Data Mining of the US FDA’s Adverse Events Reporting System Database to Evaluate Drug–Drug Interactions Associated with Statin-Induced Rhabdomyolysis

Abstract

With the rise in polypharmacy, it is increasingly important to identify drug–drug interactions (DDIs) that cause serious adverse events in a timely manner. The purpose of the study was to investigate the utility of systematic data mining of the US Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS) database for interactions between statins (HMG-CoA reductase inhibitors) and other drugs that underlie a higher risk for rhabdomyolysis. The strength of association for the reporting of rhabdomyolysis with each statin–drug combination was measured using Bayesian data mining scores, the Interaction Signal Score (INTSS), and simple proportions that were calculated as the reporting odds ratio (ROR). Scores >1.0 indicate disproportionately higher than expected reporting. A manual case review highlighted strengths and limitations of these measures. As expected, clarithromycin and cyclosporine produced high measures of disproportionate reporting of rhabdomyolysis with lovastatin and simvastatin. Drugs with no known predilection to contribute to statin myopathy produced scores <1.0 when paired with each statin. In contrast, in some instances INTSS values were <1.0 when measures of DDI with statin–drug pairs known to interact. This might be due to masking by high numbers of reports linked to the non-statin drug alone. The manual review identified dose as an important risk factor; however, this risk factor could not have been systematically identified without informatics enhancements. While disproportionality methods represent a promising tool for identifying a potential serious DDI, opportunities remain for improvements in both data mining algorithms and the acquisition of adequately informative data.