Assessment of Treatment Response and Re-Bleeding After Palliative Radiation Therapy for Bleeding Gastric Cancer: A Longitudinal Multicenter Prospective Observational Study

PURPOSE/OBJECTIVE(S) In the largest prospective study to date, we examined the effectiveness of palliative radiation therapy for bleeding gastric cancer, and the association of the biologically effective dose (BED) with survival, bleeding response, or re-bleeding. MATERIALS/METHODS We recruited patients with incurable gastric cancer for whom palliative radiation therapy was planned. Eligibility criteria included blood transfusion or hemoglobin levels < 8.0 g/dL. The primary endpoint was the intention-to-treat (ITT) response rate at four weeks. Response entailed all of the following criteria: 1) hemoglobin levels ≥ 8.0 g/dL, 2) seven consecutive days from enrollment to blood sampling without blood transfusion, and 3) no salvage treatment (surgery, endoscopic treatment, transcatheter embolization, or reirradiation) for bleeding gastric cancer. Re-bleeding was defined as the need for blood transfusion or salvage treatment. Response was assessed at 2, 4, and 8 weeks after enrollment; re-bleeding was assessed monthly during the 6-month follow-up. Sample size was designed with the probability of obtaining a 90% confidence interval (CI) ≤ ± 15% being set at 0.8, assuming a per-protocol (PP) response rate of 70% at 4 weeks. RESULTS We enrolled 55 patients from 15 institutions. A median total radiation dose of 20 Gy (range, 8-45 Gy) was delivered in a median of five fractions. Of the 55 patients, 38 responded to treatment; of the 38 responders, 19 (50%) died and 12 (32%) experienced re-bleeding. The ITT response rates were 47%, 53% (95% CI, 39%-66%), and 49% at 2, 4, and 8 weeks, respectively. The PP response rates were 55%, 78%, and 90% at 2, 4, and 8 weeks, respectively. Restricted mean response and re-bleeding time up to 8 months were 2.3 months and 0.8 months, respectively. The median overall survival was 3.8 months (median follow-up, 12.1 months). The multivariable Cox model showed that neither response (a time-dependent covariate) (hazard ratio [HR], 1.04; 95% CI, 0.44-2.48) nor BED (alpha/beta = 10) (HR, 0.98; 95% CI, 0.95-1.02) were significant predictors of overall survival. The multivariable Fine-Gray model showed that BED was not a significant predictor of response (subdistribution HR, 1.02; 95% CI, 0.99-1.06). The univariable Cox model on transition hazard from response to re-bleeding showed that BED was not significantly associated with re-bleeding (HR, 0.98; 95% CI, 0.92-1.04). Grade 1, 2, 3, and ≥ 4 radiation-related adverse events were reported in 11, 9, 1, and 0 patients, respectively. CONCLUSION The PP bleeding response rate increased to 90% during the 8-week follow-up period. The frequent occurrence of death from shortly after enrollment lowered the ITT response rate and shortened the duration of response. BED was not significantly associated with survival, bleeding response, or re-bleeding.