Abstracts

Rationale: Temporal lobe epilepsy (TLE) is the most common type of epilepsy requiring surgical treatment. Identification of a clear lesion in the brain MRI is associated with a favorable post-operative prognosis. However, there is a challenging subgroup of TLE patients who show no evidence of hippocampal abnormalities in their MRI. In this group of MRI-negative TLE patients the decision for surgical treatment is more complex and the reported chance of Engel I outcome at one year varies between 40-81% (1-2). The aim of our study is to report the postsurgical outcome of a large cohort of MRI-negative TLE patients and identify presurgical factors of good outcome. Methods: This is a retrospective observational study. Access to clinical information from patient's charts has been approved by the regulatory REBs from each institution. MRI-negative TLE patients who underwent surgery from 1999 to 2014 were identified from the UCSF Epilepsy Center database and from the database from one epilepsy neurosurgeon at the Montreal Neurological Institute. Inclusion criteria were normal high-resolution pre-surgical MRI, and minimum of one-year post-surgical follow up. We excluded patients with small encephaloceles and those with pseudo-temporal or temporal-plus epilepsy defined based on intracranial EEG investigation. Demographic information, seizure frequency and EEG data were reviewed. Seizure outcomes were evaluated using Engel's classification. Results: Sixty-eight patients were included (61.8% females). Mean age at time of surgery was 36.1 years (16-61) and mean duration of epilepsy prior to surgery was 15.4 years (2-52.9). Intracranial EEG was performed in 35 (51.5 %) patients. One-year follow up outcome was: Engel I - 36 (53%) patients; Engel II - 12 (18%) patients; and Engel class III-IV - 20 (29%) patients. Four patients had surgery in the context of bilateral TLE based on seizures recorded from both sides. The presence of epilepsy risk factors (head trauma, encephalitis, febrile seizures), type of surgery (i.e., selective amygdalohippocampectony versus anterior temporal resection) and the use of intracranial EEG were not associated with seizure outcome. Patients with Engel I outcome had unilateral or no spikes more frequently in scalp EEG when compared with the Engel II-IV outcome group. [78.8% vs 51.8%, Fisher's exact test, p = 0.0269]. The outcome did not correlate with the monthly frequency of complex partial seizures prior to surgery. However, secondarily generalized tonic clonic seizures (GTCS) were more frequent in the group with poor outcomes. [Engel I-II: 3.1 seizures/month vs Engel III-IV: 7.4 seizures/month, t-test, p = 0.0143]. Conclusions: Despite the recognition of non-lesional TLE patients as a challenging group for surgical decision, more than half of these patients will favorably respond to surgery and have no disabling seizures. The use of invasive EEG did not show a prognostic implication. Patients with better outcomes had bilateral spikes less frequently in scalp EEG and a lower frequency of GTCS prior to surgery. Further evaluation of yearly outcomes following surgery will clarify predictive factors that impact long-term seizure control. References: [1] Immonen A, et al. Epilepsia 2010; 51(11):2260-2269. [2] Fong J, et al. Epilepsia 2011; 52(8):1393-1401. Funding: None