazole (−59% ± 4%) and fluvoxamine (−85% ± 2%), but not by pantoprazole (−2 ± 4%). The strength of CYP2C19 inhibitors remained unaffected by CYP2C19 genotype, as similar percental declines in CYP2C19 activity and comparable metabolism-dependent inhibitory constants (K_inact/K_I) of omeprazole were observed between CYP2C19 genotypes. However, the consequences of CYP2C19 inhibitor-mediated phenoconversion were different between CYP2C19 genotypes. In example, voriconazole converted 50% of *1/*1 donors to a IM/PM phenotype, but only 14% of *1/*17 donors. Fluvoxamine converted all donors to phenotypic IMs/PMs, but *1/*17 (14%) were less likely to become PMs than *1/*1 (50%) or *1/*2 and *2/*17 (57%).

Conclusion: This study suggests that the differential outcome of CYP2C19-mediated DDIs between genotypes are primarily dictated by basal CYP2C19 activity, that may in part be predicted by CYP2C19 genotype but likely also depends on disease-related factors.