Varenicline in prevention of relapse to smoking: effect of quit pattern on response to extended treatment

Aim While older behavioural and pharmacological approaches to preventing relapse to smoking show little efficacy, a recent randomized trial of an extended course of varenicline reported positive results. In this secondary analysis, trial data were examined to see whether smokers who manage to achieve abstinence only later in the original course of treatment are more likely to benefit from having the course extended. Methods A total of 1208 patients abstinent for at least the last week of 12 weeks' treatment with varenicline were randomized to 3 months continued varenicline or placebo. Overall, 44% of the 12-week abstainers were abstinent from the target quit date (TQD), while the rest stopped smoking later. We examined the relationship between quit pattern and the varenicline versus placebo difference in continuous abstinence rates at week 52 and contributions of baseline patient characteristics. Results With increasing delay in initial quitting, 12-month success rates declined. Participants who had their last cigarette at week 11 of open-label treatment had quit rates at 52 weeks of 5.7% compared with 54.9% in those who last smoked in week 1 [odds ratio (OR) 20.3 (6.3, 65.9); P < 0.0001]. Patients who failed to initiate abstinence in the first week benefited more from extended treatment than patients continuously abstinent from week 1 [OR 1.7 (1.2, 2.4); P = 0.0015 versus OR 1.1 (0.8, 1.5); P = 0.6995, respectively; with the interaction of the quit pattern with treatment effect reaching borderline significance (P = 0.0494)]. No other patient characteristics were related to treatment effect. Conclusions Compared with smokers who quit smoking on their TQD, those who have an initial delay in achieving sustained abstinence have increased risk of relapse even several months later, and may be more likely to benefit from extended treatment with varenicline.