Pharmacogenomics of Medically Important Adverse Drug Effects

Abstract

Introduction: Medicines with high toxicity profiles have a heightened risk of causing serious and fatal adverse drug effects (ADEs). General Practice (GP) is key in identifying and potentially preventing ADEs. While the use of genomic information has the potential to reduce ADEs, the robustness and reproducibility of genetic research findings are questionable. Hence, separating true positives from false positives and minimising the overabundance of false-positive signals is vital.

Aim: To assess the current state of the art in pharmacogenomics (PGx) of adverse drug effects and analyse whether variants previously reported to be associated with medically important adverse effects (MIADEs) replicate in the UK Biobank (UKBB).

Methods and Materials :Three separate systematic reviews of the literature were conducted to identify relevant studies. To identify high-risk medicines, data on serious and fatal ADEs from the UK pharmacovigilance was mapped onto GP prescription data in England. Previously described associations between variants and MIADEs related to high-risk medicines in GP and endocrine drugs for breast cancer were interrogated in UKBB.

Results: I created a list of variants associated with ADEs and further generated a set of variant–drug pairs significantly associated with MIADEs. I identified medicines with high toxicity profiles in GP and created comparative safety charts to support evidence-based decision-making around formulary choices. No statistically significant genotype-treatment interactions were found for either baseline measurements or incident MIADEs in UKBB. This included MIADEs related to statins, NSAIDs, antipsychotics and endocrine therapy.

Conclusions: None of the PGx findings tested were replicated in UKBB. This included associations between variants and MIADEs related to high-risk medicines in GP and endocrine agents, in relation to neither baseline measurements nor incident MIADEs. These variants are not accurate at identifying those who are at risk of developing MIADEs in patients receiving these treatments and therefore should not be considered for personalised recommendations.