Immediate vs On-Demand Endoscopic Necrosectomy in Infected Walled-Off Pancreatic Necrosis: A Randomized Controlled Trial

Abstract

Background & aims: Endoscopic drainage is established treatment for infected walled-off pancreatic necrosis (WON). However, the optimal timing of necrosectomy remains controversial, with limited tier 1 evidence comparing immediate vs on-demand necrosectomy strategies. This study compares the efficacy and safety of immediate vs on-demand endoscopic necrosectomy in patients with infected WON undergoing endoscopic ultrasound (EUS)-guided drainage.

Methods: This is a single-center, prospective, randomized controlled superiority trial. Patients presenting with infected pancreatic necrosis were randomly assigned (1:1) to immediate necrosectomy or on-demand necrosectomy after initial EUS-guided drainage. Clinical success was defined as a reduction in the collection diameter to less than 3 cm, accompanied by complete resolution of the WON-related symptoms. The primary outcome was the number of reinterventions to achieve clinical success. The secondary outcomes were procedure-related adverse events, length of hospital stay, necrosectomy procedure duration, and mortality.

Results: Between September 2022 and December 2024, we screened 84 patients, of whom 50 patients were randomly assigned to immediate necrosectomy (n = 25) or on-demand necrosectomy (n = 25). Baseline characteristics were similar between the 2 groups. Clinical success was achieved in 23 (92%) of the immediate and 22 (88%) of the on-demand necrosectomy patients (P = .5). The median number of reinterventions was 1 (interquartile range [IQR], 0-2) in both groups (difference, -0.16; 95% confidence interval [CI], -0.42 to 0.1; P = .5). The median number of necrosectomy sessions, including the index procedure was 2 (IQR, 1-3) in the immediate necrosectomy, compared with 1 (IQR, 0-2) in the on-demand group (difference: -0.44; 95% CI, -0.63 to -0.25; P ≤ .01). Forty-four percent of the on-demand group resolved WON without necrosectomy. Procedure-related adverse events were 10 (40%) vs 6 (24%) in the immediate and on-demand groups, respectively (relative risk, 1.67; 95% CI, 0.7-3.9; P = .2). Moderate to severe adverse events occurred in 4 (16%) in the immediate necrosectomy as compared with none in the on-demand necrosectomy group (relative risk, 0.84; 95% CI, 0.7-0.99; P = .1). Major bleeding occurred in 2 (8%) of the immediate but in none of the on-demand group. Hospital stay (median, 14 vs 8 days; P = .8) and mortality (8% vs 12%; P = .5) were similar between the 2 groups.

Conclusions: In patients with infected walled-off pancreatic necrosis, immediate necrosectomy was not superior to on-demand necrosectomy. However, the on-demand strategy was associated with fewer necrosectomy sessions and fewer adverse events.

Clinicaltrials: gov, Number: NCT05530772.

Keywords: Acute Pancreatitis; Endoscopic Necrosectomy; Interventional Endoscopy; Randomized Controlled Trial; Walled-Off Pancreatic Necrosis.