Abstracts

Rationale: EEG provides information about long-term neurological outcome in infants with hypoxic-ischemic encephalopathy (HIE). The pattern and severity of injury on brain MRI is also helpful in this regard. The aim of the present study is to evaluate conventional EEG (cEEG) in predicting the severity of injury on brain MRI and outcome in neonates undergoing therapeutic hypothermia (TH) for HIE. Methods: We retrospectively reviewed all HIE cases referred to a tertiary care facility and who met the criteria for TH. cEEG was done within 24 hours after birth during TH for an average of one hour recording time. A brain MRI was done after TH (median 4 days of age) as part of standard care. Two epileptologists, blind to the patient's clinical status, independently reviewed the cEEG to assess background amplitude, continuity, variability, and seizure. cEEG severity was divided into three categories; normal to mild cEEG required background amplitude ≥ 25mV with normal continuity for age and present variability, while a severely impaired cEEG was described as discontinuity ≥ 20 seconds interburst intervals for more than 50% of recording with interburst amplitude ≤ 20 mV. Moderate impairment cEEG was defined as not meeting requirements for either of the two former categories (Table-1). The pattern and severity of injury on MRI was classified in accordance with Bonafacio et al., (2011) and distinguished between central (basal ganglia/thalamus/perirolandic), and peripheral (watershed) patterns of injury. T1, T2, and DWI sequences were scored independently, with pattern and severity classified in accordance with the highest score among all (Table-2). Multivariate logistical regression analysis was conducted using SAS statistical analysis software. Results: As part of an ongoing analysis, 24 patients' data have been analyzed. There was strong correlation between cEEG severity score and MRI pattern (R=.65, p=0.0003), such that the higher cEEG severity score predicted the central pattern of injury. There was a strong correlation between cEEG severity score and MRI severity score (R=.66, p=0.0002). Based on the scoring scheme, MRI severity scores are not completely independent of MRI pattern of injury. In order to assess the relation between cEEG and MRI severity more specifically, we conducted a posthoc analysis excluding the cases with peripheral pattern and found a stronger correlation between cEEG severity and the MRI severity (R= .76, p=0.0001). Higher severity score on cEEG was predictive of mortality (R= .58, p value= 0.0017), whereas higher severity score on MRI was marginally predictive of mortality (R=.32, p value=0.07). Conclusions: Our results suggest that the severity detected on short duration conventional EEG early in the course of TH is associated with MRI abnormalities and increased mortality. The clinical availability and predictive value of cEGG may be of clinical benefit in infants with hypoxic ischemic encephalopathy early in the course of illness.