Comment on Kiefel et al, page 3761
Severe thrombocytopenia in a 1- to 2-day-old neonate is an emergency situation because of the especially high rate of intracranial hemorrhage. In this issue of Blood, Kiefel and colleagues demonstrate that random donor platelets are often effective (24 of 27 cases) in severely thrombocytopenic neonates affected by AIT.
In this issue of Blood, Kiefel and colleagues demonstrate that random donor platelets are often effective (24 of 27 cases) in severely thrombocytopenic neonates unexpectedly affected by alloimmune thrombocytopenia (AIT). The important implication is that random donor platelets are the only universally available, immediately effective treatment for these neonates who have a high (11%-21%) incidence of intracranial hemorrhage (ICH).1
It had seemed logical that thrombocytopenic neonates with AIT would require matched platelets, because only one in 40 to 50 donors would be HPA-1b1b. Following Pearson et al's characterization of AIT in 1964,2 Aster's group3 first demonstrated that maternal platelets could be used to provide transfusions for the newborn. However, there has never been consensus on how to prepare these platelets to achieve the goals of removing maternal antibody while minimizing the platelet dysfunction caused by excess handling. Furthermore, if the mother were asked to provide platelets for her unexpectedly thrombocytopenic neonate, she often could not donate either because the infant had been transferred to a different hospital or because she had just given birth (especially by cesarean section) and was febrile or anemic. Even if the mother was able to donate, notwithstanding waiving donor testing, the platelets would not be ready for at least 4 to 12 hours. In a neonate with a very low platelet count and a high risk of bleeding, this would be too long to wait.
Matched but unrelated donors were also proposed. However, “universal donor” platelets (eg, those that were HPA-1b/1b and HPA-5a/5a) were not widely available. In England, more than half the country can obtain HPA-1b/1b and HPA-5a/5a platelets within 2 hours of request, because there is a continuous supply of these platelets from pedigreed donors at 2 sites (Cambridge and Bristol). However, this has not been the case for the rest of the world, both because of geographic inaccessibility and because the shelf life of platelets is only 5 to 7 days. Similarly, intravenous gammaglobulin (IVIG), even if effective in 24 hours, would take too long to increase the platelet count. Therefore, the dogma has been to infuse random donor platelets as a temporary, limited measure while awaiting the effects of IVIG and/or the availability of matched platelets.
Kiefel et al's figures, illustrating individual responses to random platelets, are worth a thousand words. Case 17 is a spectacular success of one random donor platelet transfusion, although case 7 is a clear failure. The other cases almost all demonstrate at least some response with IVIG and/or matched platelets, which were often administered subsequently. These findings are confirmed by our recent study, which demonstrated that 11 of 18 random donor platelet transfusions in HPA-1a incompatibility increased the platelet count more than 20 × 109/L.1
What is a “very low platelet count” for which random donor platelets and perhaps other treatment should be infused? Since the choroid plexus is fragile in the newborn, it would appear that anyone with a platelet count below 30 × 109/L soon after birth would benefit from a platelet transfusion. A higher number, such as 50 × 109/L, could be used if there was fetal or neonatal distress (eg, low Apgar score or an abnormal fetal monitor tracing). A radiologic examination of the head should be obtained soon after birth in any infant with a platelet count below 50 × 109/L, to determine whether there has been an ICH, which would require vigorous platelet replacement and maintenance of a high platelet count (eg, > 50-100 × 109/L). ▪
