No evident association between efavirenz use and suicidality was identified from a disproportionality analysis using the FAERS database

Objective: To assess the potential association of selected antiretrovirals (ARVs), including efavirenz, with suicidality. Design: Retrospective analysis of the Food and Drug Administration Adverse Event Reporting System (FAERS), by performing a Multi-Item Gamma Poisson Shrinker (MGPS) disproportionality analysis. Methods: MGPS disproportionality analysis, a technique to identify associations between drugs and adverse events, was performed using cumulative data from the FAERS database collected up to August 2012. This method yields an Empirical Bayesian Geometric Mean score and corresponding 90% confidence interval (EB05, EB95). EB05 scores >= 2 were pre-defined as a signal for a potential drug-event association. The FAERS database includes spontaneous adverse-event reports from consumers and healthcare professionals. All FAERS reports of suicidality (including suicidal ideation, suicide attempt and completed suicide or a composite of these) in patients taking efavirenz (as single agent or in fixed-dose combination), atazanavir, darunavir, etravirine, nevirapine and raltegravir were identified. A number of parallel analyses were performed to assess the validity of the methodology: fluoxetine and sertraline, antidepressants with a known association with suicidality, and raltegravir, an ARV with rhabdomyolysis and myopathy listed as "uncommon'' events in the US-prescribing information. Results: A total of 29,856 adverse event reports were identified among patients receiving efavirenz, atazanavir, darunavir, etravirine, nevirapine and raltegravir, of which 457 were reports of suicidality events. EB05 scores observed for the composite suicidality term for efavirenz (EB05 = 0.796), and other ARVs (EB05 = 0.279-0.368), were below the pre-defined threshold. Fluoxetine and sertraline gave EB05 scores for suicidality >2. Raltegravir gave EB05 scores >2 for myopathy and rhabdomyolysis. Conclusions: The pre-determined threshold for signals for suicidality, including suicidal ideation, suicide attempt, completed suicide and a composite suicidality endpoint, was not exceeded for efavirenz and other ARVs in this analysis. Efavirenz has been associated with suicidality in clinical trials. Further studies that adjust for confounding factors are needed to better understand any potential association with ARVs and suicidality.