Abstracts

Subclinical seizures (SCS) are electrographic seizures without behavioral symptoms. In a previous study, SCS offered favorable prognosis after temporal lobectomy. The aim of the present study is to further characterize SCS, study their prognostic significance of SCS after epilepsy surgery, and assess their value in localizing for surgical purposes.
Intracranial EEG reports for patients who underwent epilepsy surgery from 1987 to 2001 were reviewed. Seizure types, localization, seizure duration, and propagation routes were registered. Seizure outcome was classified as seizure free or not two years after surgery. We categorized the relationship between localization of SCS and clinical seizures as: 1) complete co-localization when both SCS and clinical seizures originated from the same focus, 2) Partial or no co-localization when some or all SCS and clinical seizures originated form different lobes. Ratio of SCS to clinical seizures were calculated for seizure duration and rate of occurrence. Chi-square and t-test for paired two samples and for unequal variances were used.
109 patients were included, mean age 32.7 years. 67 (61%) patients had 2727 SCS; of these, 65 patients had both clinical and SCS and 2 patients had only SCS. SCS began in the right hemisphere in 30 patients, from left hemisphere 30 patients and independently from both hemispheres in 7 patients. 56 (84%) patients had temporal lobe SCS onset; the rest had extratemporal or multilobar onset. 83% of temporal lobe SCS emanated from the mesial temporal lobe and rest emanated from temporal neocortex or simultaneously in mesial and temporal neocortex. 87% SCS and complex partial seizures (CPS), 83% of SCS and simple partial seizures (SPS) and 75.5% of SCS and generalized seizures originated in the same cortical area. Of 65 patients with both SCS and clinical seizures, complete co-localization was seen in 49 (75.3%) patients. Partial or no co-localization was seen in 16 (24.6%) patients. 34 (69.3%) completely co-localized patients were seizure free compared with 6 patients (24.6%) in the partial or non-co-localized group (p = 0.03). 76% of SCS did not spread beyond onset area. 23% spread within the same lobe or to other lobes (p = 0.0001). SCS duration was shorter compared to CPS (p = 0.005) and GTC (p = 0.0004) but were not different than SPS. Mean duration ratio for SCS to CPS was 0.57, and for SCS to GTC was 0.38. Lack of SCS did not adversely influence outcome. SCS duration and frequency ratios did not predict outcome.
SCS are observed in most patients undergoing intracranial EEG monitoring, arising in limbic structures and neocortex. They infrequently propagate beyond their area of origin. SCS are closely related to clinical seizures, and therefore aid in clinical localization for surgical purposes. Since the relationship between SCS and clinical seizures has prognostic significance, they can guide the surgical decision. The absence of SCS did not correlate with surgical prognosis.