Novel approaches to identify NICU patients who would benefit from platelet transfusions

Introduction

The underlying reasons for bleeding are multifactorial and likely include a combination of hematologic (i.e., platelet, von Willebrand factor [vWF], coagulation factors) and environmental factors such as vessel wall integrity, inflammation, and hemodynamic status. However, it is unknown to what extent these components contribute to bleeding. Since platelets have a major role in primary hemostasis, a low platelet count is considered a risk factor for bleeding.

Thrombocytopenia is common in the neonatal intensive care unit (NICU), and if severe it is treated with platelet transfusions to prevent bleeding. Approximately 5% to 15% of preterm neonates admitted to the NICUs develop major bleeding (most commonly intraventricular hemorrhage [IVH]) during their NICU stay, potentially leading to lifelong consequences. ^, Of thrombocytopenic neonates, 25% have one or more platelet counts below 50 × 10 ⁹ /L (severe thrombocytopenia), and 9% of these infants experience significant bleeding. Platelet transfusions are frequently given to neonates whose platelet counts fall below an arbitrary (and highly variable) trigger, in an attempt to prevent bleeding. ^{, ,} However, several studies from the past decade have shown that platelet count is a poor predictor of bleeding in neonates, ^{, ,} suggesting that other hematologic and clinical factors (i.e., platelet function, hematocrit, vWF concentrations, gestational age, postnatal age, diagnosis, severity of illness) may be stronger determinants of bleeding risk than platelet count alone. Most of these factors are highly dynamic in neonates, whose hemostatic balance is also different from that of adults. Specifically, neonatal platelets are hyporesponsive to most agonists compared with adult platelets, and this hyporesponsiveness is more pronounced in preterm infants. In healthy full-term and preterm neonates, the effects of hypofunctional platelets are counteracted by their higher hematocrits, mean corpuscular volumes, and concentrations of vWF (all prothrombotic factors that accelerate clotting), paradoxically leading to shorter bleeding times in neonates compared to adults. Any alteration in these compensatory factors can disrupt the delicate primary hemostatic balance of neonates. Neonatal platelet function also changes over time and reaches near-adult levels by 10 to 14 days of life in preterm as well as in full-term neonates. ^{, , ,}

The recent recognition of the central effects of platelets on inflammation also raised the question of whether platelet transfusions could be harmful to some infants, depending on the underlying etiology of the thrombocytopenia and the clinical context. This hypothesis was supported by a number of observational studies that reported an association between number of platelet transfusions and increased mortality and morbidity among NICU patients. At least one study reported an association between number of platelet transfusion and worse outcomes in necrotizing enterocolitis (NEC), and it was postulated that this could be related to inflammatory mediators released by platelets. More recently, the largest randomized controlled trial of platelet transfusions in thrombocytopenic preterm neonates showed a higher rate of bleeding and death among infants transfused for platelet counts below 50 × 10 ⁹ /L compared to those transfused for platelet counts below 25 × 10 ⁹ /L.

Currently, over 75% of neonates with severe thrombocytopenia (platelet count <50 × 10 ⁹ /L) are treated with platelet transfusions to prevent bleeding; however, only 9% to 11% of these neonates develop major bleeding. ^, In addition, there is no evidence supporting that platelet transfusions can prevent neonatal bleeding. The facts that (1) platelet count is a poor predictor of bleeding, (2) platelet transfusions in preterm neonates have been shown to be harmful, and (3) platelet transfusions may not reduce bleeding risk in thrombocytopenic neonates highlight the need for revised and individualized approaches to neonatal platelet transfusion decisions.

Bleeding prediction