In current practice the transfusion of red cells is almost always ABO identical, except in emergencies when a recipient's blood group is unknown or there is a shortage of ABO identical red cells. However, it is routine in many transfusion services to administer platelets that are ABO non-identical with the recipient and to use AB FFP as universal donor plasma. Dr. Neil Blumberg, Professor of Pathology & Laboratory Medicine, Director of Clinical Laboratories at the Strong Memorial Hospital, and Director of Transfusion Medicine/Blood Bank at the University of Rochester Medical Center asks us to reconsider using ABO non-identical platelet and plasma transfusions, given the evolving understanding of the biology of the ABO blood group system.

Dr. Blumberg points out that in our efforts to avoid hemolytic reactions, we do not administer group O whole blood as universal donor blood in emergencies, but we frequently ignore the 20-30 mL of incompatible plasma and administer group O red cells as universal donor, even in non-emergent, non-shortage situations, to group A, B or AB patients. In his experience, the administration of even small amounts of group O plasma to a non-O individual can cause severe hemolysis. Thus, in his opinion, Group O red cells should not be routinely given to non-O recipients, except in emergencies or shortages. In addition, there are other reasons to avoid infusing even a small amount of ABO incompatible plasma, as discussed below.

Unlike the minimal amount of plasma in group O red cells, group O platelet concentrates contain about a quarter of a liter of incompatible plasma, and such transfusions are often repeated daily or every other day.  Dr. Blumberg argues that here there is good evidence from controlled clinical trials that infusion of incompatible plasma or cells on a repeated basis leads to poorer platelet count increments, increased transfusion requirements for both red cells and platelets, and perhaps increases the risk of more serious complications, such as multi-organ failure and death in surgical patients, and recurrence of leukemia (data reviewed in Transfusion 39: 1155, 1999, Vox Sanguinis 88:207, 2005 and an upcoming mini-review paper in Bone Marrow Transplantation). Recent unpublished data from his center suggests that administration of ABO non-identical platelets may be associated with increased bleeding and red cell utilization in both surgical patients and patients with leukemia. Exposure of antigen-bearing platelets to isoagglutinins impairs in vitro platelet function in some cases. There are sound theoretical reasons for not routinely infusing incompatible soluble antigen (e.g. AB FFP to an O recipient) or antibody (e.g. group O platelets to a group A recipient). This practice leads to the creation in vivo of high molecular weight circulating ABO immune complexes. Speculatively, abundant high molecular weight immune may stimulate inflammation, suppress cellular immune function, and interfere with platelet function. He reminds us that ABO antigens, in addition to their presence on red cells, and to varying degrees, on platelets, also are present in soluble form in plasma and blood fluids, and are present on all of the body's cells, including endothelial cells, macrophages, etc. that come in contact with the transfused plasma antibody.

Dr. Blumberg adds that while the data are still preliminary, and have not been replicated as yet by others, he has initiated policies in his transfusion service such that only ABO identical red cells, platelets, FFP and cryoprecipitate will be transfused except in emergencies or shortages. His transfusion service does not transfuse patients ABO non-identical components for inventory or logistic purposes unless there are critical shortages of the component in question. They use plasma depleted group O red cells and platelets (by washing in their case) when time permits if they need to use group O cells for a patient who is not group O. He acknowledges that many, if not most transfusion medicine practitioners, and the existing standards of practice, would suggest these practices are unnecessary. Nonetheless, while not suggesting that the data to date prove that these practices are superior, he believes the data are sufficient to argue that transfusing ABO identical components, or plasma depleted when necessary, is the safest practice for patients. This will not always be possible, but with some additional effort and expense, patients can be protected against rare episodes of hemolysis, and additional serious, if still hypothetical risks can be reduced.

The following comments have been received:

ADDENDA June 20, 2005

1. A director of a hospital blood bank in New York City strongly objects to the summarization at the beginning of this discussion that "In current practice the transfusion of red cells is almost always ABO identical". In his experience it is absolutely routine to give type O to non-O patients, and what is more, AB patients almost never receive type AB blood. To quote the director, "These are facts on the ground. I suggest your site issue a clarification immediately. If Dr. Blumberg wants to make a case about isoagglutinins, fine. However, there is still no reason not to give A or B plasma (or cryo) to O patients, and a simplistic statement about using ABO identical for all products is unconscionable. Regarding platelets, there is some info that A or B platelets may survive less well in O patients, but there is no such data about O platelets in A or B patients. The isoagglutinin issue has been thoroughly debated for years (consensus being that it is only a problem in a few type O's with high titer anti-A,B) and should only be revived with peer-review publication of new and convincing data."

Please submit comments to the e-Network Forum.

Ira A. Shulman, MD
CBBS e-Network Forum Editor & Moderator