ABSTRACT
The aim of this study was to evaluate whether there is a relationship between bleeding response and radiotherapy dose to palliate patients with local recurrence or progression of gastric cancer (GC). To this end, we conducted a systematic review and meta-analysis of observational studies that evaluated the bleeding response in patients with GC with local recurrence or progression. A meta-regression analysis between biological effective dose (BED) and bleeding response was performed, as was subgroup analysis to evaluate the outcome by BED level and radiotherapy (RT) technique. A p-value <0.05 was considered significant. Ten non-comparative retrospective studies and one prospective study were included. In general, RT was effective at controlling tumor bleeding, and the bleeding response rate was 0.77 (95% confidence interval (CI), 0.73-0.81). Meta-regression analysis demonstrated a linear correlation between BED Gy 10 and bleeding response (p=0<0001). Studies using conformational RT had a significant bleeding response rate compared to those using 2D (0.79; 95%CI, 0.74-0.84 vs 0.65; 95%CI, 0.56-0.75; p=0.021). In terms of the BED level, a significant difference in BR was identified on comparing BED Gy10 ≥40 (0.79; 95%CI, 0.7-0.8), BED Gy10 30-39 (0.79, 95%CI, 0.71-0.86), and BED Gy10 <30 (0.64; 95%CI, 0.5-0.7; p=0.0001). The mean survival time was 3.31 months (95%CI, 2.73-3.9) months, and the responders had a significantly longer survival (longer by 2.5 months) compared to the non-responders (95%CI, 1.7-3.3; p<0.0001). Palliative RT is effective at controlling bleeding due to local recurrence/progression from GC. Our findings reveal a relationship between BR and BED. BED <30 Gy 10 should not be recommended, and 3DRT should be indicated instead in order to improve the result.