Reverse Endoscopic Ultrasound-Guided Neogastrojejunostomy Here are the author disclosures. Gastrojejunostomy is a surgical procedure that can treat various causes of gastric outlet obstruction. However, this operation carries a high morbidity and mortality risk. As such, endoscopic ultrasound-guided gastrojejunostomy has emerged as an alternative for surgery, with initial data showing high clinical and technical success rates. Here we describe a case of endoscopic ultrasound-guided gastrojejunostomy being utilized to manage an outlet obstruction from a hiatal hernia in the setting of a colonic interposition. Our case involves a 71-year-old female with a history of a GI stromal tumor treated with total gastrectomy and partial esophagectomy with a colonic graft interposition 17 years prior. She presented for evaluation of postprandial regurgitation of food that occurred 3 hours after eating and an associated 20-pound weight loss. Prior attempts at upper endoscopy failed due to a prominent cricopharyngeal bar, which was confirmed on upper GI series and video photoscopic swallow study, although for the patient it was asymptomatic. The images also showed significant delay across the interposition, or the neo-stomach, into the proximal jejunum. Given these findings, we decided to proceed with an upper endoscopy to further assess and potentially treat a suspected anastomotic stricture. Here is an initial CT scan of her abdomen, which was notable for a large hiatal hernia. Here is the video swallow study with findings consistent with a cricopharyngeal bar. Additionally, we can see her colonic interposition, or the neo-stomach, with retention of contrast and no emptying into the jejunum. During her initial EGD, a hypertonic cricopharyngeus bar was seen, which impeded passage of a therapeutic gastroscope. This required balloon dilatation, after which the endoscope was able to be advanced into the distal esophagus. Approximately half of the colonic interposition was above the diaphragm by fluoroscopy. The collo-jejunal anastomosis was at the level of the diaphragmatic pinch and appeared to have benign stenosis. Initially, a 20mm by 10mm lumen-opposing metal stent was placed. Endoscopically, this did not appear to address the full length of the stricture, so an additional 20mm by 150mm fully covered esophageal stent was placed coaxially in a napkin ring fashion. To prevent stent migration, the proximal end of the esophageal stent was secured to the esophageal wall using an over-the-scope clip. Symptoms of postprandial regurgitation improved in the initial days following the EGD. However, after about 3 weeks, the patient saw recurrence of her symptoms and developed abdominal pain. A CT scan showed distal migration of her stents into the small bowel. A pediatric colonoscope was advanced down the rulum of the jejunum, traversing the cricopharyngeal bar as before. Several catered, clean-based non-bleeding ulcers were visualized, suspected to be stent-induced. Both the esophageal stent and the lumen-opposing metal stent were reached and removed. We surmise that the obstruction at the colojejunal anastomosis was due to extrinsic compression from her diaphragm from a sliding hiatal hernia. To more definitively address this endoscopically, the decision was made to perform an endoscopic ultrasound-guided neogastrojejunostomy to allow the neostomach to empty into the jejunum, further downstream. A 7 French nasobiliary drain was placed into the jejunum. The linear echoendoscope was advanced into the neostomach. Contrast and saline were used to distend the rulum of the jejunum via the nasobiliary drain. A location near the lateral margin at the blind end of the neostomach, away from the hiatal hernia that was in close proximity to a loop of jejunum, was chosen as the target. A 20mm cautery-assisted lumen-opposing metal stent was then advanced across the neostomach. The view of the jejunum was obscured on an endoscopic ultrasound, and it initially appeared that the distal end of the stent was within the lumen of the jejunum. The distal flange of the lumen-opposing metal stent was deployed, but at this point, we suspected an intraperitoneal deployment given the appearance. A wire was advanced through the stent to see if it would end up in the jejunum, but this just coiled in the peritoneum. The proximal flange was deployed in the neostomach. Contrast was injected through a stone balloon over the wire, which again confirmed intraperitoneal deployment of the distal portion of the lumen-opposing metal stent. The echoendoscope was exchanged for a therapeutic gastroscope. An endoscopic exam confirmed a full thickness perforation with visible omental fat through the lumen-opposing metal stent. The jejunal rulum was explored endoscopically, and no obvious bowel injury was seen. The lumen-opposing metal stent was removed. Afterward, the perforation was successfully closed with an over-the-scope clip utilizing an omental patch. The rulum was again explored endoscopically, and CO2 insufflation on fluoroscopy did not show evidence of a visible defect. After some further thought, the neogastro-jejunostomy creation was then attempted again, this time in reverse fashion from the rulum of the jejunum into the neostomach. The neostomach was distended with saline installation. A linear echoendoscope was advanced past the collo-jejunal anastomosis into the rulum of the jejunum using a savory wire to facilitate guidance. During this process, some of the distension of the neostomach was lost. Thus, the neostomach was punctured transjejunally with a 19-gauge needle to instill more saline into the neostomach to create a larger target. A 20-mm cautery-assisted lumen-opposing metal stent was then advanced transjejunally into the neostomach. Care was taken to avoid disrupting the recently placed over-the-scope clip and omental patch, which could be seen on endoscopic ultrasound. The distal flange was deployed in the neostomach and the proximal flange in the rulum of the jejunum. The lumen-opposing metal stent was then balloon dilated to 20 mm. Here, you can see the final anatomic position of the lumen-opposing metal stent across the newly created neo-gastro-jejunostomy anastomosis, which sits below the diaphragmatic pinch. Contrast insulation endoscopically into the lumen did not show any extravasation. A nasogastric tube was placed and attached to the low-intermittent suction overnight. The patient was admitted to the hospital post-operatively for observation. An upper GI series with small bowel follow-through the following day did not show any extravasation of contrast in the neostomach or small bowel and confirmed stent patency. The patient was discharged home after two days. At one month follow-up, the patient had resolution of her clinical symptoms. She reported weight gain and no regurgitation. Endoscopic ultrasound-guided gastro-jejunostomy may provide utility for clinical scenarios that resemble gastric outlet obstruction, such as in this patient with extrinsic obstruction of a neo-gastro-jejunal anastomosis from the diaphragm related to a hiatal hernia. Misdeployment of the distal flange of a lumen-opposing metal stent into the peritoneum is the most common type of stent-related complication when performing this type of procedure. Performing the procedure in reverse fashion allows the use of a less mobile target organ, the clonic interposition graft, compared to the small bowel, thereby decreasing the chances of the target being pushed out of the way. Our case highlights the recognition of a complication, endoscopic treatment of that complication, and the use of a reverse approach to circumvent the challenges of a complex anatomy when the traditional approach is unsuccessful. Further follow-up is needed in our case to better evaluate for long-term effectiveness.

endoscopic ultrasound-guided neogastrojejunostomy

gastric outlet obstruction

hiatal hernia

colonic interposition

lumen-opposing metal stents