A Systematic Review of EEG and MRI Features for Predicting Long-Term Neurological Outcomes in Cooled Neonates With Hypoxic-Ischemic Encephalopathy (HIE).

Neonatal hypoxic-ischemic encephalopathy (HIE) represents a significant global disease burden, but more importantly, it leaves a lasting impact of disability on individual children and their families. HIE outcome prognostication is important for guiding clinical interventions and counseling families. The objective of this study was to systematically review early electroencephalogram (EEG) and magnetic resonance imaging (MRI) features associated with long-term neurological outcomes in infants after perinatal HIE. Articles were extracted from PubMed, CINAHL, and Scopus. Twenty studies were included that assessed EEG and/or MRI patterns in neonates who underwent therapeutic hypothermia and were followed to determine long-term outcomes. Articles that did not meet the inclusion criteria were excluded. Covidence review manager (Melbourne, Australia: Covidence) was used to extract, evaluate, and synthesize review results. Of the articles included, eight focused on EEG features, eight on MRI features, and four on assessments using both EEG and MRI. Abnormal EEG background and burst suppression severity were associated with poor outcomes. Higher MRI injury scores in the basal ganglia and thalamus were also correlated with poor outcomes. Finally, studies also revealed restricted diffusion and greater lesion size in the subcortical gray matter correlated with poor outcomes. We also identified limitations in the included studies which primarily involved sample size, potential for MRI pseudonormalization, and the potential tradeoff between retention of infants able to receive long-term follow-up and attrition of those lost to follow-up. We conclude that EEG background patterns, MRI scoring, subcortical lesion burden, and MRI diffusivity are sensitive metrics for predicting outcomes. Both early EEG and MRI features may serve as high-fidelity biomarkers for secondary energy failure and for counseling families of neonates at high risk for devastating neurologic outcomes. Additionally, there is a paucity of information on the impact of HIE on brain areas outside of the standard clinical basal-ganglia and watershed patterns, especially in locations like the corpus callosum. Finally, MRI pseudonormalization may underestimate the extent of injury in these studies.