Effect of the number of lymph nodes examined on the survival of patients with stage I non–small cell lung cancer who undergo sublobar resection

Early stage lung cancer is being detected at a higher frequency with the implementation of screening programs. At the same time, medically complex patients with multiple comorbidities are presenting for surgery, with a concomitant rise in rates of sublobar resection. We sought to examine the effect of sampling lymph nodes on the outcomes of patients who undergo sublobar resection for small (<2 cm) stage I non-small cell lung cancer (NSCLC).All patients in the Surveillance, Epidemiology, and End Results database from 2004 to 2013 with small (<2 cm) stage I NSCLC who underwent sublobar resection (wedge/segmentectomy) and no other cancer history were included. The association of the number of lymph nodes examined (LNE; categories none, 1-3, 4-6, 7-9, >9) with the overall survival as well as disease-specific survival were examined using univariate as well as multivariate analyses while controlling for covariates such as age, size (<1 cm, >1 cm), grade, histology (adenocarcinoma vs others), and extent of resection (wedge/segmentectomy).Data from 3916 eligible patients were analyzed. Seven hundred fifteen patients (18.3%) had segmentectomy. No lymph nodes were examined in 49% and 23% of wedge resection and segmentectomy patients, respectively. Among all eligible patients, 1132 (29%), 474 (12%), 228 (6%), and 328 (8%) patients had 1 to 3, 4 to 6, 7 to 9 and >9 LNE, respectively. Univariate analyses showed significant associations between overall and disease-specific survivals with age, grade, histology, sex, extent of surgery, and LNE. The association between the number of LNE and survival remained significant even after adjusting for significant covariates including extent of sublobar resection (hazard ratio for groups with LNE 1-3, 4-6, 7-9, and >9 compared with 0 LNE were 0.79, 0.77, 0.68, and 0.45 for overall survival; P < .001) and 0.85, 0.77, 0.71, and 0.44 for disease-specific survival (P < .05), respectively. In multivariate modeling, LNE was retained as a significant variable and extent of resection was not. In patients in whom at least 1 lymph node was examined, extent of resection was not predictive of outcome.Many patients having sublobar resection for early stage NSCLC in the United States do not have a single lymph node removed for pathologic examination. The number of LNE is associated with improved survival, presumably due to avoidance of mis-staging. This association seems greater than the association with extent of resection (segmentectomy vs wedge resection). Appropriate lymph node examination remains an important part of resection for lung cancer even if the resection is sublobar.