Prognostic value of a three-dimensional spatial classification system for clinical T1a-bN0 non-small cell lung cancer in the superior segment

Why clinicians should know about it

• Picked for Anatomy (paper of the day, 9 November 2026): 3D lung segment anatomy informs surgical margins and resection planning
• Picked for Surgical Oncology (paper of the day, 9 November 2026): Segmentectomy margins and resection extent linked to recurrence

Abstract

BACKGROUND: Recurrence after segmentectomy in non-small cell lung cancer (NSCLC) remains a clinical challenge, particularly for tumors in the superior segment of the lower lobe (S6). We hypothesized that this was attributable to the unique spatial location of S6 tumors and the difficulty in achieving adequate surgical margins. Therefore, we developed a three-dimensional (3D) spatial classification system for S6 tumors and investigated its prognostic value. METHODS: This retrospective cohort study included 916 patients with clinical T1a-bN0 S6 NSCLC who underwent thoracoscopic segmentectomy from August 2015 to April 2023. Tumors were classified into four groups (S6-I to S6-IV) based on the relationship between a simulated 2-cm margin sphere and adjacent anatomical structures: S6-I (margin sphere located above the intersegmental plane), S6-II (margin sphere extended below the intersegmental plane but did not involve main subsegmental structures of the basal segments), S6-III (margin sphere extended below the intersegmental plane and involved main subsegmental structures of the basal segments), and S6-IV (margin sphere involved main trunks of the lower lobe). Surgical methods were categorized as S6 segmentectomy (S6-Seg; including isolated and extended S6 segmentectomy) or combined S6 subsegmentectomy (S6-Comb; combined resection of S6 or its subsegments plus adjacent subsegments, such as resection of S6+8a or S6b+8a). The primary endpoint was recurrence-free survival (RFS). Counterfactual mediation analysis was conducted to assess the mediating role of margin distance. RESULTS: The 5-year RFS rates were 96.9% for S6-I, 95.3% for S6-II, 81.5% for S6-III, and 76.6% for S6-IV (log-rank P<0.001). In multivariable analysisadjusted for age, consolidation-to-tumor ratio (CTR), and tumor size, spatial classification was independently associated with RFS [S6-III vs. S6-I: hazard ratio (HR) =5.30, P<0.001; S6-IV vs. S6-I: HR =5.11, P<0.001]. Counterfactual mediation analysis demonstrated that the association between spatial classification and RFS was largely mediated by margin distance [proportion mediated (PM), 84.0% and 92.0% for S6-III vs. S6-I and S6-IV vs. S6-I, respectively]. In the S6-III subgroup, S6-Comb was associated with superior RFS compared with S6-Seg [HR =0.23, 95% confidence interval (CI): 0.07-0.78, P=0.02], an effect partly mediated by margin distance. In the S6-IV subgroup, surgical method was not associated with RFS. CONCLUSIONS: The spatial classification system effectively stratifies recurrence risk in S6 NSCLC. Its association with recurrence is largely mediated by surgical margin distance, but its primary strength lies in preoperative risk prediction. We propose a preliminary, anatomy-guided hypothesis-generating algorithm: S6-Seg for S6-I and S6-II tumors, S6-Comb for S6-III tumors, and consideration of lobectomy for S6-IV tumors,which requires validation in prospective studies.

Abstract as published, via PubMed.