Abstracts

Rationale: Epilepsy surgery candidates with unsatisfactory scalp EEG localization often require invasive intracranial EEG monitoring to accurately determine seizure focus localization. Among those patients with bilateral hippocampal seizure activity, we commonly assume that the relative frequency of right and left hippocampal seizure onsets obtained from inpatient EEG recording closely represents the pattern of seizure onset that would occur in an outpatient setting. The purpose of this study was to evaluate this hypothesis by comparing the patterns of seizure lateralization obtained in an epilepsy monitoring unit with the patterns of seizure onset obtained by prolonged outpatient intracranial EEG recording. Methods: This study analyzed information from four patients with intractable temporal lobe epilepsy. Ictal scalp EEG recordings suggested bilateral temporal activity in all cases. Intracranial EEG recordings supported bilateral temporal lobe seizure activity in 3 of 4 patients. Incongruent data indicated bilateral temporal lobe seizure activity in the fourth case. All subjects received prolonged outpatient intracranial hippocampal EEG recording as participants in a FDA approved multi-center study of a responsive neurostimulator system (RNSTM). The neurostimulator system includes two intracranial recording subdural or depth hippocampal electrodes. This implantable device is capable of detecting and storing epileptic seizures. The recording duration ranged from 3 to 39 months. We analyzed 1,503 3-minute ictal outpatient intracranial EEG recordings. We compared inpatient and outpatient EEG recordings. Results: The pattern of seizure onsets obtained in the inpatient monitoring unit corresponded closely to the prolonged outpatient intracranial EEG recordings for two subjects. The third subject showed a significant discrepancy between the inpatient and outpatient recordings. In this subject, the intracranial inpatient EEG recording of 13 right onset seizures and 5 left onset seizure indicated a 0.72 probability of right sided seizure onset (0.49-0.87, 95% confidence interval). The outpatient recordings recorded a total of 1158 seizures over 37 months. The outpatient recordings showed a significantly lower 0.31 probability for right sided onset (p < 0.05). The fourth subjects had inadequate number of outpatient ictal recordings for evaluation. Conclusions: Physicians commonly assume that inpatient intra-cranial EEG recording of many seizures will correctly represent the frequency of right or left seizure onsest in the outpatient setting. Our results indicate that there are individual cases where this assumption may fail. A more detailed statistical evaluation of the pattern of seizure onsets may provide additional understanding of this process in patients with bilateral hippocampal epilepsy.