Leveraging population-based clinical quantitative phenotyping for drug repositioning

Computational drug repositioning methods can scalably nominate approved drugs for new diseases, with reduced risk of unforeseen side effects. The majority of methods eschew individual-level phenotypes despite the promise of biomarker-driven repositioning. In this study, we propose a framework for discovering serendipitous interactions between drugs and routine clinical phenotypes in cross-sectional observational studies. Key to our strategy is the use of a healthy and non-diabetic population derived from the National Health and Nutrition Examination Survey, mitigating risk for confounding by indication. We combine complementary diagnostic phenotypes (fasting glucose and glucose response) and associate them with prescription drug usage. We then sought confirmation of phenotype-drug associations in un-identifiable member claims data from Aetna using a retrospective self-controlled case analysis approach. We identify bupropion hydrochloride as a plausible antidiabetic agent, suggesting that surveying otherwise healthy individuals cross-sectional studies can discover new drug repositioning hypotheses that have applicability to longitudinal clinical practice.