Experience with RHD*weak D type 4.0 in the USA

Several years ago, in an effort to begin to standardise practice for patients whose red blood cell type is weaker than expected or whose cells show variable reactivity with different anti-D reagents, a Working Group, representing multiple professional organisations in the USA, was convened to draft recommendations on phasing in the use of RHD genotyping. The effort was targeted to the treatment of females of child-bearing potential and obstetric patients to guide administration of Rh immune globulin to avoid unnecessary use, and also to select blood for patients to avoid overuse of D-negative units. Observational data (primarily from Europe) formed the basis for the recommendations¹. The Working Group felt confident that individuals with weak D phenotypes encoded by alleles designated as weak D type 1, 2, or 3, were not at risk of producing anti-D when exposed to D-positive red blood cells during pregnancy or following transfusion, and could be treated as D-positive individuals. Although we have encountered rare cases of anti-D in individuals with weak D types 1 and 2, the antibodies are often cross-reactive with the patient’s own red blood cells, i.e. show “auto” reactivity, or sometimes demonstrate characteristics of both allo- and auto-reactivity (personal observations). No clinically significant transfusion or pregnancy complications have been reported.

In this issue of Blood Transfusion, investigators with experience in Germany and France propose that individuals with D-positive phenotypes encoded by weak D type 4.0 and 4.1 alleles should also be treated as D-positive² based on the paucity of reports of anti-D in their collective experience, and based on the lack of negative outcomes in their countries and in Tunisia where this allele is indicated to be “of highest prevalence world-wide, i.e. 1 in 105 haplotypes”. In Tunisia, anti-D has been reported in only one patient with weak D type 4.0 and it showed characteristics of “auto” reactivity; only one allo-anti-D has been reported in the French haemovigilance system².

Here we share our experience in the USA, where RHD*weak D 4.0 is not uncommon in the African-American population, with an allele frequency of 0.0298 in patients with sickle cell disease (personal observations), and 0.019 in African-American blood donors³. In a large hospital sickle cell disease centre served by the Authors, among 857 patients there are 45 (1 in 19) with this allele, and 15 are D-positive due to homozygous or hemizygous inheritance of RHD*weak D 4.0⁴. To date, three have made anti-D identified by the hospital laboratory and are now receiving D-negative blood for transfusion.

In the past 6 years, 15 samples from patients with anti-D submitted for RHD genotyping to the New York Blood Center were found to be hemizygous or homozygous for RHD weak D type 4.0. Among these, 12 were females of child-bearing potential (ages from 6 to 37 years). Six were pregnant; five had no history of transfusion but previous pregnancies were documented (n=2, 3 and up to 8) in all except one. Seven were patients who had been multiply transfused for sickle cell disease and six of these had several additional alloantibodies. There was no indication from direct testing that the anti-D were autoantibodies, but sufficient samples were not provided for adsorption studies. During this same time frame, 23 samples from patients with a history of anti-D submitted for RHD genotyping to the American Red Cross were associated with RHD weak D type 4.0. Of these, 21 were female, 15 were of child-bearing age and seven were pregnant. Past pregnancy or transfusion histories were not available and the details of the antibody workups were not provided, or did not include efforts by the hospitals to differentiate between allo- and auto-antibodies.

Autoreactivity versus alloreactivity: challenges and relevance

The fact that anti-D is very easy to identify in almost any hospital setting complicates adequate data collection of auto- versus allo-specificity in these patients. Since it is well known that individuals with D-positive red blood cells may make anti-D due to partial D phenotypes, most hospital blood bank staff do not see a need to refer samples to our reference laboratories for adsorption studies to differentiate allo- from auto-anti-D, and would not support the extra cost to do so for a patient for whom compatible D-negative blood is easily found. Sufficient serum or plasma and pre-transfusion autologous red blood cells are needed to do these studies, but the volumes of samples submitted for RHD genotyping are not adequate to do the multiple adsorption tests required for the appropriate investigations, and patients have often been recently transfused. Hence, the anti-D in samples referred to us and found to be associated with weak D 4.0 phenotypes have not been widely studied to prove or disprove whether the antibody is or is not an allo-antibody. Furthermore, the clinical relevance of the antibody will not have been determined, only the relative specificity. To study the clinical relevance, in vitro (monocyte monolayer assay) or in vivo (⁵¹Cr or biotin survival studies) would need to be performed to predict the clinical significance in the absence of documented haemolytic transfusion reactions or evidence of decreased survival of transfused red blood cells. Although we often use cross-reactivity, i.e. “auto” reactive characteristics of an antibody as a surrogate for lack of clinical significance, this may or may not be appropriate in every situation. Experience with auto-antibodies with other specificities has shown that not all auto-antibodies are insignificant. At times, transfusion of antigen-negative blood is required, and, as in other systems, Rh “auto-antibodies” have been associated with decreased red blood cell survival.

Clinical practice considerations

There are two situations to be considered in clinical practice. The first is when a patient found to be D-positive encoded by RHD*weak D 4.0 or 4.1 presents for transfusion or pregnancy workup with no history of an antibody. The second is when a patient who is D-positive due to RHD*weak D 4.0 or 4.1 demonstrates anti-D in the plasma and requires transfusion. In the hospital transfusion service, if it were assumed that the anti-D made by a person with RHD*weak D 4.0 or 4.1 was to be disregarded and the patient treated as D-positive, the units would be incompatible in the full crossmatch (required in patients with antibodies). Issuing an incompatible unit, rather than giving a D-negative compatible unit, or issuing a D-positive unit for a patient with a history of anti-D no longer reactive, is not likely to be embraced by clinicians or laboratorians. A number of anti-D associated with weak D 4.0 alleles were in patients with sickle cell disease. Clinicians would be reluctant to consider transfusing against anti-D, regardless of the laboratory characteristics of the antibody, given the history of transfusion reactions and hyper-haemolysis in this population of patients.

The evidence considered by the Working Group on treating individuals with weak D type 1, 2, 3 as D-positive included: (i) the absence of reports of anti-D associated with transfusion and pregnancy; (ii) the apparent “auto” or cross-reactivity of anti-D in the rare or few anti-D observed; and (iii) the fact that no clinically significant outcomes had been reported. In contrast, we have observed 38 examples of anti-D in individuals with weak D 4.0 in the last 6 years, and 13 were pregnant women. No negative outcomes have been communicated to us and, although it is not certain that they would be reported, these pregnancies required monitoring as high-risk pregnancies as standard of care.

Selection of blood for transfusion for patients with weak D 4.0 phenotypes who do not have anti-D is complicated by the fact that RHD*weak D 4.0 is common in samples submitted for RHD genotyping to resolve serological weak D or D typing discrepancies in the USA, especially among individuals of African descent. Over this 6-year time period, 114 samples submitted to the American Red Cross associated with weaker than expected D typing or typing discrepancies were homozygous or hemizygous for weak D 4.0 alleles. In the last year alone, 62 of 352 samples (17.6%) referred to the New York Blood Center due to serological weak D phenotypes were hemizygous or homozygous for weak D 4.0 alleles. Supporting every patient who is D-positive due to weak D 4.0 with D-negative units to prevent anti-D would be a burden for the Rh negative blood supply. Consideration could be given specifically to vulnerable females, who could potentially be negatively affected by a pregnancy classified as high risk. For patients with sickle cell disease requiring long-term transfusion support, it is anticipated that, in the near future, RHD genotyping of minority donors would enable patients with RHD*weak D 4.0 to receive D-positive units from donors with the corresponding weak 4.0 allele to avoid anti-D alloimmunisation.

Summary

The number of examples of anti-D associated with weak D 4.0 alleles referred to our laboratories has caused reluctance to classify individuals with weak D 4.0 as not at risk of anti-D, as is currently done for weak D types 1, 2, and 3 in the USA. The apparent greater incidence of anti-D observed in the USA among individuals of African ethnicity compared to the lack of reports of anti-D in France in North Africans with weak 4.0 alleles as well as in Tunisia may be due, in part, to differences in reporting criteria, or to differences in genetic background, with the majority of African-Americans being of West African descent.

We strongly encourage facilities to perform RHD genotyping of patients who are D-positive with identified anti-D. If due to RHD*weak D 4.0 or 4.1, additional serology should be considered to characterise the antibody. Reactivity of the antibody with allogeneic weak D 4.0 red blood cells is helpful in this regard. As the use of RHD genotyping increases, data collection from hospitals regarding D-positive individuals who have made anti-D and the RHD genotypes associated would be informative. This issue could benefit from a hospital-based survey of patients who are D-positive with anti-D which could collect information on RHD genotyping data, practice and clinical outcomes of transfusion and pregnancy.