Transpyloric lumen-apposing metal stent for management of persistent post-esophagogastrectomy leak

A 54-year-old man with a history of prior vertical sleeve gastrectomy underwent esophagogastrectomy for curative intent of invasive adenocarcinoma arising in gastroesophageal junctional mucosa in the background of Barrett’s esophagus.

On postoperative day 5, his Jackson-Pratt drain fell out, and he developed right-sided chest pain and respiratory distress. His vital signs revealed oxygen saturation of 85% requiring a non-rebreather mask and tachypnea with a respiration rate of 25. His blood pressure and heart rate remained normal. An urgent CT scan revealed evidence of an anastomotic leak with extravasation of contrast and free air adjacent to the lesser curvature of the stomach (Fig. 1). The patient underwent EGD, which showed a 1.5-cm defect at the esophagogastric anastomosis (Fig. 2). An endoscopic suturing device was used to place 2 running sutures to close the defect and achieve apposition of the defect edges. Intraoperative esophagram with contrast and carbon dioxide insufflation suggested adequate closure without evidence of leakage prior to placement of a covered metal stent. Then, an 18- × 155-mm fully covered stent was placed and sutured in place (Fig. 3). A barium esophagram the following day unfortunately revealed a persistent leak along with poor transpyloric drainage of contrast. A repeat endoscopy was performed and showed adequate closure of the leak site. However, there remained a possibility of persistent leakage around the suture insertion sites. Thus, the decision was made to upsize the stent to a 23- × 155-mm fully covered metal stent (Fig. 4). To address the poor transpyloric drainage, the pylorus was dilated to 20 mm with a through-the-scope (TTS) balloon dilator.

An esophagram the following day again revealed a persistent leak and poor transpyloric drainage despite dilation (Fig. 5). Thus, during the following endoscopy, attention was placed on the pylorus, which demonstrated significant spasm and upward angulation related to the gastric pull-up anatomy. A 20-mm LAMS was deployed successfully and followed by a 20-mm TTS balloon dilation. Endoscopic suturing was used to further anchor the stent in the antrum to prevent migration (Fig. 6). A subsequent barium esophagram showed significantly improved transpyloric drainage of contrast and resolution of the anastomotic leak (Fig. 7). Four weeks later, the sutures and stents were removed using endoscopic scissor forceps and rat-tooth forceps, respectively. There was no evidence of residual leak on the follow-up esophagogram.

Improvement of transpyloric drainage should be considered one of the options for management of persistent leak despite usual endoscopic strategies such as primary closure and covered metal stenting (Video 1, available online at www.giejournal.org).