DNB: a partial D with anti-D frequent in Central Europe

Anti-D immunizations may occur in “partial D” carriers. This designation relates to more than 30 alleles differing in molecular structure, phenotype, population frequency, and lenience to anti-D production.1 The molecular bases were gene conversions,2 single missense mutations in the exofacial protein segments,3 and multiple missense mutations dispersed throughout the RhD protein.4 A partial D needs to be considered for transfusion strategies if its carriers are frequent, easily anti-D immunized, and likely to receive D-positive units in the case of transfusion. D category VI 5 is the classical example. In several countries, typing methods6 were introduced that assured D-negative transfusion and anti-D prophylaxis for DVI mothers to prevent anti-D immunizations. Despite the large number of known partial D alleles, a relevant portion of anti-D immunizations is still occurring in individuals who did not carry any known partial D.7-9Thus, the characterization of the unknown alleles underlying such case reports is remaining important.

Two D-positive samples (RIR-2 from Bern, Switzerland; RIR-3 from Bavaria, Germany) with anti-D were referred to our laboratory to elucidate the cause of anti-D immunization. Later, 1 additional blood sample (RIR-41 from Denmark) and 2 DNA samples (RIR-7, RIR-8 from Austria) were referred to the Rhesus Immunization Registry (RIR) because of anti-D in these D-positive probands. The Danish proband RIR-41 had an Austrian grandfather.

DNA was handled as described previously.10 The 10RHD exons were sequenced by an RHDallele–specific method11 in the 2 index samples (RIR-2, RIR-3). In the remaining samples, exon 7 was sequenced only.

A polymerase chain reaction with sequence-specific priming (PCR-SSP) to confirm or detect the 1063G>A substitution in DNB was devised as modular extension of a PCR-SSP system previously developed for RHD typing.10,12 Specific primers were re77 (TCTCCACAGCTCCATCATGGG) and DNBb (CAGTGACCCACATGCCATTACT) at a concentration of 3.5 μmol; HGH control primers were used at 0.75 μmol.

Routine D typing was done in tube tests using commercial monoclonal anti-D (Seraclone anti-D 226, clone BS226; and Seraclone anti-D [232], clone BS232; Biotest, Dreieich, Germany; and ImmuClone anti-Dfast, clone D1-4E11; and ImmuClone anti-Drapid, clone RUM-1; Immucor, Norcross, GA). A commercial panel of monoclonal anti-D (D-Screen; Diagast, Loos, France) was tested in gel matrix technique (LISS-Coombs 37°C, DiaMed-ID Micro Typing System; DiaMed, Cressier sur Morat, Switzerland). D epitopes were determined by agglutination, and antigen density and Rhesus index were determined by flow cytometry as previously described.13 DIV type III 13and DNU14 controls derived from donors in Baden-Württemberg, Germany.

Ethylenediaminetetraacetic acid (EDTA)– or citrate-anticoagulated blood samples were collected from blood donors in Baden-Württemberg (Germany, 1118 donors of CcDee phenotype, 1010 donors of ccDEe phenotype), Ticino (Italian-speaking Switzerland, 500 D-positive donors), Bern region (German-speaking Switzerland, 693 donors of CcDee phenotype), and Denmark (768 donors of CcDee phenotype). The donors were tested with the immunoglobulin G (IgG) monoclonal anti-D LOR17-6C7 (Germany and Denmark) or HIRO-7 (Switzerland).15 Suspected DNB samples were confirmed by sequencing of exon 7 or by PCR-SSP. Haplotype frequencies used for calculation of DNB allele frequencies were CDe 0.431, cde 0.394, cDe 0.021, and Cde 0.011 in Germans16; 0.4236, 0.3792, 0.0145, and 0.0059 in Danes17; and 0.4442, 0.3617, 0.0232, 0.0113, and cdE 0.049 in Swiss.17 Phenotype frequencies for CcDee were compared using the Fisher exact test for a 4 × 2 contingency table; for this calculation the 500 D-positive samples in Ticino were inferred to comprise 187 CcDee samples.

The name DNB derived from DNU-like andBayern (Bavaria, Germany) or Bern, because DNB was similar to DNU,14 whose point mutations were closely adjacent.

The molecular structure of 2 D-positive samples with anti-D (RIR-2, RIR-3) was determined by RHD-specific sequencing from genomic DNA. In RHD exon 7, a single G to A exchange at position 1063 was detected that resulted in a Gly to Ser substitution at codon 355. The affected amino acid was located in the exofacial loop 6, adjacent to the mutations observed in DII (Ala to Asp at position 354)14 and DNU (Gly to Arg at position 353).14 The nucleic acid and amino acid sequence data were deposited in EMBL/GenBank/DDBJ under accession number AJ417868.

In routine D typing, both DNB samples typed as CcDee without noticeable weakening of the antigen D. All 9 monoclonal anti-Ds of a commercial partial D classification kit (D-Screen) were reactive. An antigen density of 5908 antigens per cell and a Rhesus index13,18 (RI) of 0.02 were determined in RIR-2. These results were comparable to those obtained with a DNU sample (8073 antigens per cell, RI 0.06) and a CcDee DIV type 3 sample (4544 antigens per cell, RI 0). The D epitope pattern was determined for both index samples and was unique (Table1): 6 of 83 anti-Ds tested did not agglutinate DNB, indicating a loss of epitopes epD6 and epD31 as well as part of epD18 and epD23. The crossmatch with DNU red cells was faintly positive; there are no DII red cells available anymore.

We screened random samples of CcDee blood donors in 4 European populations (Table 2). With the exception of Danes, DNB was found frequently. The highest frequency was observed in the Ticino population. DNB was the most frequent partial D in whites described so far; its frequency even exceeded that of DVII.19 Further testing of other populations may be of potential benefit.

Including the probands with anti-D, we presented 12 unrelated carriers of DNB who were whites of Central Europe; the only Danish proband had an Austrian grandfather, from whom he may have inherited theDNB allele. All DNB probands were of CcDee phenotype. TheCDe haplotype association was further corroborated by testing the parents of the Danish proband RIR-41, who inherited his DNB allele from his CCDee father.

Including the 2 index probands, 5 cases of allo–anti-D immunization submitted to the Rhesus Immunization Registry occurred in DNB probands (RIR-2, RIR-3, RIR-7, RIR-8, and RIR-41).20 These cases represented a sizable fraction of the 32 confirmed allo–anti-D immunizations reported to this registry until February 2002. Although any Rhesus-positive blood sample with anti-D—with or without known Rhesus variant—may be sent to the registry for analysis, no other RHD allele was involved in a similar number of cases. Two additional carriers of DNB with anti-D were from Slovakia (Dr Martin Pisacka, personal communication, 1999) and from Germany (submitted as RIR-49 after we had finished the data analysis).

Anti-D titers ranged from 4 to 128, indicating that DNB carriers were generally able to produce strong anti-D. However, no clinical data like signs of transfusion reactions or hemolytic disease of the newborn were reported. In concordance with the assumption that these anti-Ds were triggered by transfusion or pregnancies, 4 of the probands were female; the male and at least 2 of the female probands were transfused prior to the detection of the anti-D. The anti-D was still detectable up to 8 years following the latest possible anti-D immunization event.

The repeated observations of anti-D immunizations probably reflected the combination of a high population frequency with a phenotype that triggered D-positive transfusions and with a moderate anti-D immunization potential. Hence, we recommended D-negative transfusions if a recipient is known to carry a DNB phenotype.

A serologic recognition of DNB would be demanding because it has a normal antigen density and is agglutinated by most anti-Ds, including almost all commercial anti-D typing reagents. All tested high-affinity monoclonal IgM anti-Ds that did not bind DVI but that agglutinated most weak D phenotypes13,21 agglutinated DNB also (Table 1). Therefore, a serologic strategy for detecting DNB would have to rely on a separate anti-D especially introduced to discriminate DNB from normal D. By careful selection of this third antibody (eg, HIRO-7 or HIRO-8), the strategy could be tuned to discriminate also DII, DIV, and R$0Har$ from normal D.21 Such a major change in D typing would be costly and might induce uncertainties that outweigh the benefit by far. The anti-D immunization risk in DNB carriers may safely be estimated to be lower than 1% per D-positive transfusion. This immunization index was less than known for anti-K and anti-c 22 but may be comparable to that of anti-Fy(a) and anti-Jk(a).

Genotyping strategies are increasingly utilized for blood grouping23 and may be devised to meet predefined specificity criteria. For example, RHD PCR in whites may be tuned to a specificity of greater 0.9999 and allows the identification of D-positive units missed by routine serology.10Likewise, DNB or DIIIa alleles may be detected specifically without any limitations imposed by the lack of suitable monoclonal anti-D. Because the frequency of partial D with relevant anti-D immunization risk is low in whites, their D-negative transfusion would not compromise the D-negative blood supply. Hence, a specific recognition and D-negative transfusion strategy for DNB may be perceived advantageous and become feasible with genotyping strategies in the future.

We are indebted to the contributors of the Workshop on Monoclonal Antibodies against Human Red Blood Cells and Related Antigens in Nantes in 1996,15 who provided most of the monoclonal anti-Ds used in this study. The authors thank Letizia Caramazza, Servizio Trasfusionale della Svizzera italiana, fondazione CRS, Lugano, Switzerland, for performing the population survey in Ticino; Drs Christoph Gassner and Diether Schönitzer, Innsbruck, Austria, for supplying RIR-7 and RIR-8 DNA samples; Dr Ernst-Markus Quenzel, Hagen, Germany, for supplying the RIR-49 sample; and Drs Martin Pisacka (supported by grant IGA MZ CR NM/5952-3), Reference Laboratory for Immunohaematology, Praha, Czech Republic, and Daniela Cupanikova, Clinic of Haematology and Blood Transfusion, Bratislava, Slovak Republic, for communicating one independent observation of DNB. We acknowledge the expert technical assistance of Marianne Lotsch, Anita Hacker, Katharina Schmid, and Sabine Zahn in Ulm as well as of Lene Christiansen in Aarhus.