No abstract

Backgrounds Multiple polymorphisms affecting smoking behavior have been identified through genome-wide association (GWA) studies. Circulating levels of the nicotine metabolite cotinine is a marker of recent smoking exposure. Hence, genetic variants influencing smoking behavior are expected to be associated with cotinine levels. Methods We conducted an analysis in a lung cancer (LC) case-control study nested within the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort. We investigated the effects of SNPs previously associated with smoking behavior on i) circulating cotinine, and ii) LC risk. 894 cases and 1805 controls were analyzed for cotinine and genotyped for 10 polymorphisms on 7p14, 8p11, 10q23, 15q25 and 19q13. Results Two variants in the nicotinic acetylcholine receptor subunit genes CHRNA5 and CHRNA3 on 15q25, rs16969968 and rs578776, were associated with cotinine (P=0.001 and 0.03, respectively) in current smokers, and with LC risk (P<0.001 and P=0.001, respectively). Two 19q13 variants, rs7937 and rs4105144 were associated with increased cotinine (P=0.003 and <0.001, respectively), but decreased LC risk (P=0.01 for both, after adjusting for cotinine). Variants in 7p14, 8p11 and 10q23 were not associated with cotinine or LC. Conclusions 15q25 and 19q13 SNPs were associated with circulating cotinine. The directions of association for 15q25 variants with cotinine were in accordance with that expected of LC risk, whereas SNPs on 19q13 displayed contrasting associations of cotinine and LC that require further investigation. Impact This study is the largest to date investigating the effects of polymorphisms affecting smoking behavior on lung cancer risk using circulating cotinine measures as proxies for recent smoking behavior.