George Garratty, PhD, FRCPath ARC Blood Services, Southern California Region and University of California, Los Angeles CBBS Today, 1998;16:22-26 (Fall)

Autoimmune hemolytic anemia (AIHA) can be associated with antibodies reacting optimally at 37°C [warm type AIHA (WAIHA)]; antibodies reacting optimally at 0-5°C but of higher thermal amplitude (cold agglutinin syndrome, and paroxysmal cold hemoglobinuria); and drug induced autoantibodies.^1,2 Approximately 70% of the AIHA are of the "warm" type, thus this review will concentrate on problems associated only with WAIHA.

Problems can be encountered in obtaining a blood group and in crossmatch procedures.¹ Approaches to overcoming these problems are discussed below.

ABO Typing Problems

There are rarely problems when ABO typing patients with WAIHA, but companies may add up to 8% albumin to their antisera, leading to spontaneous agglutination of the red blood cells (RBCs) on rare occasions. Thus, it is wise to use a negative control (e.g., 10% albumin) when typing patient’s RBCs, with a positive direct antiglobulin test (DAT).

Rh Typing Problems

When adding DAT + RBCs to any reagent containing a potentiator (e.g., albumin), one runs the risk that the IgG-coated RBCs will spontaneously agglutinate. Garratty et al³ found that 23% of DAT + RBCs showed spontaneous agglutination when added to the commercial anti-Rh (diluent) control [over 50% of those showing spontaneous agglutination had strongly (2+ - 4+) positive DATs].

Thus, the first approach is to type the DAT + RBCs using the routine method. If a commercial slide and rapid tube polyclonal (e.g., anti-D) anti-Rh is used, it is important that the negative control is the diluent used by the company that makes the anti-D; one cannot substitute the diluent of another company or 20-30% albumin.³ Many laboratories are now using monoclonal anti-D and the company may not supply a diluent control. As potentiators may be present in monoclonal reagents, a 10% albumin control is not always adequate. We recently reported problems in typing DAT+ patients with monoclonal Rh antisera other than anti-D.⁴ If the company does not provide a control for their monoclonal Rh antisera, then any results that are weaker than expected, or if all antisera react with the patient’s DAT+ RBCs, the results should be suspect and confirmed. If saline tube test (IgM) Rh antisera are available, then these can usually be used without problems of spontaneous agglutination.

If problems are encountered with the negative control, then IgG can be removed from the patient’s DAT+ RBCs and the Rh typing repeated. Enough RBC-bound IgG can be removed to prevent spontaneous agglutination, by treating the RBCs with low pH glycine5 or chloroquine.^5,6 Whatever method is used, it is important to always treat control Rh(D) positive RBCs identically to the patient’s RBCs to control the weakening of Rh antigens that always occurs, to some degree, during treatment.

Crossmatching Problems

Approximately 60% of patients with WAIHA have autoantibodies in their sera that react by the indirect antiglobulin test (IAT), if saline, LISS, or albumin-suspended RBCs are used.^1,2 If commercial LISS additives, polyethylene glycol (PEG), solid-phase, column agglutination (e.g., gel), or enzyme procedures are used, then more autoantibodies will be detected. Approximately 40% of sera from these individuals will contain alloantibodies.⁷ When alloantibodies occur together with autoantibodies, they may be masked by the autoantibody and not obvious, as most or all RBCs tested react with autoantibody.

Patients who are transfused with blood that is only incompatible with autoantibody, rarely exhibit any clinical signs of a hemolytic transfusion reaction (HTR), and the incompatible blood may or may not have a shorter survival than their own RBCs. In contrast, AIHA patients who are transfused with blood incompatible with alloantibody will show the same clinical signs of an HTR as any other patient, and the blood will probably have a much shorter survival than the patient’s own DAT+ RBCs. Because of this, it is of prime importance to exclude the presence of alloantibodies in the patient’s serum before transfusing incompatible blood. Of secondary, and perhaps questionable importance, is the possible specificity of the autoantibody.

The three most popular methods used to detect alloantibodies, that might be masked by autoantibodies, are: 1) warm autoadsorption, which uses either heat-eluted enzyme-treated autologous RBCs^1,5 or autologous RBCs treated with ZZAP reagent⁸; 2); adsorption with enzyme-treated or ZZAP-treated allogeneic RBC¹; more recently, adsorption in the presence of polyethylene glycol (PEG) has been suggested to be an efficient method^7-9,11; 3) dilution of the patient’s serum.^1,12

Using adsorptions with heat-eluted autologous RBCs, Laine and Beattie,¹³ James et al,¹⁴ and Issitt et al¹⁷ showed that 27-38% (75% if patient had received greater than 5 transfusions)¹⁴ of AIHA patients had alloantibody present in their serum that were masked by autoantibody. Using adsorptions with allogeneic RBCs ^15,17 or a combination of autologous and allogeneic RBCs¹⁶ Wallhermfechtel et al,¹⁵ Sokol et al,¹⁶ and Issitt et al¹⁷ found that 14% (32% if patient had received greater than 5 transfusions¹⁶) and 41% respectively had alloantibodies present in their serum. In 1997, our reference laboratory examined sera from 694 AIHA patients for alloantibodies. We detected alloantibodies in 40% of the sera containing autoantibodies, only after performing adsorptions in the presence of PEG or with ZZAP-treated allogeneic RBCs. Issitt et al¹⁷ suggested that some of the so-called alloantibodies detected following adsoption (especially autoadsorption) were really autoantibodies mimicking alloantibodies (69% following autoadsorption and 19% following allogeneic adsoprtion).¹⁷

In summary (see Table 1), we believe that some approach to exclude underlying alloantibodies must be used, if time allows, before transfusing patients with AIHA. If one is to provide the same protection to AIHA patients as to other patients, these tedious procedures need to be repeated every three days following transfusions. If there is no time to exclude the presence of alloantibodies by the approaches below, then matching the patient’s Rh phenotype should be used. It is preferable that the blood is also K- and Jk(a-). This will avoid problems with the most common alloantibodies causing hemolytic transfusion reactions.

Providing Blood for Patients who Have Not Been Recently Transfused

If a patient has not been recently (e.g., during the last month) transfused, the warm autoantibodies should be adsorbed from the patient’s serum, using autologous RBCs; this will allow any alloantibodies present to be more readily identified. As the patient’s RBCs have been absorbing autoantibody in vivo, and the autoantigens may be blocked, one has to remove RBC-bound autoantibody before performing autoadsorptions in vitro. Once autoantibody is removed, exposing autoantigen, the RBCs are enzyme-treated so that they will more efficiently adsorb autoantibody, from the patient’s serum, in vitro.

The original procedure utilizes patient’s RBCs, that are heated at 56°C for 5 minutes, to elute autoantibody.¹ These RBCs were then enzyme-treated (with 0.1% papain or ficin). Enough RBCs are prepared for several adsorptions. One or two volumes of washed heat/enzyme –treated patients’ RBCs are incubated with one volume of patients’ serum at 37°C for 30 minutes. If the IAT was originally 1+, then only one adsorption is unusually necessary, but if the IAT was 2+ or 3+ then 2 or 3 autoadsorptions may be necessary; a 4+ IAM may need more than 3 adsorptions. If more than 3 adsorptions are used, there is a danger of diluting the alloantibody in the serum, thus great care must be taken to thoroughly pack the RBCs used for the adsorptions. We prefer an alternative procedure that uses ZZAP reagent [2.5 vols. 0.2M dithiethreitol (DTT) + 0.5 vols 1% papain or ficin + 2 vols. pH 7.3 phosphate buffered saline (PBS) final pH 6.0 to 6.5], which removes RBC-bound IgG autoantibody and enzyme treats the RBCs in a one-step procedure.

The number of adsorptions needed are determined, in relationship to the strength of the original IAT result (see above). Following autoadsorption, the patient's serum is separated and tested against a small panel of RBCs.

If no reaction occurs after either adsorption method described above, then it can be assumed that all the antibody is autoantibody. If reactions still occur, then the alloantibody specificity is determined by testing against a larger panel of RBCs. If all RBCs on the panel react, then the serum may need more autoadsorptions, or there is an alloantibody to a high frequency antigen.

One has to remember that enzymes destroy some antigens (M, N, S, Fy^a, Fy^b, En^a, Ge, JMH, Ch/Rg, In^b), and ZZAP will destroy all these plus others affected by DTT (e.g., LW, all Kell and Lutheran antigens, Yt^a, Hy, Gy, Yk^a/McC^a/Kn^a.) Thus, rare autoantibodies (e.g., anti-En^a or anti-LW) to these antigens will not be autoadsorbed. This can also be advantageous as alloantibodies of these specificities will not be adsorbed and should be readily detectable in the serum.

Providing Blood for Patients who Have Been Recently Transfused

If a patient has been transfused recently (e.g., within the last month), then one cannot perform a true autoadsorption as the patient’s RBCs are contaminated with transfused allogeneic RBCs. In these circumstances we usually perform adsorptions using selected allogeneic RBCs. The allogeneic RBCs can be enzyme- or ZZAP-treated. We use one rr sample, one R₁R₁ sample, and one R₂R₂ sample; one sample should be Jk(a-) and one should be Jk)b-). We prefer to use ZZAP-treated RBCs as these cells will not absorb any antibodies to any antigens in the MNS, Kell, Lutheran, and Duffy systems (see above), but will adsorb autoantibodies efficiently. The adsorption procedure using allogeneic RBCs is similar to that described above when using autologous RBCs. By observing the pattern of reactivity left in the serum in relationship to the phenotype of the adsorbing RBC, one can determine the specificity of the alloantibody. Disadvantages of absorbing with allogeneic RBCs are that large volumes of phenotyped RBCs are needed, and alloantibodies to high frequency antigens will be adsorbed by all three RBC samples, leading to a false assumption that no alloantibody is present.

If it is not possible to perform adsorptions with allogeneic RBCs of known phenotype (for instance, a hospital may not have sufficient quantities of such RBCs), then one can perform a dilution technique. This technique relies on the fact that alloantibody is usually present at a higher titer than autoantibody, which is being continually adsorbed, in vivo, by the patient’s RBCs. Serial dilutions of the patient’s serum are tested (e.g., by indirect antiglobulin test) against a pool of 2 screening RBCs. The dilution that gives a 1+ reaction (possibly representing the stronger alloantibody) is tested against a panel of RBCs.1 Øyen et al⁸ use only a 1:5 dilution of the patient’s serum to screen for the presence of alloantibodies. We do not find this approach to be as efficient as the adsorption procedures but it can be used by hospitals when the adsorption cannot be utilized.⁷

Recently, some investigators have described a promising new, simple approach. Liew et al,⁹ Miller et al,¹⁰ and Yiagan et al¹¹ found that warm autoantibodies were readily adsorbed by mixing the patient’s serum, allogeneic RBCs, and PEG for 30-40 minutes at 37°C. We found this new approach to be as efficient, quicker and less labor intensive, than adsorption with ZZAP-treated RBCs.⁷ Table 1 summarizes approaches for untransfused and transfused patients.

1. Phenotype RBCs (Rh, K, Kidd, Duffy)
2. Save as many RBCs as possible for future autoadsorptions (ACD, CPD, Alsevers, or frozen)
3. If serum screen is positive, test against panel
4. Warm autoadsorption if necessary.

Patient Recently Transfused

1. If patient’s pretransfusion RBCs are available, perform a "warm autoadsorption"
2. If pretransfusion RBCs are not available:
   1. adsorb patient?s serum with allogeneic RBCs (rr, R₁R₁, R₂R₂); or
   2. test selected dilution of patient’s serum against panel of RBCs
      Should the Specificity of the Autoantibody be Taken into Acount?

Defining the specificity of the autoantibody is not as important as excluding the presence of alloantibodies, but if time allows, and blood lacking the putative antigens can be found, it may help the survival of the transfused RBCs.^1,2,18 If the autoantibody shows a well-defined specificity (e.g., anti-e), and the compatible blood is readily available, then we try to provide blood lacking that antigen. These RBCs may survive much longer than antigen positive RBCs. The exception to this is when the compatible blood is of an exceptionally rare type (Rh_null, En(a-), Wr (b-), Di(b-)]; we would argue that it is better to reserve these rare phenotypes for patients with alloantibodies.

When autoantibodies show a relative specificity, that is to say, the serum reacts to a higher titer with RBCs having a certain antigen (e.g., the antigen e) than RBCs lacking that antigen, we believe it is still worthwhile to transfuse RBCs lacking the putative antigen, following the same principles as mentioned above. There is not much data to support the value of doing this, but some ⁵¹Cr-survival studies have been published to support this viewpoint.¹

Blood should never be withheld from a patient while any of the above tests (especially specificty of autoantibody) are being carried out without negotiating with the physician in charge of the patient, explaining the relative benefits of these time-consuming procedures.

1. Only transfuse if patient is symptomatic*
2. Transfuse small volumes of blood ? only enough to relieve symptoms**
3. Exclude presence of 37°C-reactive alloantibodies before transfusing incompatible blood
4. If patient has hemolytic anemia, and time allows, define specificity of warm autoantibody. BUT never put patient at risk by waiting for this result.
   
   * see reference 1 (pp 358-65)
   **see reference 1 (pp 384-86)
   

References

1. Petz LD, Garratty G. Acquired immune hemolytic anemias. New York: Churchill Livingston, 1980.
2. Garratty G. Autoimmune hemolytic anemia. In:Garratty G (ed). Immunobiology of transfusion medicine. New York: Dekker, 1994, pp 493-521.
3. Garratty G, Postoway N, Nance SJ, Brunt DJ. Spontaneous agglutination of red cells with a positive direct antiglobulin test in various media. Transfusion 1948; 24:214-7.
4. Rodberg K, Tsuneta R, Garratty G. Discrepant Rh phenotyping results when testing IgG-sensitized RBCs with monoclonal Rh reagents. Transfusion 1995;35:67S (abstract).
5. Vengelen-Tyler V, ed. Technical manual. 12th ed.Bethesda, MD:American Association of Blood Banks, 1996.
6. Edward JM, Moulds JJ, Judd WJ. Chloroquine dissociation of antigen-antibody complexes. Transfusion 1982;22:59-61.
7. Leger RM, Garratty G. Evaluation of methods for detecting alloantibodies underlying warm autoantibodies. Transfusion 1997;37:65S (abstract). And Transfusion 1999;39:11-16.
8. Branch DR, Petz LD. A new reagent (ZZAP) has multiple applications in immunohematology. Am J Clin Pathol 1982;78:161-7.
9. Liew YW, Duncan N. Polyethylene glycol in autoadsorption of serum for detection of alloantibodies. Transfusion 1995;35:713.
10. Miller R, Unger P, Shapiro A. Rapid adsorption of warm autoantibodies using polyethylene glycol. Transfusion 1996;36:20S (abstract).
11. Yiagen ES, O?Gorman F, Roller V, Moon M. Differential adsorptions using polyethylene glycol (PEG). Transfusion 1996;36:20S (abstract).
12. Øyen R, Angeles ML. A simple screening method to evaluate the presence of alloantibodies with concomitant warm antibodies. Immunohematology 1995;11:85-7.
13. Laine ML, Beattie KM. Frequency of alloantibodies accompanying autoantibodies. Transfusion 1982;25:545-6.
14. James P, Rowe GP, Tozzo GG. Elucidation of alloantibodies in autoimmune haemolytic anaemia. Vox Sang 1988;54:167-71.
15. Wallhermfechtel MA, Pohl BA, Chaplin H. Alloimmunization in patients with warm autoantibodies. Transfusion 1984;24:482-5.
16. Sokol RJ, Hewitt S, Bokker DJ, Morris BM. Patients with red cell autoantibodies: selection of blood for transfusion. Clin Lab Haematol 1988;10:257-64.
17. Issitt PD, Combs MR, Bumgarner DJ, Allen J, Kirkland A, Melroy-Carawan H. Studies of antibodies in the sera of patients who have made red cell autoantibodies. Transfusion 1996;36:481-486.
18. Garratty G. Target antigens for red cell-bound autoantibodies. In: Nance SJ, ed. Clinical and serological aspects of immunohematology. Arlington, VA:American Association of Blood Banks, 1991:33.