725 RETROSPECTIVE MULTICENTER STUDY ON ENDOSCOPIC TREATMENT OF UPPER GASTROINTESTINAL POST-SURGICAL LEAKS

Therapeutic endoscopy plays a major role in the management of upper gastrointestinal (UGI) post-surgical leaks, with multiple endoscopic techniques being available. However, no definite consensus exists on most appropriate therapeutic approach, as well as data is scarce regarding clinical success and safety. Evaluate endoscopic therapy results on the management of UGI post-surgical leaks regarding number and order of therapies performed, as well as safety, clinical success and long-term follow-up. Multicenter, international, retrospective study from 10 centers of consecutive patients who underwent endoscopic treatment of UGI post-surgical leaks. 206 patients (50.5% male) were included, with a median age of 54 years (range 18-81). Previous surgery most often performed was sleeve gastrectomy (38.8%), followed by total gastrectomy (21.8%), Ivor-Lewis esophagectomy (15.0%) and gastric bypass (10.7%). Median time from surgery to beginning of endoscopic treatment was 16 days (range 0-2405). Global leak closure was observed in 187 patients (90.8%). Endoscopic closure was achieved in 165 patients (80.1%), after a median follow-up of 52 days (range 0-693). In 8 patients (3.9%) in whom leak closure was not achieved, a stent was left-in-place, without evidence of leak persistence during follow-up. Fourteen patients (6.8%) underwent surgery after endoscopic treatment failure, and leak closure was possible in 12 of them. One patient underwent radiological leak closure and one patient had spontaneous closure. Most commonly performed first-line endoscopic treatment was SEMS placement [average ranking (AR) 5.5], followed by over-the-scope clips (AR 5.3), endoscopic vacuum therapy (AR 5.1) and endoscopic internal drainage (AR 5.1). Multimodal therapeutic endoscopy was necessary in 40.8% of patients (n=84; median number of different therapies: 2; range 1-6). The median total number of endoscopic procedures performed per patient were 3 (range 1-26). The success-rate of leak closure was 41.3% (85/206) with the first endoscopic technique, 44.3% (47/106) with the second technique, 35.6% (16/45) with the third technique, 59.1% (13/22) with the fourth technique, 80.0% (4/5) with the fifth technique. In one patient that underwent 6 different endoscopic techniques, endoscopic closure was not possible. The global rate of at least one endoscopic therapy-related adverse event (AE) was 39.3% (n=81; one AE: n=60; two AEs: n=21), being severe in 10 patients (4.9%). Leak-related mortality rate was 11.6% (n=24). Median follow-up since beginning of endoscopic therapy was 303 days (range 2-2801). Multimodal therapeutic endoscopy, despite time-consuming, allows leak closure in a significant proportion of patients, with a low rate of associated severe-AEs.