Association of perioperative allogeneic blood transfusions and long-term outcomes following radical surgery for gastric and colorectal cancers: systematic review and meta-analysis of propensity-adjusted observational studies

Abstract

Abstract Background Previous meta-analyses reporting significant associations between perioperative allogeneic blood transfusions and poor prognosis in gastric cancer or colorectal cancer had a high risk of confounding bias. This meta-analysis explored this issue using observational studies that applied propensity score analysis. Methods PubMed, Embase, and the Cochrane Central Register of Controlled Trials were searched for manuscripts published between 2013 and 2022. Studies applying propensity score analysis were included to investigate the association between perioperative allogeneic blood transfusions and prognosis in gastric cancer or colorectal cancer after radical surgery. Pooled HRs for overall survival and disease-free survival were calculated using a fixed-effect model or random-effect model according to heterogeneity. Results Twelve retrospective cohort studies with 17 607 patients reported were included. Ten studies applied propensity score matching and two applied inverse probability of treatment weighting using propensity score. A total of 5962 patients were analysed after propensity score adjustment. After propensity score adjustment, perioperative allogeneic blood transfusions did not correlate with disease-free survival in gastric cancer (HR 1.16; 95 per cent c.i. 0.96–1.39; heterogeneity was assessed by the chi-squared test and inconsistency index (I2) = 57 per cent) or colorectal cancer (HR 1.12; 95 per cent c.i. 0.84–1.49; I2 = 54 per cent). However, after propensity score adjustment, perioperative allogeneic blood transfusions were significantly associated with worse overall survival in gastric cancer (HR 1.20; 95 per cent c.i. 1.08–1.32; I2 = 25 per cent) and colorectal cancer (HR 1.40; 95 per cent c.i. 1.06–1.85; I2 = 52 per cent). Subgroup analyses showed that perioperative allogeneic blood transfusions did not correlate with overall survival in colorectal cancer when major postoperative complications were balanced after propensity score. Conclusion Perioperative allogeneic blood transfusion is not correlated with recurrence of gastric cancer and colorectal cancer. Perioperative allogeneic blood transfusions are significantly associated with worse overall survival in gastric cancer and colorectal cancer, which may be attributable to unbalanced major postoperative complications after propensity score adjustment.