Comparison of conventional and modified endoscopic mucosal resection methods for the treatment of rectal neuroendocrine tumors

Background/aim Small rectal neuroendocrine tumors (NETs) confined to the submucosa are endoscopically resectable. Because most rectal NETs are submucosal tumors, conventional endoscopic mucosal resection (cEMR) may not result in a complete resection. This study investigated whether modified EMRs, namely endoscopic submucosal resection with ligation (ESMR-L), EMR with precutting (EMR-P), and strip biopsy are superior to cEMR for achieving histologically complete resection (HCR) of rectal NETs. Methods Medical records of 215 patients who were treated with endoscopic resections for rectal NETs between January 2011 and July 2019 were retrospectively enrolled. Of the patients, 110, 33, 29, and 43 underwent cEMR, ESMR-L, EMR-P, and strip biopsy, respectively. For each method, HCR and en bloc resection rates, procedure times, and complication rates were measured. Results HCR was achieved with cEMR, EMR-P, ESMR-L, and strip biopsies for 74.5%, 90.9%, 93.1%, and 90.7% of cases, respectively. The HCR rate for cEMR was inferior to those of the modified EMRs ( p = 0.045 for cEMR vs. EMR-P; p = 0.031 for cEMR vs. ESMR-L; p = 0.027 for cEMR vs. strip biopsy). Among the three modified EMRs, there was no significant difference in achieving HCR ( p = 1.000). En bloc resection ( p = 0.096) and complication rates ( p = 0.071) were not significantly different among the four EMR methods, although EMR-P required the longest procedure time ( p = 0.000). Conclusions All three modified EMRs are superior to cEMR and are equally effective for achieving HCR of rectal NETs.