Long-term Survival Outcomes by Smoking Status in Surgical and Nonsurgical Patients With Non-small Cell Lung Cancer Comparing Never Smokers and Current Smokers

Background: Survival outcomes of never smokers with non-small cell lung cancer (NSCLC) who undergo surgery are poorly characterized. This investigation compared surgical outcomes of never and current smokers with NSCLC. Methods: This investigation was a single-institution retrospective study of never and current smokers with NSCLC from 1975 to 2004. From an analytic cohort of 4,546 patients with NSCLC, we identified 724 never smokers and 3,822 current smokers. Overall, 1,142 patients underwent surgery with curative intent. For survival analysis by smoking status, hazard ratios (HRs) were estimated using Cox proportional hazard modeling and then further adjusted by other covatiates. Results: Never smokers were significantly more likely than current smokers to be women (P < .01), older (P < .01), and to have adenocarcinoma (P < .01) and bronchioloalveolar carcinoma (P < .01). No statistically significant differences existed in stage distribution at presentation for the analytic cohort (P = .35) or for the subgroup undergoing surgery (P = .24). The strongest risk factors of mortality among patients with NSCLC who underwent surgery were advanced stage (adjusted hazard ratio, 3.43; 95% CI, 2.32-5.07; P < .01) and elevated American Society of Anesthesiologists classification (adjusted hazard ratio, 2.18; 95% CI, 1.40-3.40; P < .01). The minor trend toward an elevated risk of death on univariate analysis for current vs never smokers in the surgically treated group (hazard ratio, 1.20; 95% CI, 0.98-1.46; P = .07) was completely eliminated when the model was adjusted for covariates (P = .97). Conclusions: Our findings suggest that smoking status at time of lung cancer diagnosis has little impact on the long-term survival of patients with NSCLC, especially after curative surgery. Despite different etiologies between lung cancer in never and current smokers the prognosis is equally dismal. CHEST 2010; 138(3):500-509