Abstract
Background/Objectives: While sublobar resection (SLR) is accepted for selected small, early non-small-cell lung cancers (NSCLCs), its efficacy for tumors with visceral pleural invasion (VPI) remains debated. This study aimed to compare lung-cancer-specific survival (LCSS) between SLR and lobectomy in pT2a (tumor ≤ 3 cm with VPI) N0M0 NSCLCs from a nationwide population-based database. Methods: This retrospective study utilized Taiwan Cancer Registry data from 2011 to 2018, selecting patients with pT2a (tumor ≤ 3 cm with VPI) N0M0 NSCLC that underwent SLR or lobectomy, with specific exclusion criteria. Propensity score matching (1:1) was performed using a greedy algorithm with a 0.01 caliper width. LCSS was compared using Kaplan-Meier analysis with log-rank tests and Cox proportional hazards models before and after matching. Results: In the 2460-patient cohort (624 SLR, 1836 lobectomy) before matching, lobectomy was associated with significantly better overall (p = 0.01) and PL1 VPI subgroup (p = 0.009) LCSS. In the matched cohort (523 pairs), no significant difference in LCSS was observed between SLR and lobectomy, either overall (p = 0.21) or when stratified by PL1 (p = 0.11) or PL2 (p = 0.94) status. Multivariate Cox analysis in the matched cohort confirmed no significant association between surgery type and LCSS (adjusted HR 0.75, 95% CI 0.52-1.08, p = 0.124), but older age (>75 years), PL2 VPI, and lymphovascular invasion were independent predictors of worse LCSS (all p < 0.001). Conclusions: This large population-based study, after rigorous adjustment for confounders, found that SLR and lobectomy provided comparable LCSS. SLR may be an alternative for selected patients, but prospective validation is recommended.