The prognostic role of perioperative allogeneic blood transfusions in gastric cancer patients undergoing curative resection: A systematic review and meta-analysis of non-randomized, adjusted studies.

The impact of allogeneic perioperative blood transfusions (APTs) on the prognosis of gastric cancer patients undergoing curative-intent gastrectomy is still a highly debated topic. Two meta-analyses were published in 2015, and new studies report conflicting results. A literature review was conducted using PubMed, Scopus, the Cochrane Central Register of Controlled Trials and the Cochrane Database of Systematic Reviews, updated to March 1, 2016. Thirty-eight non-randomized studies reporting data on overall survival (OS), disease-free survival (DFS), disease-specific survival (DSS) and postoperative complications (PCs) were included. An inverse variance random-effects meta-analysis was conducted. APTs showed an association with worse OS, DFS, DSS and an increased number of PCs. The hazard ratio (HR) for OS was 1.49, with a 95% confidence interval (95% CI) of 1.32-1.69 (p < .00001; Q-test p = .001, I-squared = 56%). After outlier exclusion, the HR for OS was 1.34 (95% CI = 1.23-1.45, p < .00001; Q-test p = .64, I-squared = 0%). The HR for DFS was 1.48 (95% CI = 1.18-1.86, p = .0007; Q-test p = .31, I-squared = 16%), and the HR for DSS was 1.66 (95% CI = 1.5-2.19, p = .0004; Q-test p = .96, I-squared = 0%). The odds ratio for PCs was 3.33 (95% CI = 2.10-5.29, p < .00001; Q-test p = .14, I-squared = 42%). This meta-analysis showed a significant association between transfusions and OS, DFS, DSS and PCs. The quality of the evidence was low. Aggregation, selection and selective reporting bias were detected. The biases shifted the results towards significance. Further studies using accurate adjustment methods are needed. Until such additional studies are performed, caution in administering transfusions and optimization of cancer patient blood management are warranted.