Stent indwell duration and adverse events in endoscopic ultrasound-guided lumen-apposing metal stent drainage of peripancreatic collections: systematic review and meta-analysis

In the drainage of peripancreatic fluid collections (PFCs), current practice recommends the removal of lumen-apposing metal stents (LAMSs) at 3–4 weeks, despite limited evidence. Many patients require longer indwell times (e.g. for serial necrosectomies or owing to logistical constraints). This study aimed to evaluate the incidence of adverse events (AEs) in patients undergoing PFC drainage with a LAMS, focusing on the timing of these events relative to the date of stent placement. A systematic review with meta-analysis of prospective studies evaluating the use of LAMSs in PFC drainage was conducted. Of the 1242 references identified, 15 studies, encompassing 662 patients, were included. From these studies, 128 AEs (19%; 95%CI 17% to 23%) were reported, with timing ascertainable in 122 cases (95%). The most frequently reported AEs were stent obstruction (n = 39; 32%, 95%CI 22% to 37%) and bleeding (n = 31; 25%, 95%CI 15% to 29%); 12 AEs (9%, 95%CI 6% to 18%) occurred during stent removal; 41 AEs (32%; 95%CI 25% to 42%) were classified as severe. Mortality associated with the LAMS was reported in 5/662 cases (0.8%, 95%CI 0 to 2%). No association was found between stent indwell time and the incidence of AEs in the time–event assessment. No association was observed regarding the publication year, type of collection, type of stent, or the use of cautery. In patients undergoing endoscopic ultrasound-guided drainage of PFCs, evidence suggests no significant association between LAMS indwell time and the incidence of AEs. This challenges rigid early removal policies and supports individualized, clinically driven timing, with structured follow-up and risk mitigation.