New Meta-Analysis Shows Varenicline is Safe

1Feb2012-blog144-photo1

In July 2011, Singh et al. published a meta-analysis in the Canadian Medical Association Journal indicating an increase in serious cardiovascular events in patients taking varenicline for smoking cessation (1). This paper raised concern in the medical community and caused quite an impact in the media.

In a careful reading of the paper, some details — such as the definition of serious cardiovascular events, the exclusion of clinical trials and the calculation of the number needed to harm (NNH) — challenged the validity of this meta-analysis and its results.

Concern in the scientific community, and the fact that varenicline was one of the few effective drugs for smoking cessation, led my colleagues and me to perform a new systematic review and meta-analysis without the aforementioned limitations.

New study

The new study was recently published in the Revista Uruguaya de Cardiología (2). It included 19 randomized trials from a large search of three databases (PubMed, clinicaltrials.gov and clinicalstudyresults.org) and looked at a total of 5,200 patients treated with varenicline and 3,565 who received a placebo. The outcome variable was defined more precisely in this study; “serious cardiovascular events” were considered to be death, myocardial infarction and stroke.

The authors observed that the overall incidence of serious cardiovascular events associated with varenicline was very low — 0.03% (18/5,200) with varenicline and 0.02% (8/3,565) with placebo. These were not statistically different (odds ratio of 1.91, 95% confidence interval 0.84 to 4.24).

When looking only at patients with stable cardiovascular diseases, the highest incidence of events was observed — 2.8% (10/355) with varenicline and 1.7% (6/359) with placebo. Again, these were not statistically different (odds ratio of 1.71, 95% confidence interval 0.61 to 4.74). The authors wrote that this occurrence may have been just by chance or because of the longer time spent in the studies by the varenicline patients, as they mentioned in another publication (3).

Finally, the authors proposed that even when the increased incidence would happen in the magnitude reported, the NNH would be 821 (increased absolute risk of 1.22 for every 1,000 patients treated) — and it would be largely counterbalanced by the reduction of events that occur if 1 in every 8 smokers on varenicline for 12 weeks quit smoking (NNT of 8). These arguments were consistent with the reports of the European Medicines Agency in Europe and the Food and Drug Administration in the U.S.

The authors concluded that the results of this meta-analysis support that varenicline is a safe drug with an appropriate risk-benefit profile to help people — both without cardiovascular disease and with stable cardiovascular disease — quit smoking.


  1. Singh S, et al. Risk of serious adverse cardiovascular events associated with varenicline: A systematic review and meta-analysis. Canadian Medical Association Journal. 2011;183:1359.
  2. Sandoya E, et al. Efectos cardiovasculares de la vareniclina: Revisión sistemática y metaanálisis. Revista Uruguaya de Cardiología. 2011;26:244.
  3. Jimenez-Ruiz C, et al. Lo que sabemos, lo que dudamos y lo que no sabemos. Prevención del Tabaquismo. 2011;13:98.