Challenges in genome-wide association analysis of drug-induced toxicity data from clinical trials

A key goal of cancer pharmacogenomics in the context of clinical trials is the identification of genetic loci, such as single-nucleotide polymorphisms or copy number variants, implicated in the development of drug induced adverse events. These findings may lead to identification of subsets of patients genetically predisposed to drug induced toxicities. Genome-wide Association Studies (GWAS) have provided a powerful and practical platform enabling researchers to extend their scope of discovery from a handful of candidate markers to millions of markers covering the human genome. Conducting the requisite analyses in the context of clinical trials poses both practical as well as methodological challenges. The main objective of this talk is to highlight some of these challenges in the context of toxicity GWAS studies conducted as correlate studies of Cancer and Leukemia Group B (CALGB) clinical trials.