Endoscopic ultrasound-guided transmural gallbladder drainage using lumen-apposing metal stents versus percutaneous cholecystostomy for gallbladder drainage in high-risk cirrhosis patients with acute cholecystitis: a multicenter, comparative trial
摘要
Patients with cirrhosis and acute cholecystitis who are high risk for surgery undergo percutaneous cholecystostomy (PGBD) tube placement. EUS-guided gallbladder drainage (EUS-GB) using lumen-apposing metal stents (LAMS) is a newer intervention. We performed a multicenter cohort study to compare clinical outcomes and adverse events (AEs) of EUS-GB and PGBD in patients with liver cirrhosis who are unfit for surgery.
MethodsPatients with cirrhosis and acute cholecystitis who were unfit for surgery underwent gallbladder drainage by EUS-GB or PGBD. Technical success, clinical success, and adverse events were evaluated. Long-term AEs were stratified to biliary tract-related events and catheter-related/device-related AEs. Clinical success was reported on both an intention-to-treat (ITT) basis and a per-protocol basis. Inverse probability-of-treatment-weighted (IPTW) propensity-adjusted models were used to address baseline imbalance.
ResultsSixty-three patients with cirrhosis and acute cholecystitis underwent gallbladder drainage: 20 underwent EUS-GB and 43 underwent PGBD. The mean duration of follow-up was 19.9 weeks (range, 1–52). The etiology of the cholecystitis was gallstones (63%) and acalculous (37%). Technical success was achieved in 19/20 (95%) EUS-GB patients and 43/43 (100%) PGBD patients (p = 0.32). Procedure-related AEs were similar between groups (15% vs. 9.3%, p = 0.69). On ITT analysis, clinical success was 19/20 (95%) for EUS-GB and 40/43 (93%) for PGBD (p = 0.99). Long-term biliary tract-related AEs were lower in the EUS-GB group (10.5% vs. 23.3%, p = 0.018). When the secondary device-burden composite was added, the EUS-GB group continued to have significantly fewer long-term events (10.5% vs. 41.9%, p = 0.013). Among the 12 PGBD patients who underwent subsequent cholecystectomy, 7 (58%) were urgent operations driven by drainage failure and 5 (42%) were elective interval cholecystectomies. Surgery was required in a higher proportion of PGBD patients than EUS-GB patients (28% vs. 5%; p = 0.03).
ConclusionIn this multicenter retrospective cohort, EUS-GB with LAMS appeared to be an effective and safe alternative to PGBD for cirrhotic patients with acute cholecystitis who cannot undergo surgery, with fewer long-term biliary tract and catheter/device-related events.