Early characterization of hypoxic-ischemic encephalopathy: contribution of biochemical, imaging and electrophysiological markers to patient profiling in the ancillary cohort LyTONEPAL

Objective To identify and characterize different profiles of newborns with HIE according to clinical, biochemical, electrophysiological and brain imaging data. Study Design An analysis was conducted on the LyTONEPAL ancillary cohort, including 204 newborns with mild to severe HIE. Multiple factor analysis and hierarchical ascendant classification were performed using clinical, biochemical, electrophysiological and imaging data to establish clusters representing the cohort. Results The combination of electrophysiological, biochemical, and imaging data provided the best representation of the ancillary cohort, accounting for 36.5% of the explained variance. Two clusters of patients were identified using these factors. Patients allocated to cluster 2 were more likely to present severe HIE (70.8%; P < 0.001), severe electroencephalogram (EEG) patterns (65.2%; P < 0.001), brain injuries (66.3%; P < 0.001) and biochemical abnormalities in pH (6.91 ± 0.18; P < 0.05), lactate levels (14.7 ± 5.5; P < 0.001), and glycemia ( P < 0.05). By contrast, cluster 1 was predominantly represented by neonates with mild HIE symptoms, better EEG patterns and biochemical markers and no brain lesions. There were no significant differences in maternal and obstetrical data, particularly sentinel events, between the clusters. Conclusions Early consideration of the association of biochemical, electrophysiological and brain imaging data in neonates with HIE would allow a better and more precise characterization of the HIE profile of the newborn and its severity, compared to the sole clinical evaluation through Sarnat or Thompson score in this population.