Study Comparing Platelet Transfusion Volumes in Premature Babies

There are almost 6000 babies born weighing less than 1500g every year in the United Kingdom (UK) and Ireland. One in four will have low platelets. Studies prior to 2018 suggested that three quarters of those with severe thrombocytopenia would receive at least one platelet transfusion, 98% of those were given prophylactically. Clinicians regularly transfused platelets to babies with low platelets in order to try and prevent bleeding and platelet transfusion practice in this population was common despite no evidence of benefit. The PlaNeT-2/MATISSE (Platelets for Neonatal Transfusion-2/Managing Thrombocytopenia in a Special Subgroup) trial, published in 2018, questioned the validity of this practice when it demonstrated that using a lower transfusion threshold of <25 x10^9/L compared to <50 x10^9/L prevents mortality/major bleed in 7 out of 100 preterm babies with thrombocytopenia, and bronchopulmonary dysplasia in 9 out of 100.

There is little known about the full haemostatic and immunological effects of platelet transfusions in preterm babies and have yet to establish many basic features of their appropriate use given a paucity of randomised trials in many areas of neonatal transfusion practice: are they necessary; when should babies be transfued? Although platelets comprise 10% of transfused blood components they are responsible for 25-50% of reported serious adverse reactions including transfusion-related acute lung injury (TRALI). Many neonates with severe thrombocytopenia are already critically ill, may have been exposed to chorioamnionitis in utero, or have a co-existing hyper-inflammatory state due to co-morbidities such as sepsis, necrotising enterocolitis (NEC) and bronchopulmonary dysplasia. Evidence has shown that in babies there is critical and synergistic interaction between infection/inflammation and hypoxia-ischaemia. Platelet transfusions have been independently associated with infection in critically ill patients either through direct infection (stored at room temperature) or through their effects on immunity. Inflammation and the production of systemic cytokines can also disturb cerebrovascular autoregulation increasing the likelihood of brain injury. Platelets are as important for host immunity, inflammation and angiogenesis as they are for haemostasis. There are also stark haematological differences between preterm recipient and adult donor platelets. The neonatal haemostatic system is a finely balanced system, where differences in platelet function are counterbalanced by a relative hypercoagulability of neonatal blood. Neonates have shorter bleeding times despite the hyporeactivity and decreased adhesive capacity of their platelets. This balance might be disturbed when adult platelets are transfused, potentially leading to increased thrombosis.

Neonatal platelet transfusion volumes are based on long-standing historical practice. Neonatal platelet transfusion volumes administered are large in relation to recipient weight by comparison to adult volumes and have not been assessed in a randomised trial. The standard platelet transfusion volume (15 ml/Kg, within the range 10-20ml/kg) represents neonatal practice going back several decades. These volumes/doses have no evidence base and comparative dose of platelets for body weight greatly exceed those routinely used in adults. Non-bleeding adults are typically transfused a maximum of one pack of platelets per transfusion episode and therefore receive approximately 2 - 6 mL/kg platelet transfusion volumes depending on pack volume, specification and recipient weight. A neonatal platelet transfusion volume of 10-15mL/kg has been quoted as raising the platelet count of an infant by approximately 100x10^9/L within 12 hours of transfusion. Therefore neonatal platelet increments targeted in routine clinical practice greatly outweigh those targeted in adult patients (15-25x10^9/L). Reduced volume/dose could potentially have a less harmful effect on babies but conversely could also increase the number of platelet transfusions received and platelet donor exposure.

The trial hypothesis is that lowering the volume of platelet transfusions is safe without increasing the risk of morbidity due to bleeding and mortality. The total volume of platelet transfusion administered and number of platelet transfusions will also be assessed, as well as donor exposure. Changes in the levels of inflammatory markers implicated in neonatal inflammation will be assessed before and after (2 hours 30 minutes) platelet transfusion to help determine pathophysiology of potential harm. A sample of platelet transfusate will also be taken which will give an indication of donor related variation. The pilot data demonstrated differences pre- and post- platelet transfusion in numerous biomarkers, with a statistically significant increase in CXCL5, CD40, and TGF β. The significance of CXCL5, CD40 and TGF β is that they are known key thromboinflammatory modulators that are released by platelets. They could potentiate existing inflammation, NEC, lung injury or cerebral white matter injury which, could all potentially explain long term harm from platelet transfusion in babies.

Another possibility is that some characteristics of the platelets transfused are also creating morbidity. Platelets are heterogeneous and non-standardised components at transfusion. There will be variability in numbers of platelets, volume and storage age of individual donations. Blood group/compatibility may also be an important factor. These factors will be recorded as part of the exploratory outcomes.

It is important that neonatologists and haematologists understand this common neonatal therapy better in order to improve its efficacy and mitigate against potential harm. If this research findings demonstrate that use of a lower volume transfusion reduces the chance of death/major bleeding this change will have major implications for neonatal transfusion practices internationally potentially affecting a large population of vulnerable preterm infants.

Inclusion criteria

• Infants born at or before 32+0 weeks' gestation or less than 1500g at birth
• Written informed consent obtained from parent
• Admission to a participating NICU (includes postnatal transfers)
• Severe Thrombocytopenia (platelet count <25 x109/L)
• Cranial ultrasound scan undertaken less than 24 hours prior to randomisation shows no recent major intraventricular haemorrhage (IVH) or significant extension of previous haemorrhage.

Exclusion criteria

• Major congenital anomalies including neural tube defects, major structural cardiac anomalies (excluding PDA/ASD/VSD), abdominal wall defects and congenital diaphragmatic hernia and major dysmorphic features with an abnormal karyotype e.g., T21, T13, T18, Fanconi's anaemia, Thrombo-cytopenia Absent Radius syndrome).
• Infants on a palliative care pathway
• Neonates within 72 hours of major bleed
• All fetal intracranial haemorrhages excluding subependymal haemorrhage from any antenatal ultrasound scan.
• Known immune thrombocytopenia or family history of allo-immune thrombocytopenia or maternal anti platelet antibodies or maternal idiopathic thrombocytopenic purpura
• Neonates who did not receive parenteral vitamin K