Chantix Linked to Heart Risks

The analysis, published online in CMAJ, showed that 1.06% of varenicline-treated study participants had serious adverse cardiovascular events compared with 0.82% in placebo-treated participants. The difference translated into an odds ratio of 1.72 for varenicline versus placebo (95% CI 1.09-2.71).

First author Sonal Singh, MD, Johns Hopkins, stated that, "In absolute terms, it is a small difference." "However, when you consider the difference from a population perspective, you get a very different picture.

"It is estimated that 7 million people in the U.S. were using Chantix last year. If you apply our results to the total number of users, I estimate that there were 62,500 cardiac events that were linked to use of Chantix. That is not a small number, by any standards."

The report follows an FDA announcement last month that prescribing information for varenicline will be revised to include a warning about an increased risk of second heart attack and peripheral artery disease in people with a history of cardiovascular disease.

All but one of the 14 trials in the current analysis excluded patients with a history of cardiovascular disease. The one exception included patients with stable cardiovascular disease but excluded those with unstable conditions.

The trials were all randomized and placebo-controlled and involved a total of 8,216 participants, consisting of smokers and users of smokeless tobacco. Two of the trials also had a bupropion (Zyban) comparator group.

The trials' sample sizes ranged from 250 to 1,210 participants, and treatment duration ranged from seven to 52 weeks. The primary outcome in 12 of the trials was the continuous abstinence rate. One trial each evaluated the long-term quit rate and long-term safety of varenicline.

The primary outcome was any ischemic or arrhythmic adverse cardiovascular event reported by trial investigators. All-cause mortality was a secondary endpoint.

In their sensitivity analysis, Singh and coauthors included an open-label comparison of varenicline and nicotine replacement.

Singh and coauthors found that investigators in five of the trials reported a total of 14 deaths, too few to perform a pooled analysis. Sensitivity analyses yielded odds ratios similar to the principal results (OR 1.67-1.77).

Exclusion of the trial involving patients with stable cardiovascular disease produced an odds ratio of 2.54 for varenicline versus placebo.

Analysis of trials that used varenicline doses less than 1 mg BID resulted in an odds ratio of 1.80. An attempt to analyze the risk associated with specific types of clinical events produced an odds ratio that did not achieve statistical significance.

J. Taylor Hays, MD, of the Mayo Clinic in Rochester, MN, wrote that the large increase in relative risk "must be tempered by the rarity of these events among participants in both treatment groups." Other reasons for a cautious interpretation of the data include a higher number of placebo-treated participants lost to follow-up -- introducing potential bias -- and the lack of a significant difference in event rates in the one trial that included patients with stable cardiovascular disease.

"The best outcome from this analysis would be more rigorous and adequately powered studies evaluating the safety of using varenicline among smokers who have known cardiovascular disease," Hays wrote.

"The worst outcome would be for healthcare providers to abandon the use of varenicline, which has proven to be among the most efficacious pharmacotherapies used for the treatment of tobacco dependence."

• Note that the absolute increase in the rate of serious cardiovascular events associated with varenicline versus placebo appears to be less than 1%.
• Point out that studies indicate that smoking kills more than half of persistent smokers and to the extent that varenicline enhances smoking cessation, the potential benefits would appear to far outweigh the risk for serious cardiovascular events.

Singh S et al "Risk of serious adverse cardiovascular events associated with varenicline: A systematic review and meta-analysis" CMAJ 2011; DOI: 10.1503/cmaj.110804.