A colleague outside the USA asks about the characterization of the DVI Rh variant. He would like to know what policy exists in the US or at least in California. He wonders if there is any guidance from the AABB or from other national organizations? According to what he has read (eg: Transfusion 2004;44:1282 (France); Transfus Med 1995; 5:171 (UK)), patients (blood recipients) and pregnant women should be typed with monoclonal IgM anti-D that does NOT detect the DVI Rh variant whereas donors and newborns should be typed with polyclonal anti-D reagent that DOES detect the DVI Rh variant. He wonders what would be the policy with the ambulatory individual who comes to be tested just because he or she wants to know their Rh blood type. The inquiring colleague says that he would either type them for Rh with both kinds of Rh typing reagents, or if he could only pick one reagent per person, he would type a woman with a reagent that does NOT detect the DVI Rh variant and he would type a a man with a reagent that DOES detect the DVI Rh variant.

A distinguished immunohematologist in Southern California reports that in the United States there are no regulations that relate specifically to the DVI Rh variant. The DVI Rh variant lacks many of the D epitopes (i.e., partial D) and is the most common partial D to present as a weak D (Du). AABB Standards mention that Rh-negative donors and babies born to mothers being considered for Rh prophylaxis should be tested for weak D, and that testing for weak D is not necessary for transfusion recipients. The area of weak D and Rh(D) variants is complex and becoming more complex as we learn more at the molecular level. He adds that weak D (about 0.2 ­ 1% of Caucasian population) may react weakly with anti-D, or react only by specific tests (e.g., antiglobulin test). This may be because of a suppressive effect of the C gene when in trans to the D gene (e.g., Dce/Ce), or because part of the D antigen is missing (partial D) and/or due to an aberrant form of D (e.g., at the molecular level). Aberrant forms of D can be associated with low D antigen sites. Modern monoclonal anti-D reagents react strongly (e.g., without antiglobulin test) with most RBCs that would have previously been classified as weak D. It became generally accepted that most weak D possess a normal D antigen and that there was only a quantitative difference. Because of this, the possibility of D immunization was disregarded (e.g., in typing pregnant women). It is now becoming clear this is not true. The weak D phenotype can be caused by many different RHD alleles encoding aberrant Rh(D) proteins. D antigen density on weak D RBCs ranges from 70 ­ 4000 D sites/RBC (although most have >500 sites); qualitatively distinct D antigens can often be defined (e.g., at molecular level). Wagner et al (Blood 2000;95:2699) showed that most weak D, including the more commonly encountered types, carry altered D antigens; some of these individuals can be immunized to make anti-D and others (luckily this includes the most common types) cannot.

He adds that at the practical level, there is one major question - “Can we harm a patient if we do not take into account what we have learned about weak D and Rh variants”? If the answer is yes, then a further question might be “How rare are these phenotypes and what are the cost-benefits of further defining D phenotypes by performing extra testing routinely”?

To answer the first question we should consider patients who are to be transfused, pregnant women and blood donors separately. If we type a patient (transfusion recipient), who has a weak D or a variant phenotype as D+, they will be given D+ RBCs and there is a small possibility that they could make anti-D. If the anti-D typing serum classifies the patient as D-neg, then no harm will come to the patient, but we will have wasted D-neg blood. Similarly, a D+ pregnant woman could make anti-D that could affect a D+ fetus. Similar to the patient above, a woman with a weak D can, on rare occasions, make anti-D. If classified as D+, the woman would not be given RhIg following delivery of a D+ baby. If classified as D-neg, but the woman really has a “weak D phenotype”, then the RhIg is often an unnecessary cost. Weak D in blood donors should be detected and classified as D+ so that those units are not used for D-neg recipients.

Only about 5-10% of the weak Ds are partial Ds, most of which are DVI Rh variant. The DVI Rh variant is associated with hybrid alleles and can be associated with low D antigen density where the D antigen may react with many routine anti-D by antiglobulin test but not by direct agglutination. Many partial D phenotypes present as normal D types (e.g., react strongly with routine anti-D) so they are often not classified until they make anti-D. Most weak D phenotypes will not make anti-D when receiving D+ blood because they have all the D epitopes, but some can make anti-D. Most D+ individuals making anti-D are of the DVI phenotype (the frequency of DVI is 0.02 - 0.05% of Caucasians), and typically present as weak D (DVI lack many of the D epitopes); other partial D who can make anti-D (e.g., Vb and FPTT) may appear to have normal strength D antigens.

In the experience of the California immunohematologist, the Europeans have had a greater interest in weak D and Rh variants than in the US. This may be because they have had access to many more monoclonal antibodies (e.g., panels of anti-D to determine which D epitopes are present). They also have applied molecular techniques more frequently than investigators in the US. For instance, Dr. Flegel, in Germany has tested all his D-neg donors at the molecular level, since 2000, and several ventures are underway to use high input molecular testing of all donors for multiple RBC, platelet and HLA antigens. Guidelines in several European countries and Australia suggest routinely using 2 anti-D reagents, one capable of detecting DVI and the other not reacting with DVI (e.g., use of an anti-D test that can detect DVI for blood donors and newborns, and an anti-D that does not detect DVI for pretransfusion tests of patients and pregnant women). Apart from changing the term Du to weak D, and the mention of weak D in several parts of the AABB standards, as mentioned above, there has not been as much interest in the US. Judd et al [Transfusion 2002;42:20S (abstr)] showed that current FDA-licensed anti-D will react with most D variants (DII, DIII, DIIIa, DIVa, DIVb, DVa), but not DVI, by direct agglutination or the gel test. DVI will be detected by indirect antiglobulin test (IAT). Thus, if the IAT is not used, DVI+ pregnant women and recipients will be typed as Rh(D)-neg and given Rh prophylaxis, and not transfused with D+ blood respectively. Blood donors will be labeled as Rh+ as an IAT, or equivalent testing is performed on all Rh-neg donors; thus DVI blood would not be transfused to D-neg recipients. RoHar is still a problem as it reacts strongly with anti-D in the gel test distributed by Ortho, but Ortho’s anti-D does not react with RoHar by tube test (direct agglutination or IAT). Other companies’ anti-D (Gamma and Immucor) give variable reactions with RoHar by direct agglutination but react by IAT.

He continues warning that another emerging problem is the weakest of the weak D, named DEL, because serologically the D antigen can only be demonstrated by adsorption and elution of anti-D. Until recently, this was thought to be mainly a problem in Japan but recent molecular studies shows that this phenotype is more common in Europeans than first thought. This means that some Rh-neg donors are truly D+ and evidence is emerging that such units can alloimmunize D-neg recipients to make anti-D. This rare phenotype (DEL) can be easily detected by testing at the molecular level.

He concludes saying that although D testing in the US appears acceptable (at present), it is mainly coincidental, due to the particular reactivities of current anti-Ds. There are no official requirements (e.g., AABB or FDA) for anti-D activity in regard to Rh variants such as DVI, although FDA requires that manufacturers of monoclonal anti-D specify their ability to detect DVI and put that information into the package insert. Cost-benefit calculations will obviously be needed before molecular testing becomes routine in the USA, and thus I believe that we will be relying on serological approaches for some years to come. As far as the question of typing an individual who only wants to know their blood group, he would agree with the questioner’s approach to determining the Rh type.

ADDENDA Mar. 16, 2005

The following comments have been received.

1. A distinguished immunohematologist in Michigan reports that his institution currently handles the partial D issue by testing with two different anti-D reagents, as is required in some European countries. He acknowledges that his institution has followed this practice for well over 30 years, not because they wanted to recognize partial D individuals, but rather to prevent mistyping an Rh-negative woman as Rh positive due to sample misidentification. In the past they tested apparent Rh-negatives (by direct testing) for weak D; if positive, they considered them to be Rh-positive for the purpose of blood transfusion and, if pregnant, they have withheld Rh immune globulin prophylaxis. With automation and positive sample identification via bar-coded labels, there is less concern with regard to sample mis-identification. However, with the advent of reagents formulated with IgM monoclonal anti-D, and in light of the molecular data generated by Flegel WA and Wagner FF. Clin Lab 2002;4:53-9, they now test patients initially on gel cards; since they perform an electronic crossmatch, a second ABO/Rh (if only one is on file) is done by tube testing.

Tests with panels of monoclonal anti-D's and molecular analysis of discrepant or weak (<2+ by gel) reactions has revealed five cases of partial D, and 11 weak D Types (Judd et al. Unpublished data, 2005). They have treated these patients as though they were Rh negative.

He asks the rhetorical question "Why is prevention of D alloimmunization in partial D individuals so sacrosanct?" He then answers the question saying "One has to question if our concerns with partial D, and our efforts to manage DVI patients as though they were Rh negative, are justified when the DVI phenotype is found in such a small percentage (0.02 in some studies) of patients, yet we do not select c-negative blood (or K-blood) for R1R1 (or K-) pre-menopausal women. To do so would, over time, prevent a significant number of delayed HTR's and cases of severe HDN as well as decrease the number of positive tests for unexpected antibodies that require time-consuming and costly identification studies. We have to accept that our testing strategies are not perfect."

He concludes saying that there are some partial D's (DII through DV) that give 3+ and 4+ reactions in direct tests with most currently available monoclonal anti-D reagents. (Judd et al. Transfusion 2002;42(S):20). Molecular methods are required to detect these partial D individuals if we want to prevent them from making anti-D. The Michigan colleague is not convinced they need to go to that extreme, especially if they don't make similar efforts to prevent alloimmunization as he has discussed above. In his mind, the test for weak D on pregnant women and potential transfusion recipients is redundant. One should err on the side of safety and manage patients as Rh-negative when their red cells give discrepant or weak reactions with anti-D by direct testing.

Please submit comments to the e-Network Forum.

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