Real-world surgical outcomes of segmentectomy vs. lobectomy in older adults with small stage IA NSCLC: long-term cancer-specific and competing mortality in an overlap-weighted SEER cohort

This is the authors' abstract. We don't add key points for this paper.

Background Segmentectomy preserves more pulmonary parenchyma than lobectomy, but whether this translates into lower competing mortality in older adults is uncertain. Objective To compare lung cancer-specific and other-cause mortality after segmentectomy vs. lobectomy in older adults with small stage IA non-small cell lung cancer (NSCLC) treated in routine US practice. Methods We used Surveillance, Epidemiology, and End Results data to identify adults aged ≥70 years with a first primary, 1–20 mm, stage IA NSCLC treated with segmentectomy or lobectomy. The primary 5-year cohort included diagnoses from 2010 to 2018 and the secondary 3-year cohort diagnoses through 2020. Propensity-score overlap weighting, competing-risk methods, cause-specific Cox models, Fine-Gray models, and 1,000-replicate bootstrap inference were used. Additional sensitivity analyses quantified unmeasured-confounding robustness with E -values and repeated the primary analysis after restricting to patients with ≥1, ≥6, or ≥10 regional nodes examined, with propensity scores re-estimated within each restricted cohort. Results The primary cohort included 4,072 patients (430 segmentectomy; 3,642 lobectomy). Five-year lung cancer-specific mortality was 15.23% after segmentectomy and 11.15% after lobectomy (RD, +4.08 deaths per 100 patients; 95% percentile bootstrap CI, +0.63 to +7.31); the cause-specific HR was 1.31 (95% CI, 1.05–1.63). Other-cause mortality was similar. The primary E -value was 1.71 for the point estimate and 1.23 for the confidence-limit estimate. With increasingly stringent nodal-evaluation restrictions, the lung cancer-specific association attenuated and became less precise; for ≥6 examined nodes, the 5-year RD was +4.18 per 100 (95% CI, −1.04 to +9.58), and for ≥10 nodes it was +1.88 per 100 (95% CI, −4.58 to +8.59). Conclusions In routine practice, segmentectomy was associated with higher 5-year lung cancer-specific mortality in the primary analysis, but attenuation under more extensive nodal evaluation and modest E -values indicate that patient selection and procedural quality may materially contribute. These findings should not be interpreted as evidence of an intrinsic causal disadvantage of segmentectomy in appropriately selected patients.