Clinical Efficacy of Endoscopic Ultrasound-Guided Gastroenterostomy, Enteral Stenting, and Surgical Gastrojejunostomy for Malignant Gastric Outlet Obstruction: A Network Meta-Analysis.

BACKGROUND AND AIMS: Treatment options for malignant gastric outlet obstruction (MGOO) include endoscopic ultrasound-guided gastroenterostomy (EUS-GE), enteral stenting (ES), and surgical gastrojejunostomy (SGJ). We aim to provide a comprehensive comparative assessment of these interventions.

METHODS

We performed a systematic review and Bayesian network meta-analysis (NMA) of RCTs and cohort studies with causal adjustment comparing EUS-GE to ES, EUS-GE to SGJ, or SGJ to ES for MGOO (CRD420251133918). We searched PubMed, Embase, and Web of Science. Outcomes included re-intervention, clinical success, adverse events, length of hospitalization, time to oral intake, and technical success. Transitivity was assessed by comparing key effect modifiers, and consistency through Bayesian node-splitting. Sensitivity analysis restricted to RCTs only.

RESULTS

Six RCTs and ten causally adjusted cohort studies involving 1585 patients (376 EUS-GE, 726 ES, and 483 SGJ) were included. Baseline characteristics were comparable. EUS-GE demonstrated lower re-intervention rates than ES [risk ratio (RR) = 0.17; 95% credible interval (CrI), 0.07-0.35) and SGJ (RR = 0.3; 95% CrI, 0.11-0.65). Clinical success was higher with EUS-GE compared with ES (RR = 1.20; 95% CrI, 1.13-1.29) and SGJ (RR = 1.17; 95% CrI, 1.10-1.26). EUS-GE also shortened length of hospitalization (MD = -6.00; 95% CrI, -8.38 to -3.14) and time to oral intake (MD = -2.18; 95% CrI, -3.93 to -0.63) compared with SGJ. Based on direct pairwise comparisons, technical success did not differ among the three groups; however, network meta-analysis showed a higher technical success rate for SGJ compared with ES and EUS-GE. SUCRA rankings favored EUS-GE across most outcomes (re-intervention, clinical success, adverse events, length of hospitalization, and time to oral intake). Only minor inconsistency was detected, and sensitivity analyses were consistent with the primary analysis.

CONCLUSIONS

EUS-GE is associated with better clinical outcomes to ES and SGJ for MGOO. Future studies are needed to evaluate outcomes outside expert centers.