Endoscopic submucosal dissection performed in livE pig models and in humans with large, laterally spreading gastric and colorectal adenomas – Assessment of the initial learning curves, success rate and complications

Abstract

Endoscopic submucosal dissection is a promising new endoscopic technique for en block resection of early, superficial, flat, and laterally spreading gastrointestinal neoplasms. The primary aims of this prospective study were to summarize the first Hungarian experience with endoscopic submucosal dissection, and to compare the experimental in vivo procedures in pigs with the human application of this technique in two academic centers. The ex vivo Erlangen pig model was used until initial competence was achieved. Thereafter 15 endoscopic submucosal dissection procedures were performed in up to 5 sessions in living pigs under general anesthesia. After submucosal injection with saline Fujinone ball tip, a Flush knife was used with a transparent hood for circumferential incision and submucosal dissection. The following variables were analyzed: specimen size, complete and en bloc resection rate, total duration of the procedure, and complications. Furthermore, between 2009 and 2012, 14 endoscopic submucosal dissections were carried out in humans; 1 duodenal, 3 gastric and 10 colorectal interventions were performed due to large laterally spreading intramucosal neoplasms. The specimen size, en bloc resection rate, total duration of the procedure and complications were next analyzed. 87% complete resection rate, 29.4 ± 19.5 cm2 specimen size and 136.2 ± 26.8 min of procedure duration were achieved in the live pig models. Nontransmural damages of the muscular layer occurred in 3/15 (20%) cases which were successfully closed with hemoclips, except in one pig that died due to fatal esophageal perforation. All bleedings were easily managed with hot biopsy forceps coagulation and hemoclips without hemodynamic instability. In humans, endoscopic submucosal dissection was accomplished with an 11/14 (79%) en block resection rate, with average resected specimen size of 11.56 ± 4.9 cm2 and with 157 ± 55 min of operating time. Two cases of bleeding and 2 cases of perforation occurred (14%), all of which were managed endoscopically. After 1.5 year of average follow up, local recurrence occurred in one case only. In addition, due to the 3 incomplete resections and unsuccessful endoscopic submucosal dissection, surgical intervention was required in 4 cases altogether. This study demonstrates the first Hungarian experience with endoscopic submucosal dissection. Training in live pig models could help endoscopists to overcome the learning curve and minimize the risk of complications before starting the procedure in humans. Reduction in the resection time and low risk of complications, especially bleeding, could be achieved by the application of a flush knife.