Abstracts

RATIONALE: The ideal invasive technique for patients with bitemporal epilepsy on scalp EEG is unclear. Some epilepsy programs advocate hippocampal depth electrodes. The aims of this study were: 1) to determine the outcome in patients with bitemporal epilepsy investigated with subtemporal subdural strip electrodes; 2) identify features of the non-invasive assessment that were associated with a good seizure outcome following temporal resection despite bitemporal scalp EEG ictal onsets.
METHODS: 43 patients evaluated at the University of Washington and found to have bitemporal epilepsy on scalp EEG, underwent invasive monitoring with bilateral subtemporal subdural strip electrodes. Non-invasive tests examined were the inter-ictal scalp EEG, MRI, neuropsychological evaluation and Wada test. Non-invasive test results were classified as concordant, discordant or non-lateralised in relation to ictal onset identified during invasive monitoring. Ictal invasive recordings were classified as either unilateral temporal or bitemporal with either [gt]80%, [gt]50% or no predominance to one temporal lobe. Tailored temporal lobectomies were performed and all patients were followed for [gt]1 year. Seizure outcome was graded as follows: Class I, seizure free or only a single post-operative seizure; Class II, [gte]75% reduction in seizure frequency; and Class III, [lt] 75% reduction in seizure frequency.
RESULTS: 27 patients (63%) had unilateral temporal seizure onset on invasive monitoring and 26 underwent temporal resection. 17 (65%) had a Class I, 3 (12%) a Class II and 6 (23%) a Class III outcome. 16 (37%) patients had bitemporal seizure onsets on invasive recordings. 7 underwent surgery on the side of seizure predominance. All 5 patients with [gt]80% of seizures arising from one temporal lobe had a Class I or II outcome. 2 patients with a Class III outcome had [lt]80% of seizures arising from one side. 9 patients with bitemporal seizure onsets were not offered surgery. In patients with unilateral seizures on invasive monitoring, concordant non-invasive test results significantly increased the number having a Class I outcome. A concordant MRI abnormality or neuropsychological evaluation was associated with a Class I outcome in 78% and 83% patients respectively. All except one of the patients with a Class III outcome had no concordant non-invasive tests.
CONCLUSIONS: Our results demonstrate that despite bitemporal epilepsy on scalp EEG, subdural strip electrode recordings can reliably lateralise ictal onset in most cases. Outcome following invasive monitoring with subdural strip electrodes in our series is similar to reported results using depth electrodes. A good outcome depends not only on unilateral ictal onset but also on concordant non-invasive tests. The most important aspect of the assessment of patients with bitemporal epilepsy on scalp EEG is not the invasive test used but rather the concordance of non-invasive and invasive data.