Usefulness of maternal anti‐HPA‐1a antibody quantitation in predicting severity of foetomaternal alloimmune thrombocytopenia

Abstract

To study the clinical usefulness of maternal anti-HPA-1a antibody levels in predicting severe foetomaternal alloimmune thrombocytopenia (FMAIT). Recent studies using an international anti-HPA-1a standard have shown a correlation between maternal antibody levels and neonatal thrombocytopenia. Cut-off values for identifying high-risk pregnancies have also been suggested. In 1986-2010, HPA-1a alloimmunisation was confirmed in 84 women with 129 pregnancies. Maternal samples were obtained at delivery and during subsequent pregnancies. Anti-HPA-1a was quantified using a MAIPA assay with a detection limit of 0·8 IU mL(-1) (WHO reference serum 03/152). Antibody levels were compared with the severity of neonatal disease in the index and in the subsequent pregnancies. In the index cases, the correlation between an anti-HPA-1a level and neonatal platelet count did not reach statistical significance (n = 77, P = 0·074). However, the platelet counts and antibody levels in cases with cutaneous (n = 45) or intracranial haemorrhage (n = 7) were significantly different from cases with no evidence of bleeding (n = 20). In the subsequent pregnancies, there was a stronger association between the second trimester anti-HPA-1a level and the foetal platelet count (n = 16, P = 0·046). The positive predictive value of the maternal antibody level for a foetal platelet count <20 × 10(9) L(-1) was 90%, but the negative predictive value only 31%. Although a higher anti-HPA-1a level correlated with a more severe neonatal disease, barely detectable antibody levels were also observed in severely affected pregnancies. Cut-off values with sufficient sensitivity and specificity to identify these foetuses could not be found. A previous obstetric history still remains the most useful predictive parameter for severe FMAIT in clinical practice.