Objectives
It is widely accepted that surgical resection of localized pulmonary typical carcinoid (TC) tumours remains the primary curative modality. However, the optimal extent of resection remains controversial. This study aimed to investigate the survival rates of patients with stage T1-2N0M0 TC tumours who underwent sublobar resection or lobectomy.
Methods
We queried the Surveillance, Epidemiology, and End Results database for patients who underwent surgery after being diagnosed with stage T1-2N0M0 TCs from 2004 to 2016. Propensity score matching (PSM) analysis was used to equalize baseline characteristics between the sublobar resection group and the lobectomy group. Kaplan–Meier analysis and the Cox proportional hazard model were performed for survival analysis.
Results
Of the 2469 patients included, 658 (26.65%) underwent sublobar resection and 1811 (73.35%) underwent lobectomy. All 2469 patients were analysed with PSM and, following PSM, 812 patients were included in the final analysis, divided into two groups of 406 patients. In the matched cohort, Kaplan–Meier analysis demonstrated no significant difference in survival curves between the sublobar resection and lobectomy groups in patients with stage T1-2N0M0 TC tumours (5-year OS = 90.78% versus 93.30%; HR 1.18, 95% CI: 0.77–1.80; P = 0.505). Subgroup analysis by tumour size showed that the sublobar resection group was identical to the lobectomy group in overall survival (OS) for tumours ≤3.0 cm. In addition, no difference in OS between surgical groups was observed in any subgroups. In multivariable Cox analysis, age ≤ 65y, female sex, married status, and adequate lymph node assessment (≥5) were associated with improved OS, whereas the extent of resection was not.
Conclusions
Sublobar resection seems to be associated with similar survival to lobectomy for stage T1-2N0M0 TC tumours if lymph node assessment is performed adequately. This analysis suggests sublobar resection should be considered an appropriate alternative for stage T1-2N0M0 TC tumours. However, further validations are needed in large, multicentre prospective studies.