Fabio Garofalo, MD, Maxime Noreau-Nguyen, MD, Henri Atlas, MD, Ronald Denis, MD, Pierre Garneau, MD, Radu Pescarus, MD. Hopital Sacre-Coeur de Montreal
Introduction: Laparoscopic sleeve gastrectomy (LSG) has become a widely accepted option in the treatment of mordid obesity. Gastric leaks after LSG occur in 0-7% of the patients, mostly at the gastroesophageal junction. The current treatment algorithm includes drainage, antibiotics, nutritional support and endoluminal control. Among the various endoscopic treatment options available, stents are the most common. Recently, long fully covered stents have become available in the treatment of post LSG leaks. Our hypothesis is that long fully covered stents represent a safe and effective solution for LSG leaks.
Methods and Procedures: A retrospective analysis of our prospectively collected bariatric database was performed between June 2014 to September 2015. We included all patients treated for leaks after LSG. All patients underwent a CT abdomen with oral contrast for the initial gastric sleeve leak as well as to document the fistula closure. Endoscopic treatment included partially covered metallic stent (WallstentTM, Boston Scientific, Ireland) and long fully covered MegastentTM (Taewoong Medical Industries, South Korea).
Results: During this period, a total of 596 LSG were performed. Overall, 7/596 patients (1.2 %) developed a gastric leak. One patient was an outside referral. Patients presented on average 28.6 days after the initial surgery (range, 12-75 days). All fistulas were localized at the level of gastro-oesophageal junction. All 8 patients underwent endoscopic treatment accompanied by either percutaneous or laparoscopic abscess drainage. Fistula closure was achieved in 8/8 cases with an average number of 2 endoscopic interventions (range 1-4). Average time for closure was 2.1 (range, 1-3) months and average follow-up after fistula closure was 5.5 (range, 1-9) months.
Overall, treatment with partially covered metallic stents failed in 2/5 patients and these patients were eventually successfully treated with a long fully covered stent. Stent failure was attributed to ongoing leak around the proximal end of the stent. The initial use of long fully covered stents was successful in 3/3 cases. No stent migrations or perforations were noted. Only 1/8 patients (12.5%) developed a stent-related esophageal stenosis, treated successfully with serial endoscopic dilatations.
Conclusions: Leaks after sleeve gastrectomy are difficult clinical problems. Several endoscopic procedures and multimodality treatments are often necessary. Long fully covered stents appear to be a good alternative to traditional esophageal stents either as primary endoscopic treatment or after failure of other endoscopic treatments.