Preterm Oxygenation of the Cerebrum: Key for Erythrocyte-transfusion Threshold, a randomized controlled trial

Neonatal anemia is common in preterm infants. During their stay in the neonatal
intensive care unit, most of these high-risk infants receive at least one red
blood cell (RBC) transfusion. The RBC transfusion for the anemic infant is an
intervention that rapidly improves oxygen transport to vital organs, such as
the brain, the gut, and the kidneys. There are also complications of the RBC
transfusion. Besides risks of transfusion reactions, RBC transfusions are
associated with an increased risk of necrotizing enterocolitis (NEC),
bronchopulmonary dysplasia (BPD), retinopathy of prematurity (ROP),
deterioration of intraventricular hemorrhage (IVH). These complications lead to
an increased mortality and impaired neurodevelopmental outcomes. The lack of
knowledge on the balance of potential benefits and risk of RBC transfusion for
anemic infants, have led to controversies about the optimal threshold for RBC
transfusion for this population.
The current guidelines for RBC transfusion are based on thresholds of
hemoglobin (Hb) values. According to protocol, clinical variables are not taken
into account regarding the indication, with the exception of requirement of
artificial ventilation. Sufficient oxygen saturation in vital organs, however,
may justify withholding the transfusion despite lower Hb levels. Recently,
measuring tissue oxygen saturation has proven valuable in preterm and term
infants, and has become part of routine clincial care. Tissue oxygenation can
be continuously and non-invasively measured by near-infrared spectroscopy
(NIRS).
We intend to conduct a randomized controlled trial, comparing two treatment
strategies for RBC transfusion in preterm infants. We will compare a newly
developed strategy for treatment of preterm anemia with the current treatment
strategy. In the new strategy, transfusion Hb thresholds will be lower by 1
mmol/l, provided adequate cerebral oxygen saturation values are met.

Our primary objective is to test our hypothesis that lower Hb thresholds than
the current ones, provided adequate cerebral oxygen saturation values are met,
will lead to a better neurological outcome in preterm infants at three monts
post-term.
Our second objectives are to test the hypotheses that lower Hb values for RBC
transfusion will lead to less in-hospital mortality and a smaller prevalence of
NEC, BPD, ROP, and deterioration of IVH. Furthermore, we will evaluate whether
cerebral oxygen saturation values remain within safe limits in the intervention
group. Finally, we will test our assumption that the intervention strategy
indeed leads to less RBC transfusions during the study period.

This study will be a randomized controlled trial (RCT).

The intervention group will not receive a RBC transfusion (15-20 ml/kg
leukocyte-reduced erythrocytes) in case of an Hb threshold which is 1.0 mmol/l
lower than current guidelines, i.e. at an Hb threshold of 7.0 mmol/L or 6.0
mmol/L depending on ventilatory support, unless cerebral regional tissue oxygen
saturation is lower than 72% for at least 30 consecutive minutes during the
first four weeks after birth, or until discharge. The control group will be
treated according to current clinical NICU guidelines with the threshold for
RBC transfusion: Hb < 8.0 mmol/L if the infant is ventilated, and Hb < 7.0
mmol/L if not.

This study cannot be performed in another population, as the controversies
about the threshold for RBC transfusion are typical for anemic preterm infants.
Burden: All infants participating in the study are subjected to routine
neonatal intensive care. The study intervention poses minimal extra burden on
the patients for several reasons. First, monitoring of cerebral regional
tissue oxygen saturation is routine clinical care in all preterm infants
admitted to the NICU at the UMCG, and NIRS is a continuous and non-invasive
method to use. For the purpose if this study, extra NIRS measurements will be
needed in the time periods that Hb is near intervention and standard
thresholds, when cerebral oxygen saturation dictates the need for RBC
transfusion.
Second, evaluating GMs is a widely accepted non-invasive method to assess the
neurological neonatal outcome. At 3 months post term the infant will be
video-recorded during an outpatient clinical visit or at home for 10 minutes,
with the infant in a actively awake state, comfortably dressed, with uncovered
arms and legs.
Third, the urinary samples will be collected by a small gauze in the diaper.
This is also a non-invasive method. The samples will be collected during
routine handling moments, so the infant will not be disturbed.
Benefits and risks: The supposed new treatment strategy may reduce the total
number of received RBC transfusions, and may subsequently reduce the prevalence
of RBC related morbidity. Conversely, a lower Hb threshold for RBC transfusion
might increase the severity and duration of anemia, which is also associated
with poorer outcomes. The studies investigating these lower Hb thresholds had
even lower thresholds than in our intervention group. Moreover they did not
take cerebral oxygen saturation values into account. Both evidence for serious
side effects of RBC transfusion and of severe anemia has led to equipoise and
justifies in our opinion performing our proposed RCT.