A21-09 Extent of Endobronchial Ultrasound Nodal Sampling and Risk of Pathologic N2 Upstaging in Clinical N0-N1 Non-small Cell Lung Cancer

Background Accurate mediastinal staging is essential for optimizing treatment strategies in non-small cell lung cancer (NSCLC). Although systematic endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is recommended for preoperative staging, the optimal extent of nodal sampling remains unclear. We investigated whether the extent and number of mediastinal N2 and N3 stations sampled during staging EBUS were associated with the risk of unforeseen pathologic N2 upstaging in patients with clinical N0-N1 NSCLC. Methods We retrospectively analyzed patients who underwent staging EBUS-TBNA at Asan Medical Center, Seoul, South Korea between January 2022 and December 2023, followed by curative surgical resection for clinical N0-N1 NSCLC. Patients were stratified into three groups according to the extent of mediastinal nodal sampling during EBUS-TBNA: Group 1, limited sampling of ≤ 1 mediastinal station at the N2 or N3 level; Group 2, sampling of one N2 and one N3 station; and Group 3, systematic sampling of ≥ 3 mediastinal stations including at least one N2 and one N3. The primary outcome was pathologic N2 upstaging, defined as N2 disease identified on surgical pathology despite preoperative staging as N0-N1. Logistic regression analyses were performed to identify predictors of upstaging. Results A total of 418 patients (mean age 68.6 ± 7.9 years; 64.1% male) were included. Adenocarcinoma was the predominant histology (75.4%), and most cases were cN0 (94.7%). Group distribution was 140 (33.5%) in Group 1, 192 (45.9%) in Group 2, and 86 (20.6%) in Group 3. Overall, unforeseen pathologic N2 upstaging occurred in 37 patients (8.9%). The upstaging rate decreased progressively with increasing sampling extent: 15.0% in Group 1, 8.7% in Group 2, and 4.3% in Group 3 (P for trend = 0.011). Compared with Group 1, Group 3 had a significantly lower rate of upstaging (P = 0.005; Bonferroni-adjusted). In multivariable analysis, higher primary tumor SUVmax (adjusted OR 1.007; 95% CI 1.002-1.012; P = 0.011) and clinical N1 disease (adjusted OR 1.136; 95% CI 1.004-1.285; P = 0.044) independently predicted pathologic N2 upstaging, whereas sampling ≥3 mediastinal stations was associated with a lower risk (adjusted OR 0.918; 95% CI 0.861-0.979; P = 0.010). Conclusions Comprehensive EBUS-TBNA sampling of at least three mediastinal stations—including both N2 and N3—significantly reduced the risk of occult N2 upstaging in clinical N0-N1 NSCLC. Findings from this study underscore the importance of systematic, guideline-concordant mediastinal staging to enhance diagnostic accuracy and inform optimal treatment planning. This abstract is funded by: None