Abstracts

Rationale: A number of studies conducted in newly diagnosed and refractory patients with epilepsy have demonstrated that the efficacy of antiepileptic drug (AED) treatment varies greatly according to the pathological substrate. This post-hoc analysis was conducted to evaluate whether the efficacy of brivaracetam (BRV) differs significantly when stratified according to the type of epileptogenic lesion identified on MRI.Methods: All patients in a 12-week randomized, controlled Phase 3 study N01358 (NCT01261325) were required to have a CT/MRI scan, with a report summarizing the results sent to the sponsor for review, before being enrolled into the study. Three expert consultant epileptologists, blinded to the treatment assignment and efficacy response, independently reviewed the available neuroimaging results. They initially determined if the results were normal or abnormal. Abnormal lesions, were catergorized as epileptogenic or not. Epileptogenic lesions were subsequently stratified into 1 of 9 categories. Evaluation was performed individually by the three experts. For those patients where there was no alignment, a meeting was held to gain complete agreement on the categorizations. The ≥50% responder Odds Ratio (OR) and 95% confidence intervals (CI) as well as the median percent reduction in seizure frequency were compared amongst patients with and without epileptogenic lesions and across those with specific etiologies.Results: Of the 760 patients from the ITT population reviewed, an epileptogenic lesion was identified in 281 (37%) patients, 433 (57%) did not have epileptogenic lesions and data were missing in 46 (6%). The most common etiologies were mesial temporal sclerosis (MTS; 95/281; 34%), disorders of cortical malformation (DCM; 59/281; 21%), vascular lesions (46/281; 16%), and others (81/281; 29%). BRV was significantly more efficacious than placebo for patients with and without epileptogenic lesions on MRI. The ≥50% responder rate for BRV was significantly superior to placebo for patients with evidence of MTS (OR 5.8; CI 1.8, 18.5) but not for those with DCM (OR 3.0; CI 0.6, 15.4), vascular lesions (OR 2.2; CI 0.5, 8.7) or others (OR 2.0; CI 0.6, 6.2). The efficacy of BRV for patients with MTS (OR 5.8; CI 1.8, 18.5) was substantially better than for patients with the eight other categories of epileptogenic lesions (OR 2.2; CI=1.0, 4.6) when compared with placebo. The median percent reduction results were consistent with the ≥50% responder rate analysis. A sensitivity analysis adjusted for baseline covariates (baseline seizure frequency, number of prior AEDs, levetiracetam use) confirmed the above results for the ≥50% responder rate and median percentage reduction in seizure frequency.Conclusions: In this analysis from a large Phase 3 study, the efficacy of BRV appeared to differ according to the type of pathological substrate. BRV was found to be most efficacious in patients with radiological evidence of MTS, a subgroup of patients with a high likelihood of pharmacoresistance. These findings require confirmation in a prospective trial. UCB supported