Abstracts

Rationale: The real-life efficacy and safety of Levetiracetam (LEV), an antiepileptic drug (AED) approved for reimbursement in France for the add-on treatment of partial epilepsy in 2003 has not yet been assessed. The objective was to assess one year continuation of LEV in real-life practice in France.Methods: All French ambulatory and hospital neurologists (n=1967) were asked to recruit subjects who initiated LEV from Jan. to Aug. 2005, 530 responded and 167 agreed to participate. Non participants were mainly (54%) neurologists who did not follow epileptic patients. Each subject included in this cohort was followed for one year after LEV initiation. Standardized questionnaires were used to collect demographic data, history and clinical characteristics of epilepsy (type, seizure number etc), past and actual AED use. At each visit the neurologists collected AED status in order to characterize LEV withdrawals with or without any AEDs changes. The overall one year continuation of LEV was defined as continuation of LEV regardless of a change in other AEDs. We also evaluated the rate of LEV continuation without the addition of another AED. Continuation rate was estimated by Kaplan-Meier survival analysis and LEV discontinuation predictive factors were assessed by Cox proportional hazards regression analysis by forward selection. All variables collected in the inclusion questionnaire were tested (p<0.05). Final results were adjusted on age, sex and neurologist activity. The end-point was the date of LEV discontinuation. Patients remaining on LEV at the last clinical contact were censored.Results: Among the 383 subjects included, mean age was 41 years and 55.6% were women. Mean age at seizure onset was 22.1 years. Partial epilepsy was reported for 77% (48% symptomatic, 29% cryptogenic) and generalized for 23% (15% idiopathic, 8% symptomatic or cryptogenic). The median number of seizures over the 6 months preceding the inclusion was 8. At the initiation of LEV, 13% of subjects were on monotherapy, 51% on bitherapy, 25% on tritherapy and 11% with four or more AEDs. Follow-up data were available for 368 subjects (96%): 280 continued LEV at one year, 61 discontinued LEV, 3 died and 24 were lost to follow-up. The cumulative one year probability of overall LEV continuation was 83% (78-86%) and of LEV continuation without the addition of another AED was 71% (66-75%). Prognostic factors of LEV discontinuation or addition of another AED were age<40 years: HR=1.7 (1.1-2.6) (p=0.01), seizure number over 6 months before initiation>48: HR=2.6 (1.5-4.6) (p=0.005), AED number prescribed before initiation>4: HR=1.5 (1.0-2.3) (p=0.04). Major causes of LEV withdrawal for the 61 patients concerned were insufficient efficacy (n=34), adverse events (n=30) and compliance (n=4).Conclusions: This post-licensing cohort study shows that a vast majority of patients continue LEV for at least one year. Discontinuation of LEV or addition of another AED is associated with younger age and both a high number of previous seizures or of AED used. Overall withdrawal for adverse events is rare. (Funding: unconditional grant from UCB Pharma)