Abstracts

Rationale: Sleep is known to activate interictal epileptiform discharges and widen their field. Secondarily generalized tonic-clonic seizures (SGTCS) are more common in sleep, but it is not known if sleep has an effect on the field of ictal discharges at onset or if it influences the clinical manifestations of SGTCS. Methods: We investigated the difference between waking and sleep onset SGTCS in patients that had seizures recorded in the epilepsy monitoring unit (EMU) both in waking and in sleep. Patients were identified by reviewing all reports from the Epilepsy Monitoring Unit database over 3 years. One SGTCS in waking and one in sleep were chosen for analysis based on technical factors. We recorded the clinical semiology of the partial phase and the duration of each of the partial, tonic, and clonic phases; the latency from the first EEG sign to the first clinical sign; focality of onset; degree of artifact masking onset. Statistical analysis was done using the Wilcoxon Singed Rank test to analyze the durations of each seizure phase, Mcnemar's test to evaluate the extent of masking artifact, and Bowker's test to analysis the degree of EEG onset focality. The study was approved by the Vanderbilt IRB. Results: There were 16 patients. The mean age was 36 (range 7-58 ). The localization was temporal in 6 patients, frontal in 4, general in 3, frontotemporal in 1, multifocal in 1, and unclear in 1. The mean total seizure duration was 1.62 minutes in sleep and 1.38 minutes in waking. The mean partial phase duration was 0.72 minutes in sleep and 0.5 minutes in waking. The partial phase was almost nonexistent in the patients with frontal lobe epilepsy. The mean tonic-clonic duration was 0.89 minutes in sleep and 0.87 minutes in waking. The latency from first EEG changes to first clinical changes was 0.35 minutes in sleep and 0.15 minutes in waking (p=0.2). Artifact was present at seizure onset in waking and sleep in 7 patients, absent in both in 6, present in waking but absent in sleep in 2, and present during sleep but absent in waking in one patient. The ictal EEG pattern at onset was similar in waking and sleep, except in two patients. One of these had an ictal EEG onset with generalized irregular slow activity followed by generalized rhythmic activity in waking, but generalized rhythmic activity in sleep. Another patient had an ictal EEG onset with generalized attenuation followed by focal rhythmic activity in waking but focal irregular slow activity followed by focal rhythmic activity in sleep.Conclusions: No significant differences were found between sleep onset and waking onset seizures, in the seizure semiology of the partial phase or in the duration of the seizure and its partial and tonic-clonic phases. The latency from first EEG change to first clinical change tended to be longer in sleep (p=0.2).