Bridging Genomics and Pharmacoepidemiology to Expand Treatment Options for Alcohol Use Disorder

Alcohol use disorder (AUD) is a chronic, relapsing condition and a major public health problem. However, few medications are approved to treat AUD, and those available show limited efficacy. Drug repurposing is a cost-effective strategy to identify novel therapeutic uses for existing medications. Here, we describe a pipeline that integrates genetic and electronic health record (EHR) data to identify and evaluate drugs to be repurposed for treating AUD. Our approach comprises 1) alcohol-associated gene identification and biological network generation; 2) mapping drugs to target proteins; 3) filtering promising repurposing candidates; and 4) an exemplar pharmacoepidemiologic analysis of the effect of an identified drug (i.e., baclofen) on alcohol consumption. Linking loci to genes from a genome-wide association study (GWAS) of problematic alcohol use identified 94 genes, which we expanded to 327 alcohol-related genes through network-based analyses. Across these analyses, 52 genes were linked to 195 FDA-approved drugs, including four already approved or used off-label to treat AUD. After filtering for safety, relevance, and data availability, 26 candidate drugs, including baclofen, were selected for further evaluation. An evaluation of the real-world effectiveness of baclofen using national EHR data from the United States Department of Veterans Affairs provided evidence that baclofen-exposed patients reduced alcohol consumption more than propensity-score-matched unexposed patients. This approach, which aligns genomic findings with real-world clinical data, provides an efficient method for identifying promising drug repurposing candidates and prioritizing those that merit evaluation in randomized trials to ultimately advance pharmacotherapies for AUD.