Abstinence and relapse among smokers who use varenicline in a quit attempt-a pooled analysis of randomized controlled trials

Background and aims: Varenicline increases the likelihood of long-term abstinence following a quit attempt. It has been suggested that (1) part of its benefit arises from 'recruiting into abstinence' smokers who are not able to stop on the target quit date and (2) there may be a higher rate of relapse after treatment. This study addressed these issues.

Methods: Meta-analyses of data from randomized controlled trials (RCTs) of varenicline identified from the 2012 Cochrane review of nicotine receptor partial agonists for smoking cessation were used to compare the abstinence and relapse patterns on active drug and placebo. Studies of varenicline compared with placebo in adult daily smokers with longest follow-up at either 6 or 12 months and with at least three follow-ups in the first month were included. Biochemically verified abstinence rates at each of six follow-up time-points (2, 3, 4, 12, 24 and 52 weeks) were pooled for studies reporting point prevalence abstinence. Biochemically verified abstinence rates at three follow-up time-periods (9-12 weeks, 9-24 weeks and 9-52 weeks) were pooled for studies reporting continuous abstinence. Random effects meta-analysis was used to estimate pooled proportions with 95% confidence intervals.

Results: Nineteen RCTs were included. In varenicline-treated participants, point-prevalence abstinence increased by 22 percentage points from week 2 [32%: 95% confidence interval (CI) = 25-40%] to week 12 (54%: 95% CI = 48-61%). The increase was 8 percentage points in the placebo group: 16% (95% CI = 11-21%) to 24% (95% CI = 17-33%). In varenicline-treated participants the relapse from weeks 9-12 to week 52 was 55%: 49% abstinent in weeks 9-12 (95% CI = 45-53%) versus 22% at week 52 (95% CI = 19-25%). In placebo-treated participants it was 53%: 17% (95% CI = 13-25%) in weeks 9-12 versus 8% (95% CI = 6-12%) at week 52.

Conclusions: Varenicline recruits smokers into abstinence following the target quit date to a greater extent than placebo. Relapse rates from end of treatment to 52 weeks are similar in varenicline- and placebo-treated smokers.