EUS-guided transluminal drainage using lumen-apposing metal stents with or without coaxial plastic stents for treatment of walled-off necrotizing pancreatitis: a prospective bicentric randomized controlled trial

Background and aims: Lumen-apposing metal stents (LAMSs) have proven to be effective for drainage of pancreatic walled-off necrosis (WON), although associated adverse events (AEs) have been reported. Anchoring coaxial double-pigtail plastic stents (DPSs) within LAMSs have been proposed to prevent LAMS-related AEs but have not been assessed in prospective studies. We aimed to evaluate the utility of such measures with a randomized controlled trial.

Methods: We randomly assigned consecutive patients with WON indications for drainage to EUS-guided transluminal drainage using LAMSs with (group A) or without (group B) DPSs. All LAMSs were to be removed after 3 weeks had elapsed from the index procedure with a preceding CT to decide whether additional steps needed to be taken (eg, transluminal necrosectomy or placing transluminal plastic stents in patients with incomplete resolution of WON). The main outcomes were failure of the index method, defined as necessity of reintervention (endoscopic, percutaneous, or surgical) before LAMS removal because of LAMS-related AEs and/or clinical deterioration; AE rates; and mortality with the LAMS in place. Variables were evaluated using the Mann-Whitney U test, χ² test, or Fisher exact test as appropriate. P < .05 was considered significant.

Results: Sixty-seven patients (37.3% women; mean age, 54 ± 14.4 years) underwent LAMS placement with (n = 34) or without (n = 33) DPS placement in 2 tertiary centers. Baseline characteristics including demographics, etiology, comorbidity, and clinical presentation (sterile vs infected necrosis) were comparable between both groups. The technical success rate in placing LAMSs and DPSs was 100%. The global rate of AEs was significantly lower in group A versus group B (20.7% vs 51.5%, respectively; P = .008). Stent occlusion was the most frequently observed AE (14.7% vs 36.3%, P = .042). Failure of the index method was lower in group A versus group B (29.4% vs 48.5%, respectively; P = .109); however, the difference did not achieve statistical significance. The same applied to the mortality rate with LAMSs in place (2.9% vs 12.1%, P = .197).

Conclusions: The addition of a coaxial DPS within a LAMS was associated with a significantly lower global rate of AEs and stent occlusion rate in EUS-guided drainage of WON. (Clinical trial registration number: NCT03923686.).