Surgical gastrojejunostomy or endoscopic stent placement for the palliation of malignant gastric outlet obstruction (SUSTENT study): a multicenter randomized trial

BackgroundBoth gastrojejunostomy (GJJ) and stent placement are commonly used palliative treatments of obstructive symptoms caused by malignant gastric outlet obstruction (GOO).ObjectiveCompare GJJ and stent placement.DesignMulticenter, randomized trial.SettingTwenty-one centers in The Netherlands.PatientsPatients with GOO.InterventionsGJJ and stent placement.Main Outcome MeasurementsOutcomes were medical effects, quality of life, and costs. Analysis was by intent to treat.ResultsEighteen patients were randomized to GJJ and 21 to stent placement. Food intake improved more rapidly after stent placement than after GJJ (GOO Scoring System score ≥2: median 5 vs 8 days, respectively; P < .01) but long-term relief was better after GJJ, with more patients living more days with a GOO Scoring System score of 2 or more than after stent placement (72 vs 50 days, respectively; P = .05). More major complications (stent: 6 in 4 patients vs GJJ: 0; P = .02), recurrent obstructive symptoms (stent: 8 in 5 patients vs GJJ: 1 in 1 patient; P = .02), and reinterventions (stent: 10 in 7 patients vs GJJ: 2 in 2 patients; P < .01) were observed after stent placement compared with GJJ. When stent obstruction was not regarded as a major complication, no differences in complications were found (P = .4). There were also no differences in median survival (stent: 56 days vs GJJ: 78 days) and quality of life. Mean total costs of GJJ were higher compared with stent placement ($16,535 vs $11,720, respectively; P = .049 [comparing medians]). Because of the small study population, only initial hospital costs would have been statistically significant if the Bonferroni correction for multiple testing had been applied.LimitationsRelatively small patient population.ConclusionsDespite slow initial symptom improvement, GJJ was associated with better long-term results and is therefore the treatment of choice in patients with a life expectancy of 2 months or longer. Because stent placement was associated with better short-term outcomes, this treatment is preferable for patients expected to live less than 2 months. (Clinical trial registration number: ISRCTN 06702358.)