Abstracts

Seizure susceptibility may be influenced by recurrent exogenous factors as well as rhythmic state-dependent or circadian processes. The aim of this study was to determine the temporal distributions of seizures arising from the parietal, occipital, frontal, and mesial temporal lobes., Data were retrospectively analyzed for 97 consecutive adult patients whose long-term intracranial EEG monitoring led to seizure localization and resective surgery (45 mesial temporal, 23 frontal, 13 occipital, 16 parietal). For mesial temporal, only patients with hippocampal pathology were included. Seizure occurrence was determined from continuous intracranial EEG/video monitoring with complete visual and computerized review. Only seizures with electrographic changes were included in the analysis. To prevent patients with large seizure counts from biasing the results, a maximum of 8 seizures (randomly selected) per patient were used. The final analysis consisted of 535 seizures (243 mesial temporal, 132 frontal, 83 occipital, 77 parietal). For each group, the overall temporal distribution of seizures was determined by comparing the proportions occurring in eight 3-hour bins., Both parietal and occipital seizures displayed a non-uniform distribution relative to time-of-day (chi-square: p [lt] 10^-4). A clear Gaussian-like distribution was observed in both groups. Parietal seizures peaked between 4:00-7:00, with trough between 16:00-22:00. Occipital seizures peaked between 16:00-19:00, with trough between 4:00-7:00. Mesial temporal seizures were observed to follow a bimodal distribution, with the primary peak occurring between 16:00-19:00 and the secondary peak between 4:00-10:00. Frontal seizures were found most likely to occur between 4:00-7:00 and least likely to occur between 16:00-19:00., Partial seizures arising from the occipital and parietal regions of the brain have strong preferences for specific times of the day. Both follow Gaussian-like distributions that are 180^o out of phase relative to each other, perhaps explaining why previous studies failed to find patterns for aggregates of extratemporal regions. For mesial temporal seizures, a bimodal distribution was observed with the primary peak coinciding approximately with the unimodal peak reported in previous studies. This information, if elaborated on, could aid both the localization and treatment of partial epilepsy. Future studies could expand on this work by factoring in parameters such as sleep/wake state., (Supported by NIH R01-NS044102.)