Abstracts

RATIONALE: The objective of this study was to compare the reliability of different intracranial electrodes to lateralize seizure onset in temporal lobe epilepsy. Prior studies indicate that implanted depth electrodes may provide more accurate recordings than subtemporal strip electrodes. Whether there is a difference between longitudinally placed mesiotemporal depth electrodes and orthogonal electrodes is unclear.
METHODS: Six patients were involved in this study. All patients had an identical implantation protocol. 1)Longitudinal depth electodes were placed from a posterior parietal approach. The approach was modified to intentionally place electrodes in the temporal horn of the ventricle to minimize the risk of mesiotemporal injury. 2)Depth electrodes were orthogonally placed in bilateral amygdalae and hippocampae using a frameless stereotactic system. 3)Bilateral subtemporal strips were also placed. Accurate placement of all electrodes was obtained, as confirmed by post-implantation MRI. All patients were monitored in the epilepsy monitoring unit with 24 hour video-EEG with seizure detection. Fourty-six seizures were analysed. Seizure onset lateralization was determined in a blinded fashion between pairs (right versus left) of venticlular electrodes, implanted electrodes and subdural strips and compared to the final lateralization based on analysis of all electrodes.
RESULTS: Lateralization of seizure onsets was determined in 45 of 46 seizures in the ventricular electrodes, 42 of 46 seizures in orthogonally implanted electrodes and 39 of 46 seizures in parahippocampal contacts of subdural strip electrodes. EEG seizure onsets in the ventricular electrode contacts were comparable to those recorded in orthogonally implanted depth electrodes and superior to recordings from subtemporal strips except in cases with early neocortical involvement. In 7 seizures, onsets showed significant early involvement of temporal neocortex as recorded by lateral contacts of the subdural strips.
CONCLUSIONS: These preliminary data indicate that temporal ventricular electrodes may have the EEG sensitivity of electrodes transfixing mesiotemporal structures with considerably less risk of bitemporal injury. At present we have used the ventricular electrodes as an adjunct to implanted depth electrodes. In cases of possible neocortical seizure onset, subdural strips provide EEG coverage that strictly mesiotemporal electrodes including ventricular electrodes cannot. Future indications of isolated temporal horn electrodes may include investigation for lateralization of typical mesiotemporal seizures.