Surgical management of chronic fistula after sleeve gastrectomy.

Background: There is no clear definition of the chronic leak after sleeve gastrectomy. There are several endoscopic approaches, including endoprothese, endoscopic clips, endoscopic sealing glue, or balloon dilation. In case of failure of the endoscopic treatment, a definitive surgical approach can be attempted. The objective was to evaluate the surgical treatment of chronic leak after sleeve gastrectomy.

Methods: From November 2010 through March 2012, 8 patients with chronic gastric fistula after laparoscopic sleeve gastrectomy had definitive surgical repair. The initial intervention, the diagnosis and management of the fistula, and the endoscopic approach were carefully reviewed.

Results: Five patients had their original laparoscopic sleeve gastrectomies performed at another hospital, while 3 had laparoscopic sleeve gastrectomy at our institution. The mean period of time from the diagnosis of the fistula to definitive surgical treatment was 14.4 months (range 5-44 months). Seven patients initially had surgical drainage by laparoscopy (5) and by laparotomy (2), with concomitant feeding jejunostomy in 6 patients. The endoscopic treatment consisted of endoprothese in 4 patients, endoscopic sealing glue in 2 patients, and sequential approach with glue and prosthesis in 2 other patients. One patient was treated exclusively by endoscopic approach with no surgical drainage. The surgical procedures performed for chronic fistula were gastrojejunal lateral anastomosis (4), Roux-en-Y gastric bypass (2), and gastrectomy with esojejunal anastomosis (2). Four patients presented with postoperative fistula, with a mean healing time of 32 days (range 22-63 days). No mortality was recorded.

Conclusion: In chronic forms of fistulas with no improvements by endoscopic approach, the surgical treatment can be a solution. It remains a difficult procedure with a high percentage of leakage, but this type of fistula is more easily tolerated by the patient and heals faster.