Association between transfusion with plasma and the recurrence of colorectal carcinoma.

Data was collected from patients with colonic or rectal tumours undergoing potentially curative surgery. Univariate analysis showed an association between perioperative transfusion with plasma and the earlier recurrence of tumour. Allowing for possible confounding factors the relative risk of recurrence for patients receiving plasma was 2.44 (95 per cent confidence interval = 1.10-5.46, P = 0.041) and for patients receiving whole blood or plasma, it was 2.39 (95 per cent confidence interval = 1.19-4.83, P = 0.021). The risk of recurrence for those receiving a transfusion of plasma was, therefore, more than twice that of those not given it. We were unable to demonstrate an independent harmful effect of transfusion with packed cells. These data suggest that the plasma protein, rather than the cellular component of whole blood, mediates the accelerated tumour recurrence.