A prospective, multicenter study of 1111 colorectal endoscopic submucosal dissections

Background: Endoscopic submucosal dissection (ESD) is accepted as a minimally invasive treatment for early gastric cancer, although it is not widely used in the colorectum because of technical difficulty. Objective: To examine the current status of colorectal ESDs at specialized endoscopic treatment centers. Design and Setting: Multicenter cohort study using a prospectively completed database at 10 specialized institutions. Patients and Interventions: From June 1998 to February 2008, 1111 colorectal tumors in 1090 patients were treated by ESD. Main Outcome Measurements: Tumor size, macroscopic type, histology, procedure time, en bloc and curative resection rates and complications. Results: Included in the 1111 tumors were 356 tubular adenomas, 519 intramucosal cancers, 112 superficial submucosal (SM) cancers, 101 SM deep cancers, 18 carcinoid tumors, 1 mucosa-associated lymphoid tissue lymphoma, and 4 serrated lesions. Macroscopic types included 956 laterally spreading tumors, 30 depressed, 62 protruded, 44 recurrent, and 19 SM tumors. The en bloc and curative resection rates were 88% and 89%, respectively. The mean procedure time +/- standard deviation was 116 +/- 88 minutes with a mean tumor size of 35 +/- 18 mm. Perforations occurred in 54 cases (4.9%) with 4 cases of delayed perforation (0.4%) and 17 cases of postoperative bleeding (1.5%). Two immediate perforations with ineffective endoscopic clipping and 3 delayed perforations required emergency surgery. Tumor size of 50 mm or larger was an independent risk factor for complications, whereas a large number of ESDs performed at an institution decreased the risk of complications. Limitations: No long-term outcome data. Conclusions: ESD performed by experienced endoscopists is an effective alternative treatment to surgery, providing high en bloc and curative resection rates for large superficial colorectal tumors. (Gastrointest Endosc 2010;72:1217-25.)