Window-assisted endoscopic resection for ≤2 cm gastric submucosal tumors: a single-center retrospective comparative study

Background: Endoscopic resection is a standard treatment for gastric submucosal tumors (SMTs). However, resecting tumors ≤2 cm presents challenges such as localization difficulty and perforation risk. This study compared outcomes between a window-assisted endoscopic resection technique and traditional endoscopic methods. Methods: This single-center retrospective study analyzed 19 patients with gastric SMTs (≤2 cm) resected endoscopically between January 2024 and September 2025. Patients were categorized into a window-assisted group (n=7) and a traditional resection group (n=12, comprising ESE, EFTR, or STER). Outcomes included operative time, wound closure time, complication rates, postoperative fasting duration, hospital stay, number of titanium clips used, and patient satisfaction. Statistical analyses used Student’s t-test or Mann-Whitney U test for continuous data, and Chi-square or Fisher’s exact test for categorical data. Results: R0 resection was achieved in all cases. The window-assisted group showed significantly shorter operative time (63.57±20.12 vs. 94.58±17.24 minutes, p=0.002) and wound closure time (16.86±8.76 vs. 32.08±11.35 minutes, p=0.005). Intraoperative perforation occurred less frequently in the window-assisted group (28.57% vs. 66.67%, p=0.045). Postoperative abdominal pain (42.86% vs. 66.67%, p=0.043) and local peritonitis (14.29% vs. 66.67%, p=0.023) were also lower in this group. The window-assisted group had shorter fasting times (2.00±0.82 vs. 4.08±1.62 days, p=0.003), shorter hospital stays (8.29±2.69 vs. 10.17±3.13 days, p=0.048), used fewer titanium clips (7.00±3.16 vs. 12.25±4.12, p=0.008), and reported higher patient satisfaction (85.71% vs. 41.67%, p=0.038). No recurrence was observed during short-term follow-up (1-3 months). Conclusion: In this retrospective study, window-assisted endoscopic resection for gastric SMTs ≤2 cm was associated with shorter operative and closure times, fewer clips used, and lower rates of certain complications compared to traditional techniques. However, these findings are limited by the small sample size and retrospective design. Further prospective studies are warranted.