Effect of perioperative blood transfusions on recurrence of colorectal cancer: meta-analysis stratified on risk factors.

Abstract

This study was undertaken to evaluate the influence of perioperative blood transfusions on colorectal cancer recurrence. All articles published up to December 1996 in English (or with an English abstract) were retrieved, both using MEDLINE and scanning their references, to be considered for this meta-analysis. One hundred thirty-one articles were identified, and 99 of them were excluded because they analyzed survival or mortality, were repetitive publications, or were reviews or letters. Thirty-two original studies (9 were prospective) on 11,071 patients were included for further analysis; 20 showed a detrimental effect of perioperative blood transfusions. Nineteen articles used also multivariable techniques, and 11 found perioperative blood transfusions to be an independent prognostic factor. Pooled estimates of the effect of perioperative blood transfusions on colorectal cancer recurrence yielded an overall odds ratio of 1.68 (95 percent confidence interval, 1.54-1.83) and a rate difference of 0.13 (95 percent confidence interval, 0.09-0.17) against patients who received transfusions. Stratified meta-analyses also confirmed these findings when stratifying patients by site and stage of disease. The effect of perioperative blood transfusion was observed in a dose-related fashion, regardless of timing and type, although some heterogeneity was detected. Data on surgical techniques were not available for further analysis. A consistently detrimental association was discovered between the use of perioperative blood transfusion and colorectal cancer recurrence. Further studies are needed to confirm that blood transfusion has a causal association.