Propensity score matched analysis for the role of surgery in stage Ⅲ small cell lung cancer based on the eighth edition of the TNM classification: a population study of the US SEER database and a Chinese hospital

•In this study, we re-staged patients with stage III small cell lung cancer (SCLC) from both SEER database and our hospital according to 8th TNM classification.•Propensity score matching (PSM) was used to balance clinical bias.•Surgery improved survival in patients with stage III SCLC, when compared with non-surgical treatment.•Lobectomy (LB) is the optimal surgery type for stage III SCLC patients. Patients with very early stage small cell lung cancer (SCLC) can benefit from surgery. However, the role of surgery in local advanced SCLC patients remains controversial. We designed this study to investigate the role of surgery on survival of this subset population. The included patients were identified from the Surveillance, Epidemiology, and End Results SEER database from 1998 to 2016 and Shanghai Chest Hospital of China from 2009 to 2016. Propensity score matching(PSM) was used to balance clinical bias. The overall survival (OS) and lung cancer-specific survival (LCSS) were compared by the Kaplan–Meier analysis. Cox proportional hazards regression was used to identify factors associated with survival. Among the 3005 stage Ⅲ patients, 570 (18.97%) patients underwent surgery. Compared with non-surgical group, patients undergoing surgery were more likely to be male, had smaller tumor size, mediastinal lymph node involvement and lower pathologic stage. The Kaplan-Meier analysis showed that surgical patients had a better OS and LCSS before and after PSM. 418 surgical patients were well matched with non-surgical patients. In matched surgical group, there were 224 (53.59%) patients who underwent lobectomy (LB), 147 (35.17%) patients who received sublobectomy (SLB), 31 (7.41%) patients who underwent pneumonectomy and 16 (3.83%) patients with unknown surgery type. The 5-year OS of the 4 subgroups were 28.80%, 12.50%, 8.70% and 13.50%, respectively (P = 0.002). In a multivariable Cox model, SLB (hazard ratio, 1.53; 95%CI, 1.20–1.96; P = 0.001) and pneumonectomy (hazard ratio, 1.72; 95%CI, 1.12–2.65; P = 0.013) were associated with worse OS compared with LB. Surgical resection significantly improved OS and LCSS of stage Ⅲ SCLC patients in our study. Furthermore, LB had advantage over other surgery type but further exploration in larger prospective clinical trials is needed.