Abstracts

Rationale: In patients with bilateral independent temporal lobe epilepsy (BiTLE), the decision to perform a temporal lobe (TL) resection is fraught with uncertainty because the relative importance of various investigations has not been systematically evaluated. We explored the association between seizure outcomes, presence of a lesion and seizure laterality on intracranial stereo EEG (SEEG) in patients with BiTLE. Methods: We searched Embase and Medline up to May 31, 2012, using sub-headings and keywords related to bilateral temporal lobe seizures, EEG, surgery and seizure outcomes. Eligible studies included 1) independent bitemporal seizures diagnosed by SEEG, 2) quantified seizure laterality, 3) description of lesion/no lesion, 4) surgical description, and 5) quantification of seizure outcome stratified by subgroups of interest. We assessed the association of independent variables in patients with a good outcome (Engel Class I and II) and in those with poor outcome (Engel Class III and IV). Inidvidual patient data were pooled where appropriate and proportions were compared using exact tests. Results: 290 abstracts were independently reviewed by two individuals who achieved consensus on 36 articles selected for full text review. Six articles fulfilled all eligibility criteria. The studies yielded 40 patients with a mean follow-up of 4.7 years for 29 patients, and >1 year for the remaining 11. Twenty-eight patients (70%) had good outcome (27 Class I), of whom 20 (71.4%) had a TL lesion identified on MRI (13) or on pathology (7). Eighteen had mesial temporal sclerosis (MTS), and two had a hamartoma. Outcome was poor in 12 patients (30%); of these, 5 (42%) had a lesion (all MTS, 3 on MRI and 2 on pathology). The presence of a lesion was significantly associated with good outcome (p=0.05). All patients had surgery on the side of the lesion. The pooled proportion of seizures on SEEG that were ipsilateral to the resection did not differ between lesional (80%, range 50%-97%) and nonlesional cases (74%, range 61-88%) (p=0.14). The pooled proportion of seizures on SEEG that were ipsilateral to the resection did not differ between patients with good (78%) or bad (82%) outcome (p=0.68). In patients with good outcome, the proportion of seizures ipsilateral to the resection in individual patients ranged from 42%-95% (average 78%) for lesional cases, and from 50%-95% (average 75%) for non-lesional cases (p=0.86). Conclusions: Good outcomes can be achieved in 70% of patients with BiTLE and this is significantly more likely when a lesion is present on MRI or pathology. Although clinicians pay much attention to the seizure lateralization ratio on SEEG in BiTLE, this had no relation with surgical outcome. Therefore, in the presence of a clear MRI lesion, SEEG may not be necessary. The available evidence suggests that seizure lateralization on SEEG in BiTLE in general is of questionable importance.