The Role of Platelet Transfusions to Manage Bleeding: Lessons From Randomized Trials (Including Neonates and Children)
Simon Stanworth, M.D.
University of Oxford
Platelet transfusions are the second most commonly administered cellular blood component, after red cells. The focus of much of the research on the appropriate use to date has been in adult patients with hematological cancers, looking at questions of optimal dose or safe platelet count threshold. The main message from these trials is that there is no benefit to the use of more liberal policies for platelet transfusions, and indeed the most recent randomized trials have been designed to compare strategies of no-prophylaxis against prophylaxis at platelet count thresholds 10 × 109/L in patients with hematological cancers.
One non-hematology clinical setting where platelet transfusions continue to be frequently administered is with infants. Traditionally, neonatologists have given platelet transfusions at much higher threshold platelet counts, often at or above 50 × 109/L, compared with practice in patients with hematological cancers. In a first multicenter randomized trial in preterm babies who developed severe thrombocytopenia, babies were randomly allocated to platelet transfusions at either restrictive thresholds of 25 × 109/L or liberal thresholds of 50 × 109/L. The primary outcome for the PlaNet-2/MATISSE trial was mortality or major bleeding within 28 days; secondary outcomes included a range of endpoints important to neonatal practice including bronchopulmonary dysplasia.
A total of 660 babies, representative of a sick cohort of preterm infants, were randomized in neonatal intensive care units across three countries in Europe. A total of 26% (85/324) of babies in the liberal group and 19% (61/329) of babies in the restrictive group developed a new major bleed or died by study day 28 (OR = 1.57, 95% CI 1.06-2.32; P = 0.02). A total of 63% (169/269) of babies in the liberal group were found to develop bronchopulmonary dysplasia versus 54% (153/281) in the restrictive group (OR = 1.54, 95% CI 1.03-2.30). The higher rate of mortality/major bleeding within 28 days of randomization in the liberal group was unexpected, although it was consistent with findings in prior observational studies.
The results are practice-changing for the use of platelets in sick preterm infants. Questions that remain unanswered include the mechanism of effect. Platelets have actions beyond primary hemostasis, which include immunological and inflammatory effects. Future research needs to better define the full impact of platelet transfusions in recipients. We can then move towards a personalized medicine approach for the use of platelets in different patient groups, where the benefits outweigh the risks of harm.