A novel <i>RHD</i> allele combining partial D (c. <scp>455A</scp> &gt;C) and weak D (c. <scp>1154G</scp> &gt;C) variants

For patients with weak or discrepant RhD phenotypes, RHD genotyping is recommended to determine needs for Rh immune globulin (RhIG) prophylaxis and RhD-negative red blood cell (RBC) transfusion. Patients with weak D types 1, 2, 3, 4.0, and 4.1 can be managed as RhD-positive, with caveats for type 4.0 in pregnancy; other genotypes are managed as RhD-negative for uncertain or known risk of anti-D alloimmunization.1 We previously described our hospital-based and reference laboratory experiences with RHD genotyping.2, 3 The proband here was a 33-year-old White woman of southern European ancestry with no history of transfusion presenting at 9 weeks gestation in her first pregnancy. We performed ABO typing, RhD typing (initial direct agglutination (DA) and follow-up weak D antiglobulin testing), and solid-phase antibody screens by the automated microplate method (Neo, Werfen, Barcelona, Spain). Manual tube DA typings were obtained five 5 other anti-D reagents and for RhCcEe antigens. Initial RHD genotype was obtained by DNA microarray (RHD BeadChip, Werfen, Barcelona, Spain). To investigate unexpected results, Sanger sequencing of exons 1–10 and flanking regions and site-directed PCR testing for the hybrid Rhesus box for RHD deletion were performed. We subsequently obtained DA neonatal RhD typing. The proband's RBCs expressed a weak/negative RhD phenotype in automated testing and 1+ to 3+ agglutination in tube DA tests (Table 1). The RhCcEe phenotype was ccEe. The DNA microarray detected hemi- or homozygous (hxm) c.1154G>C and reported weak D type 2 (RHD*01W.02). Inspection of the microarray results also revealed hxm c.455A>C. Sanger sequencing identified those two variants plus c.1154-31T>C in intron 8. The hybrid Rhesus box was present. c.455A>C (p.Asn152Thr) is a gene conversion variant from RHCE into RHD exon 3. In isolation, it comprises the allele RHD*DNT (RHD*38), expressing an apparently normal RhD phenotype but associated with anti-D.4 More often, c.455A>C occurs as one of several variants in alleles of the DIVa phylogenetic cluster. In RHD*01W.02 of the Eurasian cluster, the causative variant c.1154G>C (p.Gly385Ala) is often linked to the intron variant c.1154-31T>C and in a cis arrangement with RHCE*cE. In addition to its amino acid change, c.1154G>C is located at the 5′ end of exon 9 and predicted to disrupt mRNA splicing, accounting for the weak D phenotype.5 Our proband's phenotype was similar to three Chicago cases of RHD*01W.02 (Table 1), although this allele's phenotype is variable.6 Her RHD*455C, 1154C allele possibly arose from c.455A>C microconversion in a RHD*01W.02 allele. In the Genome Aggregation Database, none of 111,705 subjects had both c.455C and c.1154C (v2.1.1, variant co-occurrence). This novel allele was submitted to GenBank (accession PQ858709). c.455A>C was previously observed in cis with weak D type 1 variant c.809T>G (p.Val270Gly) in the RHD*62 allele (phenotype unknown) (GenBank LN554881). This case illustrates the importance of reviewing genotyping tests for unexpected findings. Because of published anti-D risk with c.455A>C, we overrode the RHD*01W.02 microarray call and recommended Rh immunoprophylaxis and RhD-negative RBC transfusions if needed. The patient received RhIG at 21 weeks for bleeding (placenta previa), 28 weeks, and postpartum after delivering an RhD-positive newborn (not genotyped) at 37 weeks. RBC antibody screens were negative at 9, 28, and 37 weeks gestation. Nondetection of antenatal RhIG in antibody screens at delivery in women with partial RHD variants has raised concern for maternal RhIG adsorption and immunoprophylaxis inadequacy.7 Glenn Ramsey is a consultant at Werfen.