A relationship between perioperative blood transfusion and recurrence of carcinoma of the sigmoid colon following potentially curative surgery.

Preoperative blood transfusions are used to improve graft survival in renal transplantation. If such an immunomodulating effect occurred in cancer surgery perioperative blood transfusion may be detrimental to patient outcome. A retrospective study of 68 patients undergoing potentially curative surgery for adenocarcinoma of the sigmoid colon, over a 10 year period was performed. Thirty-three patients (49%) had a perioperative blood transfusion of which two-thirds received either one or two units. Transfused patients had a poorer prognosis compared to non-transfused patients (0.28 and 0.53 five year product limit recurrence free fractions respectively; P less than 0.01 on generalised Savege test of entire recurrence free curves). Perioperative transfusion was the most sensitive prognostic indicator of recurrence on Cox proportional hazards regression analysis (relative risk 2.6; P less than 0.01, after adjustment for histological stage). Although a causal relationship is not proven, prospective work is urgently needed.