Robotic versus laparoscopic gastrectomy with splenic hilar lymph node dissection: short-term outcomes from a propensity score-matched study.

Spleen-preserving dissection of the No. 10 (splenic hilar) nodal station is technically demanding, given the narrow working space bordered by terminal splenic vessels and the pancreatic tail. This study aimed to compare short-term outcomes of robotic and laparoscopic No. 10 dissection. We retrospectively compared 113 patients undergoing gastrectomy with No. 10 dissection (laparoscopic, n = 69; robotic, n = 44) at a single center (November 2008-March 2026). Propensity scores (age, sex, BMI, preoperative chemotherapy, clinical T4 status, nodal positivity) were used for 1:1 nearest-neighbor matching (0.2-SD logit caliper), yielding 35 pairs; paired outcomes were compared with Wilcoxon signed-rank and McNemar tests. In the matched cohort, robotic surgery had less blood loss (0 [IQR 0-20] vs. 30 [5-150] mL; P < 0.001), lower day-1 drain amylase (248 [161-481] vs. 593 [427-1113] IU/L; P = 0.001), and higher No. 10 nodal yield (1 [1-2] vs. 1 [1-1]; P = 0.009), with a trend toward longer operative time (404 [367-438] vs. 372 [333-402] min; P = 0.050). Total nodal yield, morbidity (overall 17.1% vs. 11.4%, P = 0.683; severe 14.3% vs. 11.4%, P = 1.000), and hospital stay (14 vs. 13 days; P = 0.339) did not differ; neither arm had splenic hilar fistula or No. 10 metastasis. Robotic gastrectomy was associated with less blood loss, lower pancreatic enzyme elevation, and higher No. 10 nodal yield than laparoscopy, without added morbidity or longer stay. Prospective, multicenter studies are needed to confirm and clarify the oncological rationale for routine No. 10 dissection.Trial registration: University Hospital Medical Information Network (UMIN000032895).