Predicting success of direct endoscopic necrosectomy with lumen-apposing metal stents for pancreatic walled-off necrosis

Abstract

Direct endoscopic necrosectomy (DEN) with lumen-apposing metal stents (LAMSs) is increasingly used in the management of pancreatic walled-off necrosis (WON). However, it still remains unknown which patients will fail to respond to DEN with LAMSs and require additional surgical intervention. Therefore, the aim of our study was to explore predictors of successful DEN with LAMSs for pancreatic WON. This is a retrospective analysis of a prospectively collected database. All consecutive patients who underwent DEN with LAMSs for pancreatic WON were included. Collected data were demographics, disease severity, morphologic features, and procedure characteristics. Potential factors affecting DEN outcome were predefined and analyzed. One hundred one consecutive patients undergoing DEN with LAMSs for WON were identified, among whom 4 patients were excluded for technical failure (n= 1) and previous debridement without LAMSs (n= 3). In the 97 included patients, clinical success was achieved in 79 patients (81.4%). In logistic multivariable regression, 3 independent factors were negatively associated with success of DEN with LAMSs: increasing Acute Physiology and Chronic Health Evaluation II score (odds ratio [OR], .70; 95% confidence interval [CI], .55-.90; P= .005), >50% pancreatic necrosis (OR, .16; 95% CI, .03-.77; P= .022), and paracolic gutter extension (OR, .08; 95% CI, .02-.42; P= .003). A receiver-operating characteristic curve of the prediction model with these 3 factors demonstrated an area under the curve of .926. Paracolic gutter extension, increasing Acute Physiology and Chronic Health Evaluation II score, and >50% gland necrosis are negative predictors for success of DEN with LAMSs in WON. This prediction model with nomogram may be helpful in clinical decision-making and prognostication.