Endoscopic submucosal resection with band ligation versus endoscopic submucosal excavation for small (≤ 10 mm) gastric muscularis propria-derived lesions: a real-world retrospective cohort study with long-term follow-up

Abstract

The optimal management of small (≤ 10 mm) gastric subepithelial lesions (SELs) arising from the muscularis propria (MP) remains controversial. Endoscopic submucosal resection with band ligation (ESMR-L) is a simplified resection technique that may improve procedural efficiency compared with endoscopic submucosal excavation (ESE). However, independent real-world comparative data and longer-term follow-up remain limited. This study compared the efficacy, safety, and procedural efficiency of ESMR-L and ESE for small gastric MP-derived SELs. This single-center retrospective cohort study included 68 consecutive patients with gastric MP-derived SELs measuring ≤ 10 mm on preoperative endoscopic ultrasonography (EUS) who underwent ESMR-L ( n = 35) or ESE ( n = 33) between January 2019 and April 2025. The primary outcome was procedure time. Secondary outcomes included en bloc resection, R0 resection, endoscopic full-thickness resection (EFTR), clip use, adverse events, postoperative hospital stay, and local recurrence. Multivariable linear regression and propensity score-adjusted sensitivity analysis were performed for procedure time. Baseline characteristics were generally comparable between groups, although fundus lesions were numerically more frequent in the ESMR-L group. En bloc and R0 resection rates were high in both groups (100.0% vs. 97.0%, both P = 0.49). Procedure time was significantly shorter with ESMR-L than with ESE (median, 28.0 [IQR, 19.5–37.0] vs. 46.6 [35.9–67.0] min; P < 0.001). After multivariable adjustment, ESMR-L remained independently associated with a 38.2% relative reduction in procedure time ( P < 0.001), which was consistent with propensity score-adjusted analysis. EFTR occurred more frequently in the ESMR-L group, but the difference was not statistically significant (25.7% vs. 15.2%; P = 0.37). No major intraoperative adverse events, no surgical rescue, and no local recurrence were observed during a median follow-up of 38.5 months. In this real-world cohort, ESMR-L achieved high en bloc and R0 resection rates while substantially shortening procedure time compared with ESE for small (≤ 10 mm) gastric MP-derived SELs, with no observed local recurrence during long-term follow-up. ESMR-L may be an efficient and feasible alternative to ESE in carefully selected patients when performed by experienced endoscopists with reliable defect-closure skills.