Conti et al identified the association of rs2072661 with abstinence in analyses of a double-blind randomized controlled trial of placebo or active bupropion, e.g., a SNP OR (95% confidence interval) of 0.40 (0.25–0.67) at EOT and of 0.31 (0.18–0.55) at 6MO [46]. The associations of this SNP with a variety of smoking related phenotypes [35,43,54], including short-term cessation in a cross-over trial of NRT and placebo patch [53], suggested to us that rs2072661 might exhibit effects on abstinence with other PGs and that we might more accurately quantify its association in larger samples. However, we did not observe statistically significant association of rs2072661 with any PG when analyzing 24 arms of seven RCTs, i.e., excluding the two arms of the RCT in which the abstinence association was discovered [46]. If the main effect size of rs2072661 on abstinence in RCT participants is weaker than the effect observed by Conti et al, which is expected [81], analysis of additional RCT arms will be necessary to validate the original [46] or subsequent associations [53], or discover novel associations.
