IMPACT OF ANNUAL CASE VOLUME ON COLORECTAL ENDOSCOPIC SUBMUCOSAL DISSECTION PROCEDURAL OUTCOMES AND SAFETY IN A LARGE MULTICENTRIC PROSPECTIVE COHORT STUDY

Background: The adoption of endoscopic submucosal dissection (ESD) as the main option for large colorectal lesions is still limited in the West. A recent high-quality study showed colorectal ESD has been proved equally safe and more effective than piecemeal endoscopic mucosal resection (EMR). Reproducibility outside experts centers has been questioned. Thus, we evaluated results according to annual case volume in a multicentric prospective cohort (FECCO NCT04592003).
Methods: Patients referred for colorectal ESD were consecutively included between 09/2019 and 09/2022. 13 participating centers were classified into low-volume (LV) (<50 ESDs/year), middle-volume (MV) (50-100 ESDs/year), and high-volume (HV) centers (>100 ESDs/year). En bloc, R0, curative resection rates and dissection speed as well as complication rates and need for surgery were assessed. Univariate, multivariate and propensity score matching analyses were performed.
Results: 3770 colorectal ESDs were included. 62.4% of the lesions were colonic. HV group performed more colonic cases (68.6%) than MV and LV groups (61.6% and 40.5%, respectively) (p<0.01). Lesions were larger in HV and MV centers (50 x 40 mm for both) than in LV centers (45 x 35 mm) (p<0.01). The overall en bloc resection rate was 95.2%; HV centers achieved a greater rate (96.3%) than MV and LV centers (92.9% and 93.9%, respectively) (p<0.01). Pooled R0 resection rate was 87.4%; it was better in HV centers than in MV and LV centers (88.7% vs 83.8% and 86.8%, respectively). The curative resection rate, being overall 83.2%, was lower in MV centers (79%) than in HV and LV groups (84.7% (p<0.01) and 82.5% (p= 0.10), respectively). The median duration of procedure was 57 min and was shorter in HV centers (48 min) than in LV and MV groups (75 min and 70 min, respectively) (p<0.01). Dissection speed was 28.7 mm²/min, being greater in HV centers (34.8 mm²/min) and lower in MV (22.9 mm²/min) (p<0.01) and LV centers (18.1 mm²/min) (p<0.01). All groups exceed the recommended acceptability thresholds with good quickness. Delayed bleeding and surgery due to complications rates were 5.4% and 0.8%, respectively, with no significant difference. Perforation rate (overall 9%), was higher in MV centers (11.1%) than LV and HV groups (7.5% (p= 0.02) and 8.7% (p= 0.06), respectively). Multivariate analysis found size >50 mm, poor maneuverability, recurrent and appendiceal lesions, but not center volume were risk factors for both R1 resection (p<0.01) and perforation (p<0.01) suggesting differences of outcomes can be explained by lesion characteristics.
Conclusions: In an organized referral system, colorectal ESD can be successfully implemented in the West even in not expert contests. On those terms, ESD should be adopted as the first-line therapy for large colorectal lesions. However, tough lesions must still be referred to the experts.