A 74 year old female is going for heart surgery and her pre-operative testing reveals a warm auto-antibody with "e" specificity. No unexpected red cell alloantibodies are detected. Upon Rh phenotyping, the patient is e-positive and E-negative. The inquiring colleague wonders if any hospital transfusion service would select donor RBC units that are 'e' negative to 'respect' the anti-e autoantibody, realizing that such donor RBC units would likely be E-positive and present a risk of alloimmunization to the E antigen. Or, would any colleague select E-negative RBCs for this patient (even though such RBC units would likely be e-positive and show a reactive crossmatch due to the autoanti-e), in order to try to prevent development of a future anti-E?

The following comments have been received.

ADDENDA June 6, 2006

1. A transfusion medicine physician at an academic medical center in Boston reports that in this particular example, he would want to know if there was any evidence that the auto-antibody was clinically significant. For example, if the patient was anemic, with an elevated reticulocyte count, low haptoglobin or other signs of shortened red cell survival, then he would direct his transfusion service to use e-negative donor RBC units. However, if the autoantibody was a serologic phenomenon only, without inducing clinical evidence of shortened red cell survival, he would direct his transfusion service to use donor units that were matched at ABO and D only.

2. A colleague at a medical center in Massachusetts (different than the Boston hospital in ADDENDUM #1) reports that her hospital sees a large population of patients with warm autoantibodies. If faced with a similar situation like that described in this discussion, her transfusion service would select e-negative donor RBC units, even if it meant exposing the patient to E-positive red cells. Interestingly, their experience has been that in most cases the patient with autoanti-e does not make anti-E following transfusion of e-negative red cells. She acknowledged one case where a patient with an autoanti-e did go on to make anti-E following transfusion of e-antigen negative RBCs. When that happened, they were faced with determining whether to give e-negative or E-negative RBC units. Given the clinical situation at the time, they felt that the autoanti-e was more clinically significant, so they gave the patient e-negative RBC units.

3. Dr. Connie Westhoff, Scientific Director for the American Red Cross Molecular Blood Group and Platelet Antigen Testing Laboratory (attribution used with permission) reports that in a case like the one under discussion, it is important to establish if the anti-e is an autoantibody rather than an alloantibody. This is because a large number of variants of RHCE at the gene level are now known, suggesting that there may be polymorphic differences responsible for some or many of the anti-e previously classified as "auto" antibodies. This is especially true in individuals with Black, Hispanic or mixed race backgrounds, but Caucasian variants are also a possibility. She points out that it would be reasonable to categorize an antibody as an autoantibody if the patient has not been transfused, is e-positive, has a positive direct antiglobulin test (DAT), and an eluate study shows anti-e. Even if the DAT is not positive, if the antibody can be adsorbed completely on the patient's own autologous red cells, it would be considered an autoantibody. But the bottom line is that sometimes the distinction between autoantibody versus alloantibody is not so clear cut because of the well-known autoantibody production that often precedes alloantibody specificity. She raises a caveat, in that her laboratory has seen anti-D or anti-e (in serologic D positive or e positive people) appear to be auto (by adsorption on autologous cells initially), but down the road become alloantibody. These individuals by molecular testing carried D or e gene mutations, so were truly alloimmunized.

She also reports that her laboratory has seen several recent cases where R2R2 units have been selected for transfusion in patients with apparent autoanti-e, with different outcomes ... from production of anti-E, no antibody production, to stimulation of anti-Jka but not anti-E. In the patients who require long-term transfusion support, extended antigen matching in other systems (K, Kidd, Fy) might be prudent to prevent augmentation of anti-E.

4. An experienced transfusion medicine physician in Los Angeles comments that this case represents one of the 'classic transfusion conundrums'. In her experience, when there is evidence that auto anti-e is causing premature destruction of either the patient's autologous red cells or transfused e-positive red cells, she would have her transfusion service select e-negative (R2R2) donor RBC units for transfusion. Furthermore, in a case where a patient is going for cardiac surgery, if the need for multiple RBC units is likely, she would select e-negative R2R2 RBC units to 'get better crossmatch compatibility', although this would 'load up' the patient with E-positive donor RBC units, which might cause a delayed hemolytic reaction in case the patient did develop anti-E. She acknowledges that supply issues for R2R2 RBC units may come into play in the decision making process. In the post operative period, if a top up RBC transfusion is needed, she might try using an E-negative RBC unit (which would likely be e-positive) to see how well it survives. It has been her experience that if the patient with auto anti-e receives RBC transfusions (even if the units are e-negative) that following the transfusion the specificity of the autoantibody eventually 'evolves' into a 'panagglutinin', so that there is less of a reason to give e-negative RBC units at that point, and possibly more of a reason to give E-negative RBC units as her immune system has been stimulated to make red cell antibodies.

ADDENDA June 7, 2006

5. A transfusion medicine physician in Maryland reports that at his transfusion service, if faced with a similar situation as that described in this discussion, if the auto anti-e was behaving as a "laboratory finding" only (i.e., there was little or no ongoing hemolysis), he would not be particularly concerned about using "e" positive donor RBC units, as they should have a normal or near normal survival in this patient (provided there is no ongoing hemolysis), and the auto anti-e would not likely cause a "transfusion reaction". Thus, given the aforementioned assumptions, he would anticipate no clinical problem from the autoantibody. However, he would not be opposed to transfusing R2R2 cells if it makes the doctors happier to give blood that is 'compatible' with the autoantibody. He would be more inclined to use e-negative RBC donor units for a patient who is actively hemolyzing due to auto anti-e, and who required ongoing transfusion, since e-negative cells would have a survival advantage and decrease the need for transfusion, and alloimmunization to "E" may not occur. In addition, managing an elderly patient with an anti-E is a common, easily addressed issue. However if a patient with auto anti-e was Rh negative (i.e., cde/cde), the responding colleague would not wish to expose the patient to RhoD,  in order to avoid transfusing e-positive RBC units to a patient with auto anti-e.

6. According to a Section Head of the Red Cell Immunohaematology (RCI) laboratory of the National Blood Service (NBS), UK (attribution used with permission), it is the policy of UK NBS RCI laboratories to match a patient's own Rh phenotype even when specific Rh autoantibodies are present by AHG technique, provided all other clinically significant alloantibodies have been excluded. Consequently, regarding a patient with auto anti-e, whose Rh phenotype is E negative and e positive, they would provide RBC units that are E-negative and e-positive, to match the patient's phenotype. The autoantibody would only be accomodated if it were suspected that the patient was suffering from active hemolysis caused by the autoantibody.

7. A transfusion medicine physician in Chicago reports that in his practice, if an Rh negative (cde/cde) patient had auto anti-e, his bias would be to select D- RBC units, particularly for a woman, because of the possibility of inducing or restimulating previous latent Rh sensitization and triggering delayed immune hemolysis of D+ RBCs. If the patient is Rh positive and E-negative, and if the autoantibody is only 'relative' in its anti-e autospecificity, i.e, it reacts with e-negative cells but weaker than it does with e+ cells, he would not select e-negative donor RBCs, because in his experience in such a situation the benefit of using e-negative donor RBCs is dubious. He adds that compared to the halcyon days of immunohematology, the modern-day use of gel or solid-phase testing probably enhances the reactivity of warm autoantibodies and might make such antibodies less likely to look 'cleanly' antigen-specific. He adds that in his practice, he has not selected E-negative RBCs prophylactically in this circumstance. If the auto anti-e is truly antigen-specific, then the textbooks (e.g., Mollison's 11th, pg 269) cite past evidence that e+ RBCs have a shorter survival, and hence the dilemma. If the patient has laboratory evidence of autoimmune hemolysis, then he would strive for selecting e-negative RBC units, and 'take a chance' with later development of anti-E. If there is no obvious hemolysis, then his preference (bias?) would be to select e-negative donor RBCs, if they are available, because of the evidence for better RBC survival--but if e-negative RBCs are not readily available, then he would not insist on them. Some reference labs may be reluctant to release R2R2 RBCs for patients with autoantibodies.

8. It is the policy at an academic medical center in Michigan to release e-positive 'incompatible' random donor RBC units in this situation, once adsorption studies have been done to exclude underlying alloantibodies. The exception is when the patient is actively hemolyzing because of the autoantibody and does not respond to such transfusions. Then they consider the use of R2R2 units on a case-by-case basis. According to the respondant, the literature states that use of 'antigen-negative blood' when "relative" anti-e is present may be of benefit in the hemolyzing patient. However, they think twice about giving R2R2 blood to an Rh-negative premenopausal hemolyzing female with relative anti-e, due to the likely alloimmunization to D. The respondant recalls that a few years back they had such a case and were fortunate to obtain two units of r"r" RBCs, but the transfusion of those RBC units did not produce the expected rise in Hb/HCT.

ADDENDA June 8, 2006

9. A transfusion medicine physician in San Diego reports that they manage patients with 'auto anti-e' on a case by case basis. In his experience, the vast majority of patients with auto-antibodies do not hemolyze their own red cells or donor RBC units to any clinically significant extent, and many patients with auto-antibodies also have allo-antibodies, which limits one's options. So, if the patient is not hemolysing significantly, they would not use e-negative blood. If the patient is Rh-negative, they would not, in any case expose the patient to D-positive R2R2 blood. If, however the patient is Rh-positive, and the anti-e is truly a clinically significant antibody (rapidly dropping Hb, etc) and they are not limited in blood selection by allo-antibodies, then he would request R2R2 units, to the extent they are available (not always possible if there are major transfusion needs). Finally, with rare exceptions, he has observed that placing the patient who has a clinically significant auto-antibody on steroids (1 mg/kg prednisone or equivalent) quickly diminishes significant hemolysis, so one may be more comfortable giving e-positive RBC units. One of the major problems he faces in such patients is that they have already been transfused elsewhere when they first see them, so it may be difficult to quickly assess whether the anti-e is really an auto-antibody. Multiple allo-absorptions may be useful. They have also considered using molecular phenotyping to discover the patient's true phenotype, but that typically takes too much time to be useful. Perhaps it will be more applicable in the future.

10. A doctoral scientist in Toronto who 'invented' the popular reagent ZZAP reports that "generally, the survival of donor red cells reactive with a patient's autoantibody survive about as well as the patients own red cells and it is not recommended to provide 'antigen-negative' blood which is often hard to find." As pointed out, in the case of auto-anti-e where e-negative blood is easier to find than with most other autoantibodies, the risk is that you expose the patient to E antigen and future risk of delayed hemolytic transfusion reaction. Data on the survival of e-positive blood in patients having auto-anti-e are sparse. However, there are a few publications indicating better survival of e-negative blood in these patients. This has led some investigators to suggest that if the titer of the auto-anti-e is at least two tubes higher against e-positive versus e-negative cells that antigen-negative blood should be used. Of course, this will present the risk of alloimmunization to E and patients having AIHA have an increased ability to produce alloantibodies when exposed to the antigen. After making sure that the patient has no significant alloantibodies (auto- or differential absorption techniques), the respondent would not worry about the e-antigen and transfuse without regard for the autoantibody using smaller volume transfusions and monitor the Hb/Hct. If there is evidence of significant increased cell destruction or the rise in Hb/Hct is unacceptable, he would then switch to e-negative blood. For a complete review of this situation and additional rationale for the selection of blood in this particular situation, check:Petz LD, Garratty G, Immune Hemolytic Anemias, 2nd Edition, 2004, pp385-386.

Please submit comments to the e-Network Forum.

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