Abstract
Background Pharmacogenetics has the potential to optimise drug therapy and reduce adverse drug effects (ADEs) by tailoring treatment to a patient’s genotype, particularly for chronic disorders managed in general practice (GP). However, the adoption of pharmacogenetics in GP remains slow. Aim This study aimed to evaluate the reproducibility of previously reported associations between genomic variants and medically important adverse drug effects (MIADEs) associated with high-risk medications in GP. Design and setting A retrospective study using data from the UK Biobank (UKBB), a population-based cohort of over 500,000 community-based participants. Method We identified high-risk medications prescribed in GP by linking serious ADEs from the Yellow Card database with English GP prescription data. These high-risk medications were then cross-examined with genomic variants associated with MIADEs from the Pharmacogenomics Knowledgebase (PharmGKB), to select variant-drug pairs for investigation within the UKBB. Results From 78 high-risk medications prescribed in GP and 56 PharmGKB annotations linked to MIADE risk, SLCO1B1 rs4149056 was the only variant with guideline-based prescribing recommendations. This variant, along with others of lower evidence levels, was analysed in the UKBB. No genotype-treatment interaction was observed for SLCO1B1 rs4149056 and statin-related muscle toxicity. Similarly, no interactions were detected for the remaining variants in either secondary or exploratory analyses. Conclusion No statistically significant genotype-treatment interactions were observed for MIADE risk associated with high-risk medications in GP. However, the limited predictive value of the assessed variants may reflect underlying phenotypic imprecision and methodological limitations. Hence, further research is needed to validate these results.