Hemostatic Response to Palliative Radiotherapy for Bleeding Gastric Cancer and Pretreatment Endoscopic...

Objective As no prior studies have examined the association between the pretreatment endoscopic or histopathological findings and the hemostatic efficacy of palliative radiotherapy (pRT), we investigated the predictors of response and rebleeding in gastric cancer (GC). Methods This retrospective study included consecutive patients with bleeding GC who underwent hemostatic pRT between 2015 and 2023. The primary endpoint was the hemostatic response within 4 weeks after pRT, defined by a composite endpoint: hemoglobin ≥8.0 g/dL, ≥50% reduction in red blood cell transfusion versus the prior 4 weeks, no salvage hemostatic intervention, and clinical stabilization. The secondary endpoints included survival outcomes, adverse events, and the biologically effective dose (BED₁₀). Patients Twenty-three patients with bleeding GC treated with pRT were analyzed. Results A hemostatic response was achieved in 17 (73.9%) patients. The pretreatment endoscopic and histopathological factors were not associated with the initial response. The median BED₁₀ was higher in responders than in non-responders (28.0 vs. 14.4 Gy), indicating a non-significant trend (Hodges-Lehmann median difference, 13.0 Gy; 95% bootstrap confidence interval, 0-24.6). Diabetes mellitus was more frequent in non-responders. Among the responders, a circumferential tumor extent of ≥1/2 of the lumen was significantly associated with earlier rebleeding (median rebleeding-free survival, 9.5 months; p=0.045). No adverse events of grade ≥3 occurred. Conclusion In this exploratory analysis, pRT achieved effective hemostasis in patients with bleeding GC. No pretreatment endoscopic or histological features were significantly associated with the initial hemostatic response. Extensive circumferential involvement identified responders at an increased risk of rebleeding, and a higher BED₁₀ showed a non-significant trend toward improved response without increased toxicity.