To the Editor:
In a recent issue of BLOOD, Scaradavou et al1reported that 72% of 261 D-positive persons with immune thrombocytopenic purpura (ITP) had positive clinical responses manifested by a mean platelet count increase of 76,000/μL after treatment with intravenous anti-D (WinRho SDF; Calgene Corp, Winnipeg/Nabi, Boca Raton, FL). Clinical responders were not distinguished from nonresponders by their gender, duration of ITP, or pretreatment platelet count. Also, approximately 62% of persons treated had a decreased hemoglobin concentration by day 7. This mild treatment-related hemolytic anemia is not unexpected, because intravenous anti-D is proposed to increase platelet counts in persons with ITP by inducing a transient competitive block of macrophage Fc receptors by D-positive red blood cell (RBC)/IgG anti-D complexes.2
In the absence of a clear explanation for the unpredictable variation in clinical responses and hemolysis, we postulated that WinRho may have a variable avidity for RBC Rh phenotypes, reflecting the different number of D antigen sites expressed on the RBC membrane (serological “dosage effect”). Specifically, we considered the possibility that WinRho's formulation resulted in an anti-D product with greater avidity for RBCs whose Rh phenotypes were encoded by two Dgenes (eg, DCe/DcE) compared with phenotypes encoded by oneD gene (eg, DCe/ce). To investigate this hypothesis, we performed hemagglutination titers, selecting reagent RBCs that expressed either single or double doses of the D antigen, and used dilutions of WinRho as the anti-D typing reagent. We also performed monocyte monolayer assays (MMA), sensitizing reagent RBCs with WinRho (IgG anti-D) in concentrations that would be expected to occur in recipients after a standard intravenous dose of 50 μ/kg. The MMA is often useful for predicting the likelihood that IgG-coated RBCs will be sequestered and hemolyzed by mononuclear macrophages.3
The results of our direct hemagglutination tests did not show serological “dosage effect” (Table 1), nor did our MMA results reflect increased reactivity of mononuclear macrophages for RBC phenotypes encoded by two D genes (Table2). RBCs that expressed the uncommon weak-D phenotype had decreased responses in all assays. However, because the weak-D phenotype is found in only 1 in 500 patients in our hospital, these observations do not explain the more frequent variability observed in clinical responses and hemoglobin decreases. We conclude that WinRho does not manifest dosage effect for the D antigen when tested by these conventional serological methods. An ITP patient's Rh phenotype is unlikely to be a reliable predictor of clinical responses or hemolysis after treatment with WinRho.
