Phenotype–Genotype Correlation Applying a Cocktail Approach and an Exome Chip Analysis Reveals Further Variants Contributing to Variation of Drug Metabolism

Abstract

Although great progress has been made in the fine-tuning of diplotypes, there is still a need to further improve the predictability of individual phenotypes of pharmacogenetically relevant enzymes. The aim of this study was to analyze the additional contribution of sex and variants identified by exome chip analysis to the metabolic ratio of five probe drugs. A cocktail study applying dextromethorphan, losartan, omeprazole, midazolam, and caffeine was conducted on 200 healthy volunteers. CYP2D6, 2C9, 2C19, 3A4/5, and 1A2 genotypes were analyzed and correlated with metabolic ratios. In addition, an exome chip analysis was performed. These SNPs correlating with metabolic ratios were confirmed by individual genotyping. The contribution of various factors to metabolic ratios was assessed by multiple regression analysis. Genotypically predicted phenotypes defined by CPIC discriminated very well the log metabolic ratios with the exception of caffeine. There were minor sex differences in the activity of CYP2C9, 2C19, 1A2, and CYP3A4/5. For dextromethorphan (CYP2D6), IP6K2 (rs61740999) and TCF20 (rs5758651) affected metabolic ratios, but only IP6K2 remained significant after multiple regression analysis. For losartan (CYP2C9), FBXW12 (rs17080138), ZNF703 (rs79707182), and SLC17A4 (rs11754288) together with CYP diplotypes, and sex explained 50% of interindividual variability. For omeprazole (CYP2C19), no significant influence of CYP2C:TG haplotypes was observed, but CYP2C19 rs12777823 improved the predictability. The comprehensive genetic analysis and inclusion of sex in a multiple regression model significantly improved the explanation of variability of metabolic ratios, resulting in further improvement of algorithms for the prediction of individual phenotypes of drug-metabolizing enzymes.