Problem: Levetiracetam (LEV) is a new antiepileptic drug shown to be effective for the treatment of partial seizures in pivotal clinical trials. We investigated the long-term efficacy and tolerability of LEV as add-on therapy, regardless of seizure type, especially in persons who would not be eligible for clinical trials due to factors such as mental retardation and concomitant psychiatric disorders.
Methods: Ninety-eight patients participated and were followed for 1 year. Demographic data, seizure frequency, and side effects were recorded at baseline and during the 1-year follow-up. The first 35 patients were given LEV at a starting dose of 500 mg b.i.d. with weekly increments of 1000 mg (fast titration). The other patients were given LEV with a starting dose of 250 mg b.i.d. with weekly increments of 250 mg (slow titration).
Results: Fourteen patients were completely seizure free after titration to effective dose and 57 were responders with >50% seizure reduction for the first year. In the group with generalized seizures, 1 out of 19 became seizure free, but 8 patients had >50% decrease. Average dose at 1 year was 1900 mg (+/-900). Seventeen of 38 discontinuations were due to adverse effects and 21 were due to lack of efficacy. With fast titration, 15 out of 35 (43%) experienced tiredness during the first 12 weeks, and with slower titration 20 of 63 (32%) experienced tiredness. The difference was not statistically significant. Four out of the five patients who discontinued due to behavioral adverse events (mainly irritability) previously had behavioral problems and/or mental retardation. One patient discontinued due to psychosis.
Conclusions: Levetiracetam appears to be well tolerated in patients with severe epilepsy and shows efficacy in a long-term follow-up. Behavioral adverse events were noted in a small number of patients and occurred mainly in patients who had a history of behavioral disturbance or were mentally retarded. These data from an open population are consistent with the findings of clinical trials.