Abstracts

Rationale: A multicentre, double-blind, randomized, head-to-head comparison of levetiracetam (LEV), slow-release carbamazepine (CBZ-SR), and lamotrigine (LTG) in newly diagnosed elderly patients with focal epilepsy (NCT00438451; EudraCT Number: 2005-003324-19). Methods: Patients aged 60 years or above with new onset focal epilepsy (either at least one seizure and spike discharges on EEG or a relevant lesion on CT/MRI or a total of 2 spontaneous seizures) were eligible; those with symptomatic epileptic seizures due to acute (< 2 weeks) cerebral lesions were excluded. Patients received LEV, CBZ-SR, or LTG (initial target dose of 1000mg LEV, 400mg CBZ-SR, or 100mg LTG) in a parallel group design over 58 weeks (6 weeks titration and 52 weeks maintenance). Following titration patients were treated individually in a double-blind manner depending on efficacy and tolerability on doses between 500 - 3000mg for LEV, 200 - 1200mg for CBZ-SR, or 50 - 300mg for LTG. Results: A total of n=361 patients were randomized and n=359 included in the intention-to-treat (ITT) population. Mean (±SD) age was 71.8 ± 7.5 for LEV (n=122, 34% female), 71.7 ± 6.7 for CBZ-SR (n=120, 46%), and 70.7 ± 7.4 for LTG (n=117, 41%). The mean number of seizures prior randomization were 3.8 ± 9.9, 4.8 ± 10.8, and 2.7 ± 3.1 for LEV, CBZ-SR, and LTG respectively (p < 0.05 for CBZ-SR vs. LTG, t-test). Based on the ITT population retention rates at 58 weeks (primary endpoint) were 61% (95%CI: 53-70) for LEV, 46% (37-55) for CBZ-SR, and 56% for LTG (47-65) [p-values: overall 0.048; LEV vs. CBZ-SR 0.02, LEV vs. LTG 0.36, LTG vs. CBZ-SR 0.15, Fisher exact test]. Logistic regression revealed an odds ratio (OR) for a retention to treatment at 58 weeks of 1.838 (95%CI: 1.092-3.093) for LEV vs. CBZ-SR, of 1.169 (0,689-1,984) for LEV vs. LTG, and of 0.636 (0.377-1.073) for CBZ-SR vs. LTG. The number of concomitant diseases significantly influenced 58-week retention (OR 0.921; 95%CI 0.859-0.987), i.e. chances to remain on therapy were reduced by 7.9% with each additional concurrent disease. There were no significant differences in seizure-freedom rates at week 30 (LEV 48%, CBZ-SR 39%, LTG 49%) and 58 (LEV: 43%, CBZ-SR 33%, LTG 38%) and in time-to-first seizure from randomization or after titration (after week 6). Conclusions: Results demonstrate evidence of superiority of LEV in monotherapy to CBZ-SR in new-onset focal epilepsy in the elderly. As the trial was not powered for the comparison of LEV and LTG there was no significant difference between these treatment arms. However, results do suggest that LEV should be among the first-line treatments in new-onset epilepsy of the elderly population. Future head-to-head trials with novel or other antiepileptic drugs in the elderly population need to consider LEV as comparator.