Administration of Anti-Platelet Antibodies Prevents the Anti-Platelet Immune Response and Bleeding Complications of Neonatal Immune Thrombocytopenia in a Murine Model.

Abstract 223 Background: The human platelet antigen (HPA) 1a is a potent immunogen located on the β3 integrin. Ten % of pregnant HPA1a negative women produce antibodies against the HPA1a antigen if the foetus is HPA1a positive. Fetal/neonatal immune thrombocytopenia (FNIT) can occur if the mother develops alloantibodies against fetal platelets, with intracranial haemorrhage as the most severe complication. The current opinion has been that immunization against the HPA1a antigen takes place during the first HPA1 non-compatible pregnancy. However, results from a large and recent screening study in Norway found that the majority (75%) of women were immunized around time of delivery, and not so often during pregnancy. This indicates that FNIT could be more similar to haemolytic disease of the newborn (HDN) than previously thought. To prevent HDN, antibody mediated immune suppression (AMIS) is induced by administration of anti-D antibodies in connection with RhD-negative pregnancies. The same principle could be used to prevent FNIT by administration of anti-HPA1a antibodies in HPA1a-negative pregnancies. We have previously established a murine model of FNIT using β3 integrin-deficient (β3−/−) mice. The first aim of the current project was to test whether administration of human anti-HPA1a IgG could suppress the anti-human platelet immune response in β3−/− mice after transfusion of human HPA1a positive platelets. For the second part of the project, we used a pure murine model to test whether administration of murine anti-β3 antibodies transfused after delivery could induce AMIS and prevent bleeding complications of FNIT in the subsequent pregnancies. Methods: Human/murine model: Human IgG from 5 donors with high levels of anti-HPA1a antibodies was purified by Protein G affinity chromatography. Purified IgG from one male donor without detectable anti-platelet specific antibodies was used as control IgG. Human platelets were isolated from an HPA1a positive donor. β3−/− mice were immunized by one tail vein transfusion with 2 × 106 human HPA1a positive platelets, with or without subsequent transfusion of 900ug human IgG (100% saturation). After 7 days, the mice were bled and sera collected. The anti-human platelet immune response was analyzed via flow cytometry, using FITC-conjugated goat anti-mouse IgG as detection antibody. Six mice were injected with anti-HPA1a containing IgG. Control IgG (n=6) or no IgG (n=4) were used as negative controls. Pure murine model: High-titer anti-β3 sera were produced by 4 weekly transfusions of 108 wild type (WT) platelets to β3−/− mice. Naïve β3−/− female mice were bred with naïve β3−/− male mice. Within 24 hours of delivery, the mother was transfused with 108 WT platelets with or without immediate transfusion of anti-β3 sera. The transfusions were repeated one week after delivery and the same females were bred again with WT male BALB/c mice. The anti-β3 immune response was analyzed via flow cytometry, using FITC-conjugated goat anti-mouse IgG. The FNIT phenotype was monitored and all live pups were bled from the carotid vein to determine platelet count. Results: Administration of purified anti-HPA1a IgG significantly suppressed the anti-human platelet immune response in β3−/− mice after transfusion of HPA1a positive platelets as compared with control IgG (p < 0.05). In the pure murine model of FNIT, the anti-β3 immune response was markedly suppressed during the subsequent pregnancy in the mice treated with anti-β3 sera. Two out of three mice receiving anti-β3 sera treatment delivered live pups with moderate thrombocytopenia without signs of haemorrhage (mean platelet count 217 ×106/mL). The third mouse receiving anti-β3 sera delivered dead pups. In contrast, all female mice (n = 3) without anti-β3 sera treatment miscarried. Conclusions: We have demonstrated in vivo that AMIS can be induced by administration of anti-platelet antibodies using a murine model of FNIT. Preliminary data indicates that bleeding complications of FNIT can be prevented with this prophylactic approach. Disclosures: Skogen: Prophylix Pharma a/s: Employment, Equity Ownership. Killie:Prophylix Pharma a/s: Equity Ownership. Husebekk:Prophylix Pharma a/s: Equity Ownership. Kjeldsen-Kragh:Prophylix Pharma a/s: Equity Ownership.