Varenicline (Chantix) is a nicotine partial agonist that is approved by the FDA for smoking cessation. Since July 1, 2009, the FDA has also required a “black box warning” based on postmarketing reports of adverse neuropsychiatric effects. The label now warns clinicians to watch for “serious [or] unusual changes in mood and behavior.”
To explore the correlation, if any, between varenicline and adverse psychiatric effects, researchers reexamined data from two large sources.
First, they obtained complete data from 17 placebo-controlled randomized trials conducted by its manufacturer (Pfizer), which included a total of 8,027 subjects. They evaluated reports of suicidal behavior, depression, agitation/ aggression, and nausea.
In parallel, they reviewed 30-day and 60-day reports of neuropsychiatric events in 35,800 patients receiving varenicline or nicotine replacement therapy (NRT) as part of a large Department of Defense (DOD) observational study commissioned by the FDA between 2006 and 2007, prior to the FDA warning.
The researchers found that the risk of suicidality, depression, and agitation, relative to placebo, was not elevated in subjects who took varenicline in Pfizer’s clinical trials, which lasted an average of 11.6 weeks. In the observational study, rates of all neurospsychiatric adverse events were actually lower for varenicline than for NRT (2.28% vs. 3.16%).
Based on these observations, the authors conclude that varenicline is not associated with a risk of neuropsychiatric events, and suggest that the FDA warning may have stimulated reports of adverse events after 2009 that may not have been due to varenicline. Indeed, they point out that other studies showing increases in suicidality (eg, Moore TJ et al, PLoS One 2011;6:e27016) may be due to overreporting since the FDA warning in 2009.
These results may appear to give varenicline a clean bill of health, but there may be more to the story than meets the eye. The study’s authors have published prior reports refuting the link between suicidality and other drugs such as SSRIs and gabapentin (Neurontin); these studies have been criticized due to possible selective use of data and the way “suicidality” is measured in clinical trials.
The same problems may plague the present study, since industry-funded trials may not code adverse events in a reliable fashion, and the observational study similarly relied upon clinicians’ reporting of diagnostic terms (such as “depressive disorder” or “suicide attempt”) which may not capture the full range of neuropsychiatric events. Several of concerns have been raised by other investigators (see, for instance, Spielmans GI et al, JAMA Psychiatry 2013;70(1):121–122).
The close relationship between the authors and varenicline’s manufacturer should also be highlighted. The lead author discloses that he “has served as an expert witness for Pfizer in a case related to varenicline,” and Pfizer was permitted to review the manuscript prior to its submission.
Furthermore, the authors relied on unpublished clinical-trial data (as they have in prior publications claiming to debunk the suicidality link) from Pfizer, meaning that an independent reassessment of their methods and calculations is not possible (Gibbons RD and Mann JJ, Am J Psychiatry 2013;170(12):1460–1467).
TCPR’s TAKE: The link between psychotropic drugs and serious adverse events such as suicide remains murky. This study suggests that varenicline is no more likely to cause suicidal behavior or other neuropsychiatric effects than placebo or NRT, and that such events may have been overreported since the FDA mandated the black-box warning on varenicline in 2009. However, the classification of what’s considered an adverse event or “suicidal behavior” must be clarified, and without free and complete access to the data used to draw these conclusions, independent analysis is not possible. Finally, financial connections between the authors and the drug manufacturer may result in a less-than-objective appraisal of the data.