- Top of page
Summary: Purpose: We analyze a series of patients with mesial temporal lobe epilepsy (MTLE) associated with hippocampal sclerosis (HS) submitted to presurgical investigation with scalp sphenoidal, followed by foramen ovale electrodes (FO), and, when necessary, with depth temporal electrodes. We sought to evaluate the clinical utility of FO in patients with MTLE-HS.
Methods: We included patients who had phase I investigation with bitemporal independent seizures, nonlateralized ictal onsets, or ictal onset initiating in the side contralateral to the side of hippocampal sclerosis. Patients whose implanted FO failed to demonstrate an unambiguous unilateral ictal onset were later evaluated with depth hippocampal electrodes.
Results: Between May 1994 and December 2004, 64 patients met our inclusion criteria: 33 female (51.5%) and 31 male subjects (48.5%). The mean age at enrollment was 37.66 ± 10.6 years (range, 12–56 years). The groups with nonlateralized surface ictal EEG onsets and contralateral EEG onsets had a greater chance of lateralization with FO when compared with the group with bilateral independent seizures on surface EEG (p < 0.01). Foramen ovale electrodes lateralized the seizures in 60% of patients. Seventy percent of patients became seizure free after temporal lobectomy. Five patients were implanted with depth temporal electrodes after FO video-EEG monitoring. The depth-electrode EEG onsets confirmed the results of FO video-EEG monitoring in all patients, and the surgery was refused.
Conclusions: In MTLE-HS, FO is a reliable method for lateralization of seizures that are not clearly recorded by surface EEGs.
Mesial temporal lobe epilepsy associated with hippocampal sclerosis (MTLE-HS) is the commonest type of pharmacologically intractable epilepsy. However, after a careful preoperative evaluation, 60–70% of MTLE-HS patients can be rendered seizure free after temporal lobectomy (1).
In many candidates for temporal lobe resection, the seizure focus can be accurately lateralized and patients selected for surgery by using noninvasive investigations, such as high-resolution magnetic resonance imaging (MRI), noninvasive video-EEG monitoring, and positron emission tomography (PET) and single-photon emission computed tomography (SPECT) studies (2). Surface interictal and ictal EEG recordings remain an essential element of the presurgical evaluation in patients with MTLE-HS to demonstrate the presence of unilateral epileptogenesis. Notably, the definition of the seizure-onset zone by surface ictal EEG generally provides an accurate estimation of the epileptogenic zone, defined as the area of cortex necessary and sufficient for the generation of clinical seizures (3).
However, in ∼30% of MTLE-HS patients, scalp-sphenoidal EEG recordings fail to demonstrate an unambiguous unilateral ictal onset, showing contralateral, bitemporal independent, and nonlateralized ictal onsets, or the surface EEG recordings are not interpretable because of movement artifacts (4,5). Conversely, a bilateral scalp onset is still compatible with a unilateral mesial temporal onset and should not deter further surgical assessment (6).
When noninvasive studies remain nonconcordant or uncertain regarding the localization of seizure onset, invasive studies such as those with depth electrodes may be needed (7). Nevertheless, the proportion of patients evaluated with depth electrodes has decreased over the last decade, because of our increasing ability to localize epileptogenic lesions such as HS with high-resolution MRI, and because invasive electrodes carry considerable costs and risks of complications (3).
To record from the mesiobasal aspect of the temporal lobe, Wieser et al. (8–10) developed a subdural electrode inserted through the foramen ovale (FO). This procedure is safe and can be an alternative to invasive implantation of depth electrodes in MTLE patients who are candidates for temporal lobectomy.
Here we describe a series of patients with MTLE-HS submitted to a progressive presurgical investigation with scalp sphenoidal, followed by FO and depth temporal electrodes when necessary. Our findings agreed with previous descriptions showing that, in patients whose surface EEG fails to demonstrate an unequivocal unilateral ictal onset, FO might provide relevant and sufficient information to give an indication for resective surgery.
- Top of page
Available data from autopsy and surgical studies suggest that patients with temporal lobe epilepsy frequently have bilateral hippocampal damage (12,13). However, HS is usually asymmetrical and, when preoperative evaluation data are concordant with the side of hippocampal atrophy, temporal lobectomy can be indicated, and postsurgery outcome is usually good (1). Clinical decision making is more complicated when neurophysiologic and neuroimaging data indicate bilateral involvement of mesial structures (1,3). Although data from the present series indicate frequent bilateral involvement of mesial structures, we found evidence that FO evaluation provide accurate neurophysiologic data about lateralization of seizures that were not clearly lateralized by surface EEGs.
In other words, despite the hypothetically poor postsurgery outcome suggested by bilateral surface EEG onsets and the high rate of patients with bilateral HS, FO evaluation substantially affected the presurgical evaluation, indicating the side of temporal lobectomy in two thirds of these MTLE patients. It is important to emphasize that this proportion was similar in patients with bilateral and unilateral HS revealed by MRI.
The signal-to-noise ratio of FO is better than that in the scalp-sphenoidal electrode recordings (7–10). Therefore many electrophysiologic events with lateralizing value are detected by FO but remain undetected by scalp-sphenoidal contacts (14,15). In our series, this permitted the indication of temporal lobectomy in 60% of cases that were not lateralized by scalp-sphenoidal electrodes, with the great advantages of being nontraumatic to the brain and of carrying much less risk for the patient than the more-invasive depth electrodes. Moreover, even in a sample of patients with bilateral HS on MRI and apparent bilateral surface ictal EEG onsets, 70% of the patients had good postsurgical outcome after temporal lobectomy, provided that unilateral EEG onsets were defined by FO. This proportion is very similar to previous descriptions of patients with MTLE (1,10,11).
It should be stressed, however, that the FO recording technique addresses only specific questions: if the seizures originate in the mesiobasal temporal lobe structures and if they are consistently lateralized. In other words, a clear hypothesis should exist regarding the location of the epileptogenic zone, derived from noninvasive studies. In previous descriptions of FO, it has been stated that, in comparison to the presurgical examination using intracerebral depth recordings, the information gained by FO electrode recording is limited, and that the FO technique cannot substitute for a proper depth-electrode evaluation (9). However, this conclusion was drawn before the recent advances of neuroimaging. The advent of high-resolution MRI has allowed the proper selection of patients that might benefit from FO recordings, namely MTLE patients, excluding those patients with epileptogenic lesions outside the mesiobasal structures.
In all the five patients in whom the result of FO electrodes showed bilateral independent or nonlateralized ictal onsets, further recordings with depth electrodes confirmed the findings. This can be explained by the evidence of the intimate relation of the properly placed FO electrode with the parahippocampal gyrus, uncus, and lingual gyrus (Fig. 1C), and by the good correlation between the ictal onsets in FO and depth electrodes (Fig. 2A).
In conclusion, our results indicate that, in properly selected patients, FO is a reliable method for lateralization of seizures that are not clearly recorded by surface EEGs, usually providing sufficient information to indicate epilepsy surgery. In addition, FO implantation is relatively simple, well tolerated, and can substantially facilitate the presurgical evaluation of patients who are candidates for temporal lobectomy surgery by decreasing the risks of invasive neurophysiologic evaluation without excessive loss of information.