Elisa Smit1, Richard Lee-Kelland1, Sally Jary1, Andrew Whitelaw1, Frances Cowan1, Marianne Thoresen1,2
Department of Neonatal Neuroscience, School of Clinical Sciences, University of Bristol, UK
Department of Physiology, Institute of Basic Medical Sciences, University of Oslo, Oslo, Norway
Introduction (include hypothesis)
The incidence of intracranial haemorrhage (ICH) in term infants with neonatal encephalopathy (NE) undergoing therapeutic hypothermia (TH) is not well documented1. We postulate that infants undergoing TH are at risk of ICH due to traumatic delivery, asphyxia, and clinical factors related to NE and TH. Cranial ultrasonography (cUS) before, during and following TH and routine magnetic resonance brain imaging (MRI) allowed us to describe the incidence of ICH, and risk factors for ICH in infants undergoing TH.
Methods (include source of funding and ethical approval if required)
Cooled infants ≥36 weeks gestation fulfilling the TOBY cooling criteria, were included (n=192). With ethical permission, demographic and clinical variables were collected prospectively in our tertiary, regional cooling centre. All infants underwent cUS on days 1-4 and brain MRI scan on median day 8; both were used to identify ICH. Post mortem results were reviewed for the presence of ICH in non-survivors. Mann-Whitney U test, t-test, and Fisher-Exact test were used to compare groups. Stepwise logistic regression was used to identify factors associated with ICH. Survivors were followed up with Bayley-III neurodevelopmental evaluation at 18-24 months.
Intracranial haemorrhage was present in 70 infants (36%) and the predominant patterns of ICH were: 16% intraventricular (IVH), 21% intraparenchymal, 56% subdural/subarachnoid, 3% cerebellar, and 2% sub-galeal haemorrhage. Seventeen infants (9%) had more than one type of ICH. Vaginal birth (80% in ICH vs 43% in no ICH group) and coagulopathy (36% vs 19%) were associated with ICH. An amplitude integrated EEG pattern before cooling of continuous normal voltage with seizures was seen in 21% of infants with ICH vs 8% in those without ICH (p<0.001). Infants with IVH showed significantly more thrombocytopaenia (<50 x10*9/L) and required more inotropic support. They also had a significantly lower cognitive Bayley outcome. Coagulopathy was associated with intraparenchymal haemorrhage. Vaginal birth, higher cord pH, and renal impairment were associated with subdural and subarachnoid haemorrhage.
More than 1 in 3 cooled infants with moderate to severe NE have at least one type of ICH on brain imaging or post mortem examination. This study shows that vaginal birth and coagulopathy are associated with ICH. Cardiovascular instability appears to be a risk factor associated with IVH, whilst coagulopathy is a factor involved in intraparenchymal haemorrhage.
References (include acknowledgement here if appropriate)
1. Intracranial hemorrhage in full-term newborns: a hospital-based cohort study. Brouwer et al. Neuroradiology (2010) 52:567–576