Abstracts

Rationale: Anti-epileptic medication therapy is not always sufficient to manage seizures. In an appropriately selected person, surgery can be an effective treatment option. Intracranial EEG monitoring is considered the gold-standard for seizure localization. The risks associated with intracranial EEG monitoring, primarily subdural grids have been documented. It has been observed that intracranial monitoring can be associated with subclinical cerebral edema which has presumably led to falsely localizing seizures. However, little is written regarding false localization associated with intracranial EEG monitoring. Methods: We undertook a retrospective review of all patients admitted to the Mayo Clinic Arizona Epilepsy Monitoring Unit (EMU) from January 2005 to May 2010. Patients who were included in the study were those who had any type of intracranial monitoring and magnetic resonance imaging (MRI)with intracranial electrodes in place. Those patient s MRI reports were then reviewed for findings of cerebral edema associated with intracranial electrodes. Results: : In the period of time reviewed, there were 50 admissions to the EMU for intracranial monitoring and 32 of those had an MRI obtained with intracranial electrodes in place. These 32 monitoring events occurred in 27 individual patients. No cases of edema were noted in patients with subdural electrodes. Six patients had subclinical cerebral edema surrounding temporal depth electrodes, representing 25% of patients with depth electrodes during the study period. In 3/6 subjects electrographic seizures without clinical manifestation were recorded from the depth electrode with surrounding edema that were distinct from typical clinical seizures and appeared to be iatrogenic. Subject 1 had typical seizures arising in the right mesial temporal lobe with subclinical seizures arising in the left temporal neocortex with left sided edema on MRI; seizure free at 8 months after right temporal lobectomy. Subject 2 had typical seizures arising in right mesial temporal lobe and subclinical seizures arising right temporal neocortical; seizure free 15 months after selective right amygdalohippocampectomy. Subject 3 had typical seizures arising right mesial temporal and inferior frontal with subclinical seizures arising left mesial temporal; with plan for more extensive right temporal and frontal intracranial implant. There were no cases of similarly discordant subclinical seizures with depth electrodes in the absence of cerebral edema on MRI. One individual had falsely localizing clinical seizures as well as PLEDs associated with a subdural grid, yet ultimately became seizure free. Conclusions: Iatrogenic, falsely localizing, subclinical, seizures were seen in 8% of patients undergoing intracranial EEG monitoring, primarily with temporal depth electrodes associated with subclinical cerebral edema on MRI. This is a concerning complication of intracranial EEG monitoring as this finding on EEG might lead epileptologists to disregard these patients as candidates for epilepsy surgery. In fact the patients in our study have done well post-operatively.