Abstracts

Rationale: The vast majority of seizures end spontaneously. Understanding the mechanisms underlying this phenomenon would be useful not only in understanding what happens when a seizure becomes prolonged, evolving into status epilepticus, but also in designing interventions to terminate seizures. The purpose of this study is to compare termination patterns of seizures with electrographically focal onset that become secondarily generalized (sGTC) with seizures appearing to have electrographically generalized onset as seen in idiopathic generalized epilepsy (IGE) or childhood absence epilepsy (CAE). Appropriate characterization will allow us to further constrain the types of mechanisms which could be at play in ending seizures. Methods: We examined the termination pattern of a total of 167 seizures in 57 patients- 71 with focal onset that secondarily generalized (scalp EEG:45, intracranial EEG: 26) and 96 seizures with generalized onset (IGE:10, CAE: 86). Preliminary analysis consisted of visual inspection of the records to determine if seizures displayed spatially coherent termination, transition to a burst-suppression pattern, and the relationship between bursts and suppressions if present. Each seizure was then formally processed using a custom seizure detection and termination algorithm designed in MATLAB to objectively capture the electrographic features of the seizure termination phase of generalized seizures with focal versus generalized onset. Results: Our results expose characteristics that occur at the end of each type of epileptic episode which suggest distinct mechanisms. We found that while the two classes of seizures under investigation - focal and generalized - seem to terminate in their own unique way, there are common characteristics between the two types. We identified a conspicuous electrographic motif within the secondarily generalized seizure group - a "burst suppression pattern" similar to those in anesthesia recordings. We observed robust synchrony amongst channels toward seizure end and a simultaneous end in which the seizure ended in all leads at the same time. The length of suppressions significantly increased toward seizure end while the length of bursts showed no significant change. Within this dataset, there were some patients with concurrent scalp and depth electrode recordings.Depth electrode data showed (1) a band of coherence that started in the theta range and steadily decreased to the delta range at seizure termination; (2) occasional activity in a few leads during an otherwise global suppression epoch; and (3) relatively more asynchrony prior to the final burst epoch than was observed on scalp electrodes.Furthermore, we analyzed generalized seizures and found a similar burst suppression pattern, though not identical to that of the focal seizures. Seizures from patients with CAE showed a general trend of slowing and decreased amplitude at seizure end. Conclusions: These characteristic ending patterns can be used to constrain mechanistic models of spontaneous seizure cessation and inform our ideas about new therapies designed to curtail seizure intensity or end them altogether.