Are HPA-5b antibodies a significant cause of FNAIT and associated bleeding or merely an incidental finding?

Abstract

Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is caused by maternal IgG antibodies produced against incompatible human platelet antigens (HPA) on fetal platelets. The IgG antibodies cross the placenta and can cause clearance of fetal platelets, thrombocytopenia and bleeding, sometimes severe. Most cases are identified in first pregnancies when bleeding, bruising and thrombocytopenia are recognized after delivery of the infant. Laboratory tests of the parent’s blood that show HPA-specific antibodies in maternal serum and incompatibility between the parents for the cognate HPA confirm a diagnosis of FNAIT. In white populations, HPA-1a antibodies are the most frequently implicated (>80%), followed by HPA-5b antibodies (˜15%). It is well-established that infants affected by FNAIT involving HPA-1a antibodies frequently have severe thrombocytopenia (platelets ≤25 000/µl) associated with organ bleeding and intracranial haemorrhage (ICH). However, our understanding of the clinical significance of HPA-5 antibodies is less clear. Case reports of infants born with FNAIT due to HPA-5a or HPA-5b antibodies generally have higher platelet counts than infants born with FNAIT due to HPA-1a antibodies, and bleeding is often less severe. One possible reason for this difference is that platelets express significantly more (˜80 000 molecules/platelet) of the glycoprotein complex (GP)IIb/IIIa that carries HPA-1a on GPIIIa (CD61) than they do of GPIa/IIa (3,000–5,000 copies per platelet) that carries HPA-5 on GPIa (CD49b). However, there are exceptional cases of FNAIT involving HPA-5 antibodies, describing infants affected by severe thrombocytopenia with or without ICH.1, 2 In their paper, Vos et al report results from a large, retrospective cohort study of suspected FNAIT cases with a focus on clinical outcomes in cases with maternal HPA-5b antibodies, in an effort to determine the clinical significance of HPA-5b antibodies in FNAIT.3 Data were collected over an 18-year period from 1,864 cases of suspected FNAIT that were submitted for laboratory evaluation to the reference laboratory Sanquin Diagnostic Laboratories, (Amsterdam, the Netherlands). HPA antibodies were detected in maternal sera from 262 (14%) cases: 161 (61%) HPA-1a antibodies and 60 (23%) HPA-5b antibodies. Complete clinical histories could be obtained for 129 HPA-1a and 40 HPA-5b cases. Severe bleeding, infant postnatal and nadir platelet counts, mortality and postnatal treatment were compared between the two groups. HPA-1a cases showed more bleeding than HPA-5b cases (76% vs. 31%, respectively), however, severe bleeding (mostly ICH) was similar (11% vs. 10%). The median platelet count after birth was lower for HPA-1a cases, but the nadir platelet counts were not significantly different between groups. There were six (5%) HPA-1a and one (3%) HPA-5b cases that resulted in perinatal death. Postnatal treatments consisting of platelet transfusions, IVIgG or both, were administered more frequently in HPA-1a cases (69%) compared to HPA-5b cases (22%). Antenatal IVIgG treatment was given in five HPA-1a and 11 HPA-5b cases. Interestingly, physicians were aware of a FNAIT diagnosis because of a previously affected infant in only one case in each group. All other cases were suspected when cerebral abnormalities were detected during routine ultrasound screens. Four of the treated HPA-1a pregnancies were either terminated or resulted in fetal death, and the infants’ platelet counts for two of the antenatally detected cases were both <30 000/µl. However, infants in all 11 of the HPA-5b IVIgG-treated cases were born with platelets >120 000/µl. Overall, these findings confirm what previous studies have indicated, that HPA-1a FNAIT is associated with more severe clinical outcomes and that HPA-5b FNAIT is often, but not always, less severe. When the clinical data were stratified by the presence of other risk factors for neonatal alloimmune thrombocytopenia there were no major differences observed in the results. Two major strengths of the study are the comprehensive clinical case histories presented and the thorough platelet antibody and genotyping workups performed on parental blood, that when combined, significantly increase the accuracy and clinical significance of the findings. There are a few other notable results that deserve discussion. As mentioned,4, 5 ≥80% of FNAIT cases are reported to be associated with antibodies against HPA-1a, so why did de Vos et al only detect HPA-1a antibodies in 61% of cases? This was probably because many hospital laboratories can detect HPA-1a antibodies using commercial platelet antibody detection kits and, therefore, reference laboratories like Sanquin receive fewer HPA-1a cases and more difficult FNAIT cases for testing. In contrast, the authors detected a higher percentage (23%) of anti-HPA-5b than has been reported in previous studies (˜10–15%).4-6 This could be due to the authors’ practice of performing repeat postpartum testing 6 weeks later to ensure that HPA-1a or HPA-5b antibodies were not missed; however, the authors did not indicate how many HPA-5b antibodies were only detected by testing later maternal samples. Last, the authors detected 10 FNAIT cases with maternal anti-HPA-5b in which there was no fetal-maternal incompatibility for HPA-5b. This is not an uncommon finding, including in our own platelet reference laboratory here at Versiti, Milwaukee, WI. As the authors discovered, this finding is usually a result of a previous pregnancy with a different father, maternal blood transfusions or in rare cases, non-paternity. Why it occurs more often for HPA-5b is not known. However, the authors questioned whether HPA-5b antibodies might be an incidental finding. So, they calculated the expected frequencies of incompatibility between mother and fetus for either HPA-1a or HPA-5b and for the actual frequencies in their study cases. They found that the study frequencies were higher than expected for both HPA-1a (86% expected and 100% study) and HPA-5b (52% expected and 79% study), in support of HPA-5b antibodies not being incidentally present. Therefore, how can physicians determine which HPA-5b FNAIT cases will result in a mild or more severely affected infant? It helps to remember that in FNAIT cases associated with antibodies against any HPA, even HPA-1a, outcomes vary from modest to severe thrombocytopenia and no bleeding, to minor signs of bleeding, and no ICH to severe ICH. As a result, I agree with the authors that for pregnancies where maternal serum has antibodies against any HPA and there is documented fetal-maternal incompatibility for that HPA, antenatal IVIgG protocols should be discussed with the patients and implemented as required. A large, prospective clinical study of FNAIT cases associated with maternal HPA-5 antibodies would be required to begin to understand their true clinical significance, but such a study of a rare disorder (<1 in 1,000 births) is near impossible to conduct. However, de Vos et al announce that they have such a study underway! I look forward to the results.