Pharmacovigilance Analysis of Adverse Psychiatric Events and Suicidality Reported for Roflumilast, an Add-On COPD Therapy
Ayad K. Ali *
Eli Lilly and Company, Indianapolis, IN, USA.
*Author to whom correspondence should be addressed.
Abstract
Aims: Roflumilast is a phosphodiesterase-4-inhibitor used as add-on therapy to long-acting bronchodilators in chronic obstructive pulmonary disease. Although roflumilast is well tolerated, there have been concerns regarding psychiatric problems, including suicide tendencies. This study aims to identify and characterize signals of adverse psychiatric events reported for roflumilast in the US FDA Adverse Event Reporting System (FAERS).
Study Design: Retrospective pharmacovigilance analysis.
Place and Duration of Study: Adverse event reports submitted to FAERS from October 1997 through September 2012.
Methodology: Multi-item Gamma Poisson Shrinker data-mining algorithm was applied to adverse psychiatric events (APE) that were submitted to the FAERS (3Q1997-3Q2012). Empirical Bayes Geometric Mean (EBGM) and 95% confidence interval (EB05-EB95) were calculated for roflumilast-associated APE compared to all drugs in FAERS. The following Preferred Terms of the MedDRA terminology were used to define the outcome of interest: “anxiety”, “depressed mood”, “depression”, “insomnia”, “suicide attempt”, and “suicidal ideation”. Signals with EB05>2 are considered significant disproportional reporting (>twice that expected) of APE.
Results:126 reports of APE were identified for roflumilast, corresponding to mutually non-exclusive events of insomnia (n=53), anxiety (n=38), depression (n=36), suicidal ideation (n=30), depressed mood (n=8), and suicide attempt (n=6). EBGM (EB05-EB95) were: APE, 3.55 (3.06-4.11); insomnia, 4.55 (3.62-5.66); anxiety, 2.96 (2.26-3.82); depression, 2.88 (2.19-3.75); suicidal ideation, 5.65 (4.16-7.52); depressed mood, 3.90 (2.20-6.53); and suicide attempt, 1.66 (0.86-2.95).
Conclusion: Roflumilast is associated with higher than expected reporting of APE, including suicidal thoughts, but not suicide attempts. Given the inherent confounding and bias limitations of spontaneous reporting systems, pharmacoepidemiologic studies are required to test these hypotheses; meanwhile, prescribers should consider alternative add-on therapies to patients with past or present depression or suicidality.
Keywords: Roflumilast, Adverse Event Reporting System, FAERS, Suicide, Psychiatric Events, COPD, Pharmacovigilance