Neonatal alloimmune thrombocytopenia (NAIT) is caused by maternal antibodies specific for fetal platelet antigens and is the most common cause of intracranial hemorrhage in full term infants. The antigen HPA-1a, carried on beta 3 integrin (GPIIIa), is the most common trigger for NAIT and the 2% of women who are HPA-1a-negative are at risk to produce such antibodies. Most HPA-1a antibodies causing NAIT can be easily detected but it is not rare for an HPA-1a negative mother who lacks detectable antibodies to give birth to an infant with thrombocytopenia. Several recent reports suggest that low avidity HPA-1a antibodies not detected by conventional serologic methods are responsible for some of these cases (Socher 2009, Bakchoul 2011).
To examine this question, we retroactively analyzed a cohort of 3478 suspected NAIT cases referred for laboratory diagnosis. Among 677 HPA-1a-negative mothers, we identified 61 in whom HPA-1a-specific antibodies were not detected by conventional methods. Surface plasmon resonance (SPR) analysis enables ligand-receptor interaction to be studied in real time without washing the target. Using this approach, we studied reactions of IgG from the 61 archived serum samples against purified GPIIb/IIIa isolated from group O platelets and identified 18 samples that reacted preferentially with the HPA-1a-positive version of the integrin in comparison with HPA-1a-negative integrin. Information defining clinical status was obtained on 13 of these cases by follow-up communication. Seven cases had been referred because of neonatal thrombocytopenia. Platelet nadirs ranged from 8,000/ul to 141,000/ul (median 38,000/ul). Five of the 7 infants had bleeding and were given maternal platelet transfusions and/or IVIgG. Normal platelet counts were achieved after 4 to 70 days (median 7 days). One infant had a normal platelet count, however IVIgG had been given to its mother throughout pregnancy. The remaining 5 cases were referred in mid-pregnancy because an HPA-1a-negative sister previously gave birth to an infant with NAIT. Four of these infants had normal platelets at birth; one had mild TP (platelets 125,000/ul). Only 3 of 12 mothers typed for HLA were positive for HLA-DRB3*0101, a marker found in >95% of women who make “conventional” HPA-1a antibodies during pregnancy (p < 0.001). The ability of human antibodies to cause destruction of human platelets in vivo can be studied in the NOD/SCID mouse, which lacks xenoantibodies normally found in other species (Newman 2007). Serum from 4 mothers was available in quantities sufficient for mouse studies; three of the four maternal sera caused accelerated destruction of HPA-1a-positive, but not HPA-1a-negative platelets.
These findings indicate that low-avidity HPA-1a antibodies not detectable by conventional serologic methods are made by a subset of women exposed to the HPA-1a antigen during pregnancy and that some, but not all of these antibodies are capable of causing NAIT, which is usually mild, but can be severe. Women negative for HLA-DRB3*0101 may be especially prone to produce antibodies of this type. Maternal-fetal incompatibility for platelet antigens other than HPA-1a is very common and many apparent NAIT cases not involving HPA-1a go unresolved. The possibility that low avidity antibodies specific for antigens other than HPA-1a are responsible for some of these cases deserves study.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.