Endoscopic Purse-string Closure And Transgastric Gastroenterostomy: The Animal Study

Backgroud: The development of natural orifice translumenal endoscopic surgery(NOTES) made the new endoscopic surgery approach possible when entering viatransesophageal, gastric, colonic, urethral or vaginal access, and the NOTESrepresents the goal of scarless, less invasive procedures with less postoperativepain and faster postoperative recovery. In one hand, the white paper released byNOSCAR group in2006stated that the reliable closure of access points remainsan obstacle before NOTES could be fully implemented in human subjects.Although various methods or equipments of closure have been described inliterature, most of them are technically challenging. A simple, effective, and safegastrotomy closure technique is still lacking. In the other hand, gastric outletobstruction is often caused by primary cancer originated from stomach,duodenum, ampulla or the head of pancreas or metastatic cancer, which istraditionally treated by open or laparoscopic surgery. The development ofNOTES made the transgastric gastroenterostomy possible. Four different groupshave reported very similar methods with the intestine being pulled into the stomach lumen and sutured to the stomach wall with the help of various specialsuturing devices. In2011, Ryou et al. developed a novel self-assembly magnets tocreate gastroenterostomy without the need for suturing. However, all theseprocedures seem to be time-consuming, technically demanding, and requireadditional suturing devices or assistances, which may increase the risk ofcomplications.Aims: To evaluate the technical feasibility and safety of the new methods ofendoscopic closure and gastroenterostomy, which aims to provide basicexperimental data for the clinical NOTES procedures in the future.Methods: Atotal of30healthy female dogs were used in two studies, lincluding5dogs for endoscopic closure using clips,10dogs for endoscopic loop-basedpurse-string closure,3dogs for endoscopic gastroenterostomy using clips, and12dogs for endoscopic gastroenterostomy using partially covered occluder. Beforethe operation, all animals were anaesthetic. Endoscope and instruments andgastric lumen were disinfectant. Cefazolin was used to prevent infection afteroperation.1. Endoscopic intraluminal closure: The gastrotomy was performed invarious regions of the stomach (greater curvature of the antrum, lesser and greatercurvature of the body, anterior and posterior wall of the body). After theprocedure of transgastric NOTES peritoneal exploration, the incision of thestomach were closed by using clips or purse-string method, respectively.Endoscopic closure using clips was performed as following: the first two clipswere released in both the margin area of the incision and then the following clipswere used from the margin area to the center to finalize closure of the incision;Endoscopic loop-based purse-string closure was performed as following: thenylon loop was fully opened and attached to the margins of the gastrotomy by using clips, then the incision was closed by ligation of loop. The endoscopic,postmortem and histological examinations were performed2weeks afteroperation to evaluate the closure of the incision. The feasibility and safety ofincision closure were compared between the purse-string closure method andtraditional clip method.2. Endoscopic transgastric gastroenterostomy: The distal part of the stomachbody along the greater curvature was chosen as the site for the gastrotomy. Afterthe endoscope was advanced into peritoneal cavity, a free-floating small bowelloop was chosen for gastroenterostomy by using clips or partially coveredoccluder, respectively. Endoscopic transgastric gastroenterostomy using clips wasperformed as following: after the double-channel endoscope was advanced intothe peritoneal cavity, a loop of small bowel was grasped and pulled into thegastric lumen through the gastric incision by an endoscopic forceps. Endoclipsthrough another endoscopic channel were used to secure the intestinal and gastricwall together, then the enterotomy was performed by using needle knife on thesurface of the small bowel loop. After that, the small bowel was then firmly fixedto the gastric wall using several other clips; Endoscopic transgastricgastroenterostomy using partially covered occluder was performed as following:after the therapeutic endoscope (single channel) was passed through the gastricincision into the peritoneal cavity, the enterotomy was created on theantimesenteric surface of the small-bowel loop. The occluder-loaded deliverycatheter was then advanced through the intestinal incision into the small bowellumen, and the distal disc was deployed in the small bowel lumen. Gentle tractionon the delivery catheter helped the expanded disc to approximate the small bowelwall against the gastric wall at the gastrotomy site, the proximal disc of theoccluder was then released and expanded in the gastric lumen. In the fully deployed position, the occluder held the intestinal and gastric walls together atthe gastrostomy site for anastomotic healing. Arepeat endoscopy was performed,and the implanted occluder was extracted using a snare. Subsequent endoscopic,radiological, postmortem and histological examinations were performed2weeksafter operation to determine the patency of the gastroenterostomy. The aim of thisstudy was to develop and evaluate a novel method for the creation of agastroenterostomy with an occluder.Results:1. In the first study, the difference of technical success rate between loopgroup and clip group was not significantly (100%,10/10vs80%,4/5; p=0.33),and the mean times for gastrotomy closure in loop group is7.3±1.8min, which isshorter than the clips group (12.1±2.7min, p=0.001). The histological healingrate of loop group is better than clip group (100%vs40%, p=0.022) and thecomplication rate of adhesion in loop group is lower than clip group (10%vs80%, p=0.017). One animal in the clip group died secondary to intra-abdominalinfection because of incomplete closure of the incision. There were no seriouscomplications in loop group.2. In the second study, the technical success rate of the clip method was66.7%(2/3), and mean operation time was61.9±20.8min. The gastric andintestinal walls were unstable fixed by using clips and the small bowel loop slipout of the gastric incision lead to failed to anastomose in one animal. Twoanimals died (66.7%,2/3) in this group on day2and day5due to peritonitis andintestinal obstruction, respectively; The technical success rate of the occludermethod was100%(12/12), and mean operation time was32.3±10.3min. The firstone dog (8.3%,1/12) in this group died on day4secondarily to severeintra-abdominal infection from incorrect deployment of the occluder and poor bowel preparation. Minor bleeding occurred at the anastomosis after removal ofthe occluder in two of the remaining dogs (18.2%,2/11). Necropsy revealedpostoperative adhesions developed at the anastomotic site in one dog (9.1%,1/11). And no intra-abdominal infection or intestinal obstruction was observed inthis group. Complete healing of the anastomosis was confirmed on histologicalevaluation.Conclusions:1. Endoscopic loop-based purse-string closure is reliably for closure ofmulti-regional NOTES gastrotomy, and seems to be easier and safer than clipsmethod. The purse-string technology may be useful for the repair of theiatrogenic gastric perforation.2. The endoscopic gastroenterostomy using clips is technically feasible,however, this procedure being time-consuming and associated with highmorbidity and mortality; The novel technique for gastroenterostomy creation byan occluder is technically feasible in animal model and may provide an effectivealternative for the creation of an endoscopic gastroenterostomy.