Su1357 A NEW COLLECTION: REMOVAL AND REPLACEMENT OF THE SAME LUMEN-APPOSING METAL STENT FOR MULTIPLE ENDOSCOPIC NECROSECTOMY SESSIONS

Abstract

The recently approved lumen-apposing self-expandable metal stent (LAMS) provides safe and effective transmural drainage of pancreatic fluid collections (PFC). Large walled-off necrosis (WON) require multiple direct endoscopic necrosectomy (DEN) through the stent. However, the amount of necrotic material extracted is limited by the inner diameter of the stent. Removal of LAMS can facilitate DEN by improving scope maneuverability through the cystogastrostomy and allow removal of larger solid necrotic debris. The aim of this study is to evaluate the feasibility and safety of this novel technique of removal and replacement of the same LAMS to facilitate endoscopic necrosectomy. A retrospective analysis of a prospectively maintained database of all patients undergoing EUS-guided drainage of PFC via cautery-enhanced LAMS was performed. Patients with WON who underwent more than one necrosectomy session with removal and replacement of LAMS from 01/2017 to 11/2017 were included. All patients underwent fluoroless EUS-guided transgastric cautery-enhanced LAMS for perigastric WON in the usual fashion. During subsequent endoscopy for DEN, the LAMS was initially removed using a snare or rat-toothed forceps. The WON cavity was entered through the cystgastrostomy, which provided enhanced access and maneuverability to perform DEN. At the end of session, the same LAMS was manually inserted inside the distal tip of the gastroscope and deployed across the cystogastrostomy at the end of each session (Fig 1). A total of 17 patients (mean age 38.4±14.7 years; 88% males) were included. The mean size of WON was 10.4±3.5 cm by 9±3.7 cm. The mean interval from EUS-guided LAMS placement to first endoscopic necrosectomy session was 20.4±13.3 days. Median number of sessions required was 2 (range 1-5; mean 2.3). Complete resolution was achieved after a median of 2 (range 1-5) DEN sessions with stent retrieval after 8.1±6.4 weeks. This modification provided better visualization of deeper recesses as well as improved scope tip maneuverability even allowing for scope tip retroflexion within the cavity as needed. Snare was used in all sessions with removal of larger necrotic pieces with each pass, leading to less number of passes, shorter duration of procedure, and less number of sessions, compared to the results of a recently published meta-analysis (2.3 vs. 4.09; p <0.05). LAMS was successfully replaced across the cystgastrostomy tract during all sessions (n=39). No stent migration was noted. There was no evidence of stent damage or tissue ingrowth. Our study demonstrates removal and replacement of LAMS during multiple endoscopic necrosectomies is feasible, safe, and facilitates DEN with reduction in procedure duration and number of sessions. Intentional removal of the stent should be considered when a large necrosectomy is done.