BackgroundSelf-expandable metal stent (SEMS) insertion and percutaneous gastrostomy (PG) feeding are commonly used for patients with esophageal cancer and dysphagia. This study aimed to compare outcomes between SEMS insertion and PG feeding for them.MethodsWe retrospectively analyzed 308 patients with esophageal cancer who underwent fully covered SEMS insertion (stent group) or PG (gastrostomy group) for dysphagia due to tumor. Patients with other causes of dysphagia, such as radiation-induced or postoperative stricture, were excluded from the study. Clinical outcomes were compared between the two groups, including overall survival and need for additional intervention and postprocedural nutritional status.ResultsAt baseline, the stent group (n = 169) had more stage IV patients, less cervical cancers, and received radiotherapy and esophagectomy less often than the gastrostomy group (n = 64). The Kaplan-Meier curves showed higher overall survival in the gastrostomy group than in the stent group. Multivariate analysis revealed that PG was associated with better survival compared with SEMS insertion (hazard ratio 0.541, 95% confidence interval 0.346–0.848, p = 0.007). In addition, the gastrostomy group needed additional intervention less often (3.1% vs. 21.9%, p < 0.001) and experienced less decrease in serum albumin levels (-0.15 ± 0.56 g/dL vs. -0.39 ± 0.58 g/dL, p = 0.011) than the stent group after procedure.ConclusionsOur data suggested that, compared with SEMS insertion, PG is associated with better overall survival in patients with esophageal cancer and dysphagia. Stabilized nutritional status by PG may play a role in improving patient survival.