Circadian patterns of generalized tonic–clonic evolutions in pediatric epilepsy patients

OBJECTIVE To investigate the sleep/wake, day/night, and 24-h periodicity of pediatric evolution to generalized tonic-clonic seizures (GTC). METHODS Charts of 407 consecutive patients aged 0-21 years undergoing continuous video-EEG monitoring for epilepsy were reviewed for the presence of GTC evolution. Seizures were characterized according to 2001 ILAE terminology. Charts were reviewed for EEG seizure localization, MRI lesion, and for seizure occurrence in 3-h time blocks, out of sleep or wakefulness, and during the day (6 AM-6 PM) or night. Analysis was done with binomial testing. Regression models were fitted using generalized estimating equations with patients as the cluster level variable. RESULTS 71 patients (32 girls, mean age 12.63 ± 5.3 years) had 223 seizures with GTC evolution. Sleep/wake seizure distribution predicted tonic-clonic evolution better than time of day, with more occurring during sleep (p<0.001). Tonic-clonic evolution occurred most frequently between 12-3 AM and 6-9 AM (p<0.05). Patients with generalized EEG onset had more tonic-clonic evolution between 9 AM and 12 PM (p<0.05). Patients with extratemporal focal seizures were more likely to evolve during sleep (p<0.001); this pattern was not found in patients with temporal or generalized seizure onset on EEG. Patients without MRI lesions were more likely to evolve between 12 AM and 3 AM (p<0.05), in the sleeping state (p<0.001), and at night (p<0.05). Logistic regression revealed that sleep and older patient age were the most important predictors of GTC evolution. CONCLUSION GTC evolution occurs most frequently out of sleep and in older patients. Our results may assist in seizure prediction, individualized treatment patterns, and potentially complication and SUDEP prevention.