In 2009 FDA issued a black box warning for varenicline and neuropsychiatric events. We studied efficacy (smoking cessation) of varenicline, and safety (neuropsychiatric events) in both randomized clinical trials (RCTs) and a large observational study. The observational study was included to determine the generalizability of the RCT findings to the general population.
RCTs: Re-analysis of all 17 placebo controlled RCTs (n=8027) of varenicline conducted by Pfizer using complete intent-to-treat person-level longitudinal data.
Analysis of Department of Defense collected adverse neuropsychiatric adverse event data in inpatients and outpatients taking varenicline versus nicotine replacement therapy (NRT) (n=35,800). The primary endpoints for the RCTs were smoking abstinence and adverse event reports of suicidal thoughts and behavior, depression, aggression/agitation, and nausea. The effect of varenicline in patients with (n=1004) and without (n=7023) psychiatric disorders was examined. The primary endpoints for the observational study were anxiety, depression, drug induced mental disorder, episodic and mood disorder, other psychiatric disorder, post traumatic stress disorder, schizophrenia, suicide attempt, transient mental disorder.
RCTs: Varenicline did not increase rates of suicidal events, depression, or aggression/agitation. Varenicline increased risk of nausea (OR=3.69, 95% CI = (3.03, 4.48), p<0.0001). Varenicline increased rate of abstinence by 124% compared to placebo (p<0.0001), and 22% compared to bupropion (p<0.0001). While having a current psychiatric disorder or history of psychiatric illness increased the risk of neuropsychiatric events, it did so equally in treated and control patients.
Following propensity score matching, overall rate of neuropsychiatric disorders was lower for varenicline versus NRT (2.28% versus 3.16%, p<0.0001).
In the RCTs, varenicline revealed no increased risk of neuropsychiatric adverse events relative to placebo. Varenicline provided greater benefit in terms of smoking cessation relative to both placebo and bupropion. The same results were observed in patients with and without a current psychiatric disorder or history of psychiatric illness. In the observational study, the overall rate of neuropsychiatric disorders was lower in patients treated with varenicline relative to NRT, revealing that the finding of no increased risk of neuropsychiatric adverse events in RCTs generalizes to the population of patients engaging in treatment with varenicline.