This analysis included 63 randomised controlled trials (total n=30,508) of any duration that evaluated nicotine replacement therapy (NRT), bupropion or varenicline for smoking cessation and that reported on whether or not any cardiovascular disease (CVD) events occurred. Major outcomes evaluated included all CVD events (diagnosis of any CV event considered in previous systematic reviews) and major adverse cardiovascular events (MACE; including CV death, non-fatal MI, or non-fatal stroke). In the absence of many head-to-head trials evaluating all interventions, a Bayesian random-effects network meta-analysis was conducted.
Eight of the studies included patients with cardiovascular disease, four included patients with COPD and one included perioperative patients – these were included in an analysis restricted to high-risk patients. The median duration of treatments was 12 weeks (interquartile range [IQR] 8-12 weeks) and the median follow-up was 12 months (IQR 6-12 months).
The results of the network meta-analysis were as follows:
• The risk of any CVD event with NRT was increased compared to placebo (relative risk 2.29; 95% Credible Interval 1.39-3.82). There was however no statistically significant increase in risk of MACE (RR 1.95; 0.92-4.30).
• For bupropion, there was no increased risk of any CVD event compared to placebo (RR 0.98; 0.54-1.73) and there appeared to be a protective effect for MACE (RR 0.45; 0.21-0.85).
• For varenicline, there was no increased risk of any CVD event (1.30; 0.79-2.23) or MACE (1.34; 0.66-2.66) compared to placebo.
When the studies examining high-risk patients were analysed separately, the direction of effect was similar to the complete analysis, but none of the results reached statistical significance (although this was based on a smaller sample).
To determine which endpoints were driving the increased risk of all CV events seen with NRT, the MACE events were removed from the analysis. The most commonly reported event was heart palpitations. When including only events associated with NRT that are considered to be of lower severity (palpitations, bradycardia, arrhythmia), the pooled OR was 2.08 (95% CrI 1.35-3.19). The authors concluded based on this that the increased risk was driven by low-risk events.
The limitations of the study include possible reporting bias (concern regarding CVD risk with smoking cessation is a relatively new issue and many studies did not report on CVD safety outcomes), the lower power of the varenicline analysis, and the possibility that study attrition (although rates were similar for the interventions) could reflect intolerability and thus some events may have been missed.
The authors say that their findings, when considered alongside the well-established CVD and mortality risks of continued smoking, suggest that the benefits of smoking cessation therapies outweigh their risks. Further study is however needed, particularly on their use by smokers hospitalised for STEMI.