The prognosis of patients with rectal neuroendocrine tumors (NETs) depends on tumor grade, lesion size, and depth of invasion. For lesions ≤10 mm without invasion of the muscularis propria, endoscopic resection is generally recommended; however, the optimal resection technique remains uncertain. Endoscopic submucosal resection with a ligation device (ESMR-L) may ensure a negative vertical margin by applying band ligation just above the muscularis propria, thereby reducing the risk of positive vertical margins. This study aimed to compare the treatment outcomes of ESMR-L, endoscopic submucosal dissection (ESD), and endoscopic mucosal resection with precutting (EMR-P) for rectal NETs, and to evaluate the resection ability of each resection method.

We retrospectively analyzed 27 consecutive patients who underwent endoscopic resection for rectal NETs at our institution between January 2010 and January 2021. The patients were divided into three groups: ESMR-L (n = 10), ESD (n = 10), and EMR-P (n = 7). Treatment outcomes were compared among the groups. R0 resection was defined as an en bloc resection with histologically negative margins. Curative resection was defined as R0 resection without lymphovascular invasion, based on immunohistochemical staining. In addition, we compared the resection depth between ESMR-L and ESD using histopathological specimens.

Baseline characteristics were not significantly different among the three groups. The overall R0 resection rate was 88.9%, with no statistically significant differences among groups (ESMR-L: 90%, ESD: 80%, EMR-P: 100%; p = 0.82). Positive vertical margins were observed only in the ESD group (ESMR-L: 0%, ESD: 20%, EMR-P: 0%). The procedure time was significantly shorter for ESMR-L (4.2 ± 1.0 min) than for ESD (32.1 ± 14.7 min) and EMR-P (13.4 ± 3.7 min) (p < 0.001). No adverse events occurred in any group. All tumors were classified as grade 1; lymphatic invasion, venous invasion, and overall lymphovascular invasion were observed in 22.2%, 30.0%, and 44.4% of the patients, respectively. The curative resection rate was 44.4%. In the pathological analysis, the mean distance from the deepest part of the tumor to the vertical resection margin was significantly longer in the ESMR-L group than in the ESD group (900 ± 680 μm vs. 379 ± 212 μm, p=0.01).

ESMR-L demonstrated resection outcomes comparable to those of ESD and EMR-P, with a shorter procedure time. Moreover, compared to ESD, ESMR-L was associated with a lower rate of positive vertical margin, suggesting that it may facilitate a deeper resection depth. For rectal NETs, ESMR-L appears to be a safe and effective endoscopic resection technique.