To the Editor: Results from large-scale phenome-wide association studies (PheWAS) allow association of genetic variants with a wide spectrum of human disorders and have provided considerable insight into disease etiologies1. The PheWAS strategy relies on electronically available phenotypic data collected from patient cohorts. PheWAS is similar to a genome-wide association study (GWAS), but whereas a GWAS asks “What genetic variants are associated with a disease?”, a PheWAS asks “What diseases are associated with a genetic variant?” In 2013, Nature Biotechnology published a study by Denny et al. 2 in which they conducted a comprehensive PheWAS on 3,144 singlenucleotide polymorphisms (SNPs) that had been previously associated with a variety of phenotypes by GWAS. This PheWAS, like many other PheWAS published1, 3, used International Classification of Diseases version 9 (ICD9) codes extracted from electronic medical record systems in large patient cohorts to define case-control groups for many phenotypes. In the United States, ICD9 coding is primarily used for billing and can have variable effectiveness in regard to describing discrete phenotypes. Regardless, Denny et al. 2 demonstrated that for many of the GWAS SNPs, PheWAS was able to rediscover expected SNP-disease associations while also identifying novel associations2. We propose that PheWAS results may also provide new opportunities to identify candidates for drug repositioning.

Drug repositioning is the process of discovering new indications for existing drugs and is becoming an important component of drug development as success rates for novel drugs in clinical trials decrease and costs increase4. Critical to drug repositioning is the initial identification of candidate drug-disease relationships. Genetic-based association studies, including GWAS, have proven to be effective for generating hypotheses related to drug repurposing5, 6. GWAS can identify disease susceptibility genes that are targets for existing drugs used to treat different conditions. For example, a large GWAS implicated flavopiridol, a cyclin-dependent kinase 4 (CDK4) inhibitor and anticancer