Lymph node yield and station count in early robot-assisted versus established video-assisted resection for non-small cell lung cancer: higher total node counts reflect yield per sampled station.

Whether robot-assisted thoracic surgery (RATS) improves nodal staging over video-assisted thoracoscopic surgery (VATS) early in adoption is unknown. We compared occult nodal upstaging and lymph node (LN) retrieval between early RATS and an established VATS practice by one surgeon, and asked whether total LN count reflects the number of stations sampled. Retrospective single-surgeon cohort of consecutive lung cancer resections with systematic LN dissection (2023-2026). Primary outcome: occult nodal upstaging (pathologic N + in clinically N0 patients). Secondary outcomes: LN yield, distinct stations sampled, and yield per sampled station, and perioperative outcomes. Eighty-seven patients qualified (RATS, n = 49; VATS, n = 38; 92.1% of VATS operations after RATS introduction). Occult upstaging was 10.6% vs. 10.5% (95% CI for the difference, - 14.8 to + 13.7% points; not significant). Total LN yield was higher with VATS (13.8 vs. 17.7; p = 0.018), whereas stations sampled were similar (4.7 vs. 4.7; equivalent within post-hoc margins of ± 1 and ± 0.75 but not ± 0.5 stations); yield per sampled station was lower with RATS (2.5 vs. 3.3; p = 0.001), mostly in mediastinal stations. This did not attenuate across the 49 RATS cases and persisted after multivariable adjustment (rate ratio 1.29; 95% CI 1.04-1.60). Operative time was longer with RATS; complications and R0 resection did not differ. No significant difference in occult upstaging was observed, but the study was underpowered and this result is inconclusive. Higher VATS node counts reflected yield per sampled station rather than more stations, for reasons that could not be determined. Station-level yield should be reported alongside total node count.