Early Prediction of the Evolution of Self-Limited Epilepsy With Centrotemporal Spikes to Epileptic Encephalopathy With Spike-and-Wave Activation in Sleep: A Prediction Model Construction Based on Quantitative Electroencephalography Characteristics

Abstract

Aims

To predict the progression of children with self-limited epilepsy with centrotemporal spikes (SeLECTS) to epileptic encephalopathy with spike-and-wave activation in sleep (EE-SWAS).

Methods

We conducted a retrospective analysis of early clinical and electroencephalography (EEG) data. Clinical parameters included demographic and epilepsy-related characteristics. EEG were qualitatively (localization, lateralization, synchrony, non-Rolandic discharges, nondipole spikes, multiple spikes, focal slow-wave activity) and quantitatively (spike–wave index [SWI], spike–wave frequency [SWF], power spectral density [PSD], phase-locking value [PLV], phase lag index [PLI], weighted phase lag index [wPLI], characteristic path length [CPL], clustering coefficient [CC], small-worldness [Sigma]) analyzed. A logistic regression-based prediction model was further formulated and evaluated.

Results

This study included 50 children with seizure-free typical SeLECTS and 76 who developed EE-SWAS. Multivariable logistic regression revealed that early EEG features—SWF, relative PSD in the alpha band, wPLI and CPL in the delta band—were associated with the risk of encephalopathic transformation. The model demonstrated good performance with an area under the curve of 0.817 (95% confidence interval 0.736–0.898). The model showed a good fit and clinical benefit.

Conclusion

Initial quantitative EEG characteristics of SeLECTS can predict the development of EE-SWAS, suggesting distinct disease characteristics and pathogeneses in children at risk of encephalopathic transformation.