Randomised comparison of leucocyte-depleted versus buffy-coat-poor blood transfusion and complications after colorectal surgery

Authors: Jensen LS, Kissmeyer-Nielsen P, Wolff B, Qvist N.
Source: Lancet. 348:841-45. September 28, 1996.
Institution: Aarhus University and Odense University Hospital, Denmark.
Financial support: None indicated.



Blood transfusion leads to a depression of immune function which is incompletely understood and which may be related to the transfusion of leucocytes in allogeneic blood and blood products. This depression of immune function can lead to a susceptibility to infection (and possibly to cancer or cancer recurrence).

Blood component transfusion is generally "buffy-coat-poor", which removes about two thirds of the leucocytes present in whole blood. Using special filters, leucocyte-depleted blood contains only about 0.1% of the leucocytes in whole blood. If the depression of immune function is indeed related to the presence of leucocytes in the transfused blood products, leucocyte-depletion using filtration should significantly lessen this problem. This study was designed to look at this question in the setting of colorectal surgery. Patients were randomized to either traditional, buffy-coat-poor blood, or filtered, leucocyte depleted blood in the event of transfusion. Infectious complications were then compared among those not transfused and those transfused with either of the two types of blood products.



Authors' Discussion

The authors note the markedly increased rate of postoperative infections in patients who had received intraoperative and postoperative buffy-coat-poor blood transfusions, but not in those who had received leucocyte-depleted blood (who had infections at a rate similar to the non-transfused patients). This effect was not noted in patients who only received pre-operative transfusions.

They note that a previous study with a similar design by Houbiers et al (1) did not show any benefit with leucodepletion, but state that "their findings have been criticized, especially because of their multicentre design".

Finally, they state that although their study was done using a bedside filter, filtration done in the bloodbank is also an option, allows quantitative quality control and can be cheaper.


This randomized trial shows a very significant reduction in infectious complications after elective colorectal surgery when patients requiring transfusion are given filtered, leucocyte-depleted blood rather than conventional, buffy-coat-poor packed cells.

These results are in accord with a previous study co-authored by one of the authors of this study (2) which found that colorectal surgery patients transfused with leucocyte-depleted blood had significantly fewer postoperative infections than those transfused with whole blood.

On the other hand, a study with a design very similar to this one by Houbiers et al (referenced above) found no benefit in terms of infectious complications when patients undergoing surgery for colon cancer were given leucocyte depleted transfusions. Objections raised by Jensen and others to the Houbiers paper in several letters to the editor (3) include the multicenter nature of that trial and the high infection rate in all arms of the study. These objections are dealt with by Houbiers et al in their response. The differing results of these trials remains to be explained.

In a review of the detrimental effects of perioperative blood transfusion (4), Nielsen offers a speculative explanation of the negative results of the Houbiers trial -- the release of immunosuppressive agents from leucocytes during storage. If, in fact, certain conditions of storage cause leucocytes to release the factors responsible for immunosuppression into plasma, then leucocyte-depleting filtration might be ineffective.

Clarification of this point is obviously important. If it turns out that certain conditions of storage negate the beneficial effects of leucocyte depletion, these conditions must be avoided to reap the potential benefits of this technique, which presumably extend to circumstances other than colorectal surgery.


Reader comments

October 28, 1996

Dr. Lone S. Jensen, corresponding author of this study, sent me a letter pointing out an error in my summary concerning the estimated blood loss in the four groups. I mistakenly added numbers that should not have been summed. This has been rectified and the numbers should now be correct.

-- mj

From the authors of the Houbiers article referenced above:

Date: Mon, 3 Feb 1997
From: Leo van de Watering <lvdwat@dsl.nl>

We've written our letter to the Lancet to try and clarify some of the confusion following the publications of our and Jensen's studies. Both studies performed a comparison between colorectal surgery patients transfused with buffy coat depleted blood and patients transfused with leukocyte depleted, filtered, blood. In contrast to Jensen and colleagues we had found an increase in postoperative infections in all transfused patients and therefore postulated that transfusion associated postoperative infections are based on two factors: One related to the transfusion of red cells and the other associated with the transfusion of high absolute numbers of leukocytes.

The generally accepted critical leukocyte load of 1 x 106 per unit, as mentioned by Jensen, is based on alloimmunisation studies in rabbits which results were extrapolated to men. Although no comparative studies about the threshold leukocyte contamination with regard to alloimmunisation have been performed, empirical studies show that leukocyte reduction below 5 x 106 per unit red cells/platelets prevents primary immunisation. Since the prevention of alloimmunisation is the main indication for using filtered blood this threshold is most commonly reported. However, for the prevention of Non-Hemolytic-Febrile-Transfusion-Reactions (NHFTR) buffy coat removal resulting in 1 x 109 leukocytes per unit may already be sufficient while on the other hand Transfusion-Associated-Graft-Versus-Host-Disease (TA-GvHD) even has been observed after transfusion of filtered blood containing less than 1 x 106 leukocytes per unit. What threshold is important for postoperative infections is still unknown.

Several studies have reported a dose-response relation between blood transfusions and postoperative infections. Jensen reported in her table 5 an average blood transfusion requirement of 90/17 = 5.3 units in infected patients, with a median of 4 units. Since the buffy coat depleted red cells used in Denmark contained 50% more leukocytes per unit than those used in Leiden (1.2 vs 0.8 x 109), to reach a same amount of leukocytes 6 to 8 units should have been given in our study. The number of patients receiving this amount of leukocytes in our study was too small to show a difference. In a new study comprising over 900 cardiac surgery patients with an average blood transfusion requirement of more than 5 units, we do find a significant difference in postoperative infections between patients receiving buffy coat depleted blood and patients receiving filtered units.

We conclude that the total number of leukocytes transfused may explain the differences in outcome and should be taken into account in future studies. Other factors that are not yet taken into account in all clinical studies are the conditions of the leukocytes transfused i.e. the presence of leukocyte/platelet aggregates and the expression of adhesion molecules and activation markers. These may all be influenced by the storage duration and storage conditions.

Leo MG van de Watering, JGA Houbiers, CJH van de Velde, A Brand

Leiden University Hospital
Dept. of Immunohematology & Blood Bank
Dept. of Surgery
Leiden, the Netherlands

April 2, 1997

For a review of the conflicting evidence concerning transfusions, leucocyte depletion and post-operative infections, see an editorial in Transfusion (February, 1997: 37:121-25) by M. Blajchman entitled "Allogeneic blood transfusions, immunomodulation, and postoperative bacterial infection: do we have the answers yet?" Dr. Blajchman summarizes the evidence and argues that the primary data from six trials should be re-analyzed in a meta-analysis.

Submit a comment about this article


1. Houbiers JG, Brand A, van de Watering LM, Hermans J, Verwey PJ, Bijnen AB et al. Randomised controlled trial comparing transfusion of leucocyte-depleted or buffy-coat-depleted blood in surgery for colorectal cancer. Lancet 344: 573-578 (1994)

2. Jensen LS, Andersen AJ, Christiansen PM, Hokland P, Juhl CO, Madsen G et al. Postoperative infection and natural killer cell function following blood transfusion in patients undergoing elective colorectal surgery. Br J Surg 79: 513-516 (1992)

3. Letters to the editor. Lancet 344: 1429-31 (1994)

4. Nielsen HJ. Detrimental effects of perioperative blood transfusion. Br J Surg 82: 582-587 (1995)

More references related to this summary from the NLM's PubMed database.

Return to Journal Club home page

Site Meter