Predictive Factors of Postneonatal Epilepsy: The Significance of Follow-up EEG Changes

Seon, Min Gyu; Park, Moon Sung; Lee, Jang Hoon; Jung, Da Eun

Abstract

PURPOSE
To describe the prevalence and severity of postneonatal epilepsy after neonatal seizures in term neonates as well as to evaluate the predictive factors of postneonatal epilepsy.
METHODS
Retrospective analysis of 50 children who experienced neonatal seizures. In children with at least 12 months of follow-up data, the univariate and multivariate logistic regression analysis was applied in order to determine the predictive factors of postneonatal epilepsy. Electroencephalography (EEG), neuroimaging studies, and other clinical variables were systematically analyzed. Infants with abnormal EEG recordings in the initial studies underwent a follow-up EEG examination at 1 or 3 months later.
RESULTS
Twelve of 50 neonates with neonatal seizures (24%) developed postneonatal epilepsy. Seventy-five percent (9 of 12) of the children with postneonatal epilepsy were eventually seizure-free without AED (antiepileptic drug), and 25% (3 of 12) had seizures at the last follow-up (modified angel classification 2 or 3, mean follow-up period; 52 months). On the univariate logistic regression analysis, abnormal EEG, Magnetic resonance imaging (MRI) findings, combined with encephalopathy and the number of AEDs were correlated with postneonatal epilepsy (P<0.05). On the multivariate analysis, the persistent abnormality on the follow-up EEG was correlated with postneonatal epilepsy (adjusted odds ratio=20.78; P=0.016).
CONCLUSION
The number of intractable cases was relatively low, indicating good prognosis in postneonatal epilepsy. Abnormal EEG, MRI findings, combined with encephalopathy, and the number of AEDs were very good predictors of postneonatal epilepsy. The persistent abnormality in the follow-up EEG was more frequently seen in postneonatal epilepsy patients.