NEONATAL SOCIETY ABSTRACTS
Correlation functional resting state networks and cerebral function monitoring in hypoxic ischaemic encephalopathy
Presented at the Neonatal Society 2018 Spring Meeting (programme).
Bravar G1,3, Vecchiato K1, Steinweg J1, Kelly C1, Hughes E1, Victor S1, Singh R2, Edwards AD1, Arichi T1,2
1 Centre for the Developing Brain, School of Biomedical Engineering and Imaging Sciences, King's College
2 Paediatric Neurosciences, Evelina London Children Hospital
3 University of Trieste, Trieste, Italy
Background: Hypoxic Ischemic Encephalopathy (HIE) is a major cause of death and neurodisability in term neonates. Cerebral function monitoring (CFM) is routinely used to assess the severity of neonatal encephalopathy, identify seizures, and as a prognostic tool to predict neurodevelopmental outcome. Brain activity can also be assessed using functional Magnetic Resonance Imaging (fMRI) by studying resting state networks (RSNs) which are established during the perinatal period (Doria et al. 2010). We therefore aimed to test the hypothesis that the degree of encephalopathy as assessed by CFM is correlated with altered RSNs in neonates with HIE.
Methods: We studied 21 infants with a diagnosis of HIE (median [range] gestational age (GA) at birth: 39⁺6 [36+6- 41+6]) treated with hypothermia. Infants were classified into two groups on the basis of CFM trace recorded at 24 hours following delivery: 11 infants with a “normal” CFM trace, and 10 infants with an abnormal CFM trace defined as the presence of an abnormal supressed background. Following rewarming, 6.5 minutes of resting state fMRI data were acquired at median 6 days of age (range 4-10 days) with a 3T Philips MR system. Written parental consent was taken before all data collection (REC: 09/H070/84). Data analysis was performed using tools implemented in FSL. RSN were defined using temporal concatenation ICA, and group averages and differences were identified using dual regression, a general linear model (incorporating the CFM appearances, the anatomical pattern of brain injury, presence of seizures, and age at scan as explanatory variables) and permutation testing.
Results: RSNs including the motor, somatosensory, visual, auditory, and components of the default mode network were identified in both groups. Infants with a severely abnormal CFM trace had a tendency towards more unilateral RSNs without additional activity in the thalamus in comparison to those with a normal CFM trace. Significant differences (p<0.05) were identified in the distribution of 3 RSNs encompassing the thalamus (figure), brainstem, and bilateral frontal lobes.
Conclusion: A severely abnormal CFM at 24 hours following delivery was associated with significantly altered RSNs at 6 days of age. Although RSNs were present in all subjects, there was reduced interhemispheric and long distance connectivity in infants with a greater degree of early encephalopathy. This effect was most significant in networks which involved the deep grey matter and brainstem, which is of particular interest as they are areas commonly injured in infants with severe HIE and are known to have a role in consciousness. This suggests that HIE disrupts the key functional connections of these important regions to areas across the cortex and may explain why injury to these regions results in poor cognitive outcome.
Corresponding author: email@example.com
Doria V et al. Emergence of resting state networks in the preterm human brain. PNAS 2010, 107 (46) 20015-20