Gastrojejunostomy versus endoscopic stenting for the palliation of malignant gastric outlet obstruction: a systematic review and meta-analysis
Journal Articles
Overview
Research
Identity
Additional Document Info
View All
Overview
abstract
BACKGROUND: Though gastrojejunostomy (GJ) has been a standard palliative procedure for gastric outlet obstruction (GOO), endoscopic stenting (ES) has shown to provide benefits due to its non-invasive approach. The aim of this review is to perform a comprehensive evaluation of ES versus GJ for the palliation of malignant GOO. METHODS: MEDLINE, Embase, and CENTRAL databases were searched and comparative studies of adult GOO patients undergoing ES or GJ were eligible for inclusion. The primary outcomes were survival time and mortality. Secondary outcomes included technical success, clinical success, reinterventions, days until oral food tolerance, postoperative adjuvant palliative chemotherapy, postoperative morbidities, length of stay (LOS), and costs. Pairwise meta-analyses using inverse-variance random effects were performed. RESULTS: After identifying 2222 citations, 39 full-text articles fit the inclusion criteria. In total, 3128 ES patients (41.4% female, age: 68.0 years) and 2116 GJ patients (40.4% female, age: 66.8 years) were included. ES patients experienced a shorter survival time (mean difference -24.77 days, 95% Cl - 45.11 to - 4.43, p = 0.02) and were less likely to undergo adjuvant palliative chemotherapy (risk ratio 0.81, 95% Cl 0.70 to 0.93, p = 0.004). The ES group had a shorter LOS, shorter time to oral intake of liquids and solids, and less surgical site infections (risk ratio 0.30, 95% Cl 0.12 to 0.75, p = 0.01). The patients in the ES group were at greater risk of requiring reintervention (risk ratio 2.60, 95% Cl 1.87 to 3.63, p < 0.001). CONCLUSION: ES results in less postoperative morbidity and shorter LOS when compared to GJ, however, this may be at the cost of decreased initiation of adjuvant palliative chemotherapy and overall survival, as well as increased risk of reintervention. Both techniques are likely appropriate in select clinical scenarios.
