Factors Predictive of Complete Excision of Large Colorectal Neoplasia Using Hybrid Endoscopic Submucosal Dissection: A KASID Multicenter Study

Abstract

BACKGROUND: Endoscopic submucosal dissection (ESD) with snaring (hybrid ESD) bridges the gap between ESD and endoscopic mucosal resection. We evaluated factors predictive of en bloc and complete resection of large colorectal neoplasms using hybrid ESD. METHODS: This was a prospective clinical study of 78 patients who underwent hybrid ESD for excision of colorectal neoplasms (≥ 2cm) between May 2015 and September 2016 at six university hospitals. We evaluated lesion and patient characteristics, endoscopist experience level (< 50 or ≥ 50 cases with colorectal ESD), and technical factors such as concurrent fibrosis, completion of a circumferential incision, degree of submucosal dissection (< 50 or ≥ 50%), and visualization during snaring (< 50 or ≥ 50%). RESULTS: Multivariate analyses showed that the en bloc resection rate was significantly related to the degree of visualization during snaring (odds ratio (OR) 7.811, 95% confidence interval [CI] 1.722-35.426; p = 0.008) and the presence of fibrosis (OR 0.258, 95% CI 0.68-0.993; p = 0.049). The complete resection rate was significantly related to the colorectal ESD endoscopist skill level (OR 5.626, 95% CI 1.485-21.313; p = 0.011) and gross lesion type (OR 0.145, 95% CI 0.022-0.936; p = 0.042). When all three technical factors, i.e., completion of circumferential incision, ≥ 50% submucosal dissection, and ≥ 50% visualization during snaring, were satisfied performing hybrid ESD, the en bloc resection rate (87.5%) was similar to that of ESD. CONCLUSIONS: Visualization during snaring, presence of fibrosis, gross lesion type, and endoscopist colorectal ESD experience level affect en bloc or complete resection of large colorectal neoplasia using hybrid ESD.