Abstracts

Rationale: It is commonly believed that in patients with potential multifocal epilepsy, seizures occuring in a cluster tend to occur from the same focus and therefore seizures in clusters may be less helpful in presurgical evaluations of seizure localization. Recent investigations of responsive neurostimulation have provided opportunities to study long-term recordings from intracranial electrodes in humans. This allows assessment of seizure onset over time in relation to seizure clusters. Methods: Two patients with known bilateral independent mesial temporal lobe onset seizures had implantation of bilateral depth electrodes arrays in conjunction with the pivotal trials of responsive neurostimulation with the NeuroPace(TM) device. Data were collected and downloaded to a remote server over a 10-month period; antiepileptic medications were not changed during this period. The side of seizure onset and date and time of occurrence of each seizure were noted. A cluster of seizures was defined by the occurrence of more than one seizure during a fixed window of time. Seizures were assigned to either Non-Cluster Seizures (NCS) or Cluster Seizures (CS) groups. The side of seizure onset was also compared to the previous seizure. Therefore each seizure was assigned a value of 0 if its onset was ipsilateral to the previous seizure and 1 if the onset was contralateral to the previous seizure. The number of seizures for each group and each value were collected in a contingency table for both patients. The group factor distinguishes seizures occuring within a cluster or not (NCS vs CS). Chi-square statistic was not corrected for continuity as the number of seizures was sufficient in all cases. Results: Using a window of 24-hour for the definition of a cluster, NCS shows a baseline L/R ratio of 22/21 for patient 1 and 6/39 for patient 2 and CS shows a L/R ratio of 17/15 and 17/38 respectively. Patient 1 shows no difference in L/R ratio when seizures occur in clusters (p=0.41) but patient 2 shows a greater occurrence of left-sided seizures (p=0.037) during cluster. Contingency tables for the change of side of onset of seizures show that neither patient has a change in the distribution of change of onset side when seizures occur in cluster (Patient 1: p=0.41 and Patient 2: p=0.88). Conclusions: n these two patients with known independent bilateral mesial temporal onset seizures, long-term recordings with stable antiepileptic drug regimens do not support the hypothesis that seizures occuring in clusters are more likely to originate from the same seizure focus. Support by Neuropace, Inc. (GKB)