Endoscopic Submucosal Dissection for Colorectal Lesions in a Community-based European Hospital: Experience and Learning Curve

Abstract

Background and Aims: Endoscopic submucosal dissection (ESD) has become an established technique for en-bloc resection of large or complex colorectal lesions. However, conventional ESD (C-ESD) can be technically challenging and time-consuming. For non-malignant lesions, ESD with snaring [hybrid ESD (H-ESD)] can facilitate lesion removal and to save time. We report the experience and learning curve in a regional hospital setting in Switzerland.

Methods: We retrospectively evaluated the outcomes of all ESD procedures performed between 2018 and 2024 in our hospital. Primary outcome was the recurrence rate at the first follow-up endoscopy. Secondary outcomes included en-bloc resections, procedural time and complication rates stratified by periods in the learning curve.

Results: Ninety patients (42 women; mean age 67 ± 12 years) were included in the study of which 51 underwent C-ESD and 39 H-ESD. The recurrence rate at first follow-up endoscopy was 8.8% (6/68) with no difference (p>0.05) between C-ESD (8.1%; 3/37) and H-ESD (9.7%; 3/31). The overall en-bloc resection rate was 63%, significantly higher (p<0.001) in the C-ESD group (98%) vs H-ESD (18%). Similarly, R0 resection rates were significantly (p<0.001) higher in C-ESD (86%) vs H-ESD (18%) groups. Procedural times were similar in the C-ESD (86 ± 44 min) and H-ESD (88 ± 45 min) groups. All parameters improved as the learning curve advanced. Complications requiring surgical intervention occurred in two cases (2.2%) one in the C-ESD and one in the H-ESD group.

Conclusions: Colorectal ESD is a feasible and can be self-taught in non-tertiary hospitals, achieving outcomes comparable to those reported in specialized centers, under appropriate case selection. Although conventional ESD achieves higher en-bloc and R0-resections than hybrid ESD, the latter may serve as a rescue option in selected cases.