Background and study aims: Duodenal perforations during upper endoscopy are rare. However, since invasive endoscopic techniques like EMR, ESD, and EUS are increasingly used, so may the incidence of duodenal perforations. Surgical repair is still the mainstay of therapy but associated with significant morbidity and mortality. Endoscopic closure of such perforations would represent a minimal invasive alternative and was therefore evaluated in a randomized controlled animal study. Material and Methods: In 24 domestic pigs a 1cm large duodenal perforation was endoscopically created using needle knife incision with the animals under general anaesthesia. After randomisation pigs were assigned to either surgical repair (n=12) or endoscopic closure (n=12). Endoscopic repair was performed using a 2T160 double channel upper endoscope (Olympus, Hamburg, Germany), a Twin Grasper and 12mm traumatic Over The Scope Clips (OTSC, Ovesco, Germany). Surgical repair was achieved by open laparotomy. Pressurized leak tests of the perforation closures were performed following necropsy. Additionally pressurized leak test were performed in an ex vivo evaluation of hand-sewn closures of 1cm large scalpel incisions (n=18) and of healthy duodenal tissue (n=18). Tissue for ex vivo evaluation was obtained from freshly killed pigs. Results: Mean time for endoscopic closure was 5 min. (Range 3-8 min., SD 2). No complication occurred during any of the closure procedures. At necropsy and macroscopic examination all OTSC and surgical closures showed complete incorporation and closure of the duodenal needle-knife incisions. Pressurized leak test showed a mean burst pressure of 166 mmHg (Range 80-260, SD 65) for OTSC closures and a mean burst pressure of 143 mmHg (Range 30-300, SD 83) for surgical sutures. Ex vivo hand-sewn sutures of 1cm scalpel incisions exhibited a mean burst pressure of 81 mmHg (Range 43-140, SD 31) and ex vivo duodenal tissue exhibited a burst pressure of 247 mmHg (Range 200-300, SD 35). Ex vivo duodenal tissue burst pressure was significantly higher compared to OTSC closure (p<0.01), in vivo surgical closures (p<0.01), and ex vivo hand-sewn closures (p<0.01). OTSC closures were comparable to surgical closures (p=0.461) and superior to ex vivo hand-sewn closures (p<0.01). In vivo surgical closures were superior to ex vivo hand-sewn closures (p<0.01). Conclusions: Endoscopic closure by means of the OTSC system exhibits comparable results to in vivo surgical sutures and seems suitable to attempt closure of 1cm large duodenal perforations.
Read full abstract