A blood banker in Canada reports that at her facility the practice has been to reduce the supernatant plasma of pooled platelets before issuing the product to ABO-incompatible adult recipients. They want to lower the level of anti-A and Anti-B in the product to decrease the amount of hemolysis due to ABO incompatibility between the donors and the recipient. The Canadian blood banker wants to know what is the practice at other institutions for transfusing ABO-incompatible platelet pools to adult recipients and why?

The following replies were received in response to the above question. Please note that there are sharp differences in opinion on the need to remove plasma from platelet units or if plasma should be removed, when it is appropriate to do so.

1. The e-Network Forum Editor suggests that the forum review a comprehensive discussion of volume-reduced platelets can be found in the archives of the e-network.

2. A hospital-based blood bank physician in Southern California is of the opinion that for adults it is unnecessary to reduce the volume of plasma in ABO-incompatible platelet units. She states that there is abundant A and B substance found on the surface of endothelial cells. Since the lining of every vessel in the body is studded with A and/or B substance (depending on the individual's blood group) it acts as a "sink" for the limited amount of anti-A and anti-B infused with incompatible plasma, effectively removing it from the circulation before significant hemolysis is able to occur. Actually, large amounts of incompatible plasma can be infused with little problem (at least from the standpoint of hemolysis due to anti-A or B).

3. A blood bank physician in New Mexico reports that most hospitals in that state are just glad to get platelets, although it is preferred that the platelets contain ABO-compatible plasma.

4. Another blood banker reported that her institution's policy is to allow up to 4 ABO-incompatible plasma transfusions via platelets per week. A platelet transfusion to an adult consists of either the administration of a pool of platelet concentrates or the administration of a unit of plateletspheresis. Transfusion number 5 within a 7-day period would be volume reduced to lower the load of isohemagglutinins. They allow the first 4 within a 7-day period due to the risk of platelet activation and actual platelet loss during the volume reduction process. The volume of the unmodified platelet product is 300 ml if a pool of 6 random platelet concentrates are used. Some of their units of plateletspheresis can have a volume in excess of 400 ml. The platelets are reduced to a volume of one half of the original volume.

5. A blood banker from the United Kingdom (UK) reports that in the UK this issue has been approached in 2 ways - guidelines for testing donations and education of hospital blood bank technologists. UK Guidelines for Blood Transfusion Services require all donations to be tested in order to identify group O donations that have a high titer of anti-A/B. This is done by testing a 1 in 100 dilution of plasma (in saline) with A₁B red cells on the automated blood grouping machine. Platelet donations containing high titer anti-A/B are labeled by the computer. Hospitals are familiar with this terminology and avoid giving donations with a high titer of anti-A/B to non-O patients. The responding blood banker reports that about 3-4% donations have a high titer of anti-A/B. Education of hospital blood bank technologists to select red cells, platelets and FFP of the patient's ABO group wherever possible. The responding blood banker reports that audits have shown that this practice is implemented, except where HLA-matched platelets of the patient's own ABO group are not available; a clinician will make a decision in these cases.

6. A blood banker from a hospital in Minnesota which performs many progenitor cell transplants reports that because they have so many patients that receive transplants, they transfuse a lot of platelets and often need to transfuse ABO plasma-incompatible platelets. Their policy is that patients may receive incompatible plasma volumes as follows:

• Age16 years and older - the patient can receive up to 1000 ml/wk
• Less than 16 years old - 40 kg or greater: the patient can receive up to 600 ml/wk; 25-39 kg: the patient can receive up to 400 ml/wk; 5-24kg: the patient can receive up to 100 ml/wk
• Neonates receive only plasma compatible platelets.

If a patient has reached their limit of incompatible platelets, plasma-incompatible platelets must be volume-reduced.

If for some reason a patient must receive compatible plasma, specific instructions are noted in patient history.

7. A blood bank physician in Rochester, NY reported (via the MEDLAB-L Digest list serv) that his hospital has an absolute policy for patients receiving multiple platelet transfusions for leukemia or lymphoma that they are given only ABO-identical transfusions or they remove most of the incompatible plasma (by washing in their case). They try to give ABO-identical platelets to everyone else, but do not always succeed because of inventory problems. Their reasoning is explained in an editorial in Transfusion entitled "The second century of ABO and now for something completely different" (Transfusion 1999; 39:1155-1159). The New York blood banker believes that it boils down to the fact that there is some evidence, not particularly conclusive, that infusing ABO-mismatched platelets is associated with increased morbidity and mortality in patients with hematological malignancies, and his group now has preliminary data that the same problem may exist in surgical patients (Blumberg, NS et al. Transfusion, June 2001). While not definitive, these results are compatible with the known deleterious effects of high levels of circulating immune complexes on inflammation and immune function. It is his hypothesis that the ABO blood group system is even more important in transfusion therapy than anyone had previously thought.

8. A blood banker from Thailand reports that several institutions in that country practice the same way as the inquiring Canadian's hospital.

ADDENDA April 3, 3002

9. Another UK blood banker reports that "the English National Blood service advises that platelet transfusions can often cross the ABO barrier. Group B platelets are often in short supply, especially group B Rh D negative, and after holidays. Under these circumstances, the choice after ABO-same (B) is ABO-opposite (A). Although this seems counter-intuitive (and may need a lot of explaining to nurses and to the patient) the rationale is that the anti-B in group A plasma, or anti-A in B plasma, is nearly always weakly reactive at 37C, and much less so than the anti-B and anti-A in the anti-A,B of group O plasma. Group O platelets are only advised if they are shown not to have a 'high-titer anti-A,B' - usually judged by the reactivity of a 1/100 dilution of donor plasma in the Olympus grouping machine (universally used for donor typing in England) in the presence of PEG or similarly 'encouraging' agglutinating agent, against A₁B cells. This is obviously a somewhat arbitrary definition. Apheresis platelets are often advocated - if available - because they tend to have fewer red cells."

10. A blood banker in Australia is also concerned about crossing the ABO barrier for platelet transfusions. This has two aspects to it as this can involve giving platelets that are incompatible by antigen, eg. group A platelets to a group O recipient, or giving platelets that are incompatible by plasma, eg. group O platelets to a group B recipient. He reports that their platelets pools are plasma-reduced and are resuspended in 'T-Sol' which lessens the impact of giving platelets incompatible by plasma group between donor and recipient. These plasma-reduced platelets have the added benefit of releasing more fresh plasma for fractionation. So if they are forced to give platelets across the ABO barrier they endeavour to use pools only and avoid apheresis platelets as far as possible. There are situations when so called 'incompatible' platelets are given by choice, especially in bone marrow or PBSC transplant situations where ABO mismatches occur between recipient and donor. In summary the Australian blood banker says they give ABO-compatible platelets whenever possible. If they are not available, they then choose pools with minimal incompatible plasma, and lastly they may give 'incompatible' platelets by choice in transplants.

ADDENDA April 8, 2002

11. A blood banker in Southern California reports that his institution has had two cases of moderately severe red cell hemolysis due to group O platelet transfusions to group A patients. In both cases the plateletpheresis donors were found to have high IgG titers of anti A.

• The first case was a 36 year old female patient with Hodgkin's disease, blood type A positive. Approximately 24 hours after receiving a 214 ml group O plateletpheresis, the patient demonstrated anti-A in the serum and eluate, drop in hemoglobin from 11.1 g/dL to 8.6 g/dL and a rise in the total bilirubin from 1.7 to 3.3. The IgG anti A titer of the plateletpheresis donor was found to be 1:2048.

• The second case was a 45 year old male patient with AML, blood type A positive. Approximately 48 hours post transfusion of a 241 ml group O plateletpheresis, the patient demonstrated anti-A in the serum and eluate, a drop in hemoglobin from 10.4 g/dL to 8.8 g/dL and a rise in total bilirubin from 0.7 to 3.1 mg/dL. The IgG anti A titer of this plateletpheresis donor was found to be 1:4096.

Based on the above two cases, the Southern California blood banker concludes that having weekly volume restrictions on the amount of incompatible plasma as suggested by previous discussants would not have protected these two patients. He wonders if the U.S. should consider adopting the U.K. standard of screening platelet donors to identify those with high anti -A or anti-B titers. Besides the risk of overt hemolysis from donors with high isoagglutinin titers, there may be other deleterious effects, evidence of which is only beginning to appear in the literature. For these reasons, the responding blood banker is of the opinion that the decision to transfuse out-of-group platelets should not be taken lightly.

ADDENDA April 14, 2002

12. A blood bank physician from Vermont reported that his transfusion service routinely volume-reduces ABO-incompatible platelets to prevent hemolysis due to anti-A or anti-B antibodies. He commented that an adverse reaction to ABO-incompatible platelets is a rare event, although dealing with a positive Coombs test and incompatible, type-specific red cells after transfusing ABO-incompatible plasma is annoying. The point is that his facility does not give patients ABO-incompatible plasma. He asks "Why should we give them ABO-incompatible plasma in the form of platelets, even if the risk of hemolysis is small? Occasional cases of hemolytic transfusion reactions in this scenario have been reported, although rarely, I admit."

Please submit comments to the e-Network Forum.

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