Abstracts

Rationale: Antiepileptic drugs (AEDs) are the mainstay of treatment for patients with epilepsy. Retention is the percentage of patients who started taking a drug who remain taking that drug after a given period and is a composite measure of efficacy and tolerability. Recognition of clinical characteristics associated with retention may help physicians to take informed decisions on the AED treatment. However, there is little evidence addressing clinical predictors of retention with AEDs. Our aim was to determine the clinical characteristics associated with retention at 12 months in patients with focal (partial-onset) seizures treated with eslicarbazepine acetate (ESL), in real-world setting (Euro-Esli study).  Methods: The Euro-Esli study was an exploratory, pooled analysis of data from 14 European clinical practice studies, that included 2058 patients. Kaplan–Meier survival method was used to evaluate retention rate at 12 months. Univariate and multivariate logistic regression adjusted for age and epilepsy duration was used to estimate the association between ESL retention at 12 months and baseline (prior to ESL initiation) predictor variables: gender, age, age at onset of epilepsy, epilepsy duration, presence of seizures at baseline, seizure types, monthly seizure frequency (<5 vs. ≥ 5 seizures/month), number of previous AEDs (<2 vs. ≥2 previous AEDs), number of concomitant AEDs. Results: Among 1559 patients included in this post hoc analysis, mean (standard deviation, SD) age was 43.9 (15.7) years, and 51% were male (baseline characteristics in Table 1). At 12 months, retention rate was 73.4%. Overall, 415/1559 discontinued ESL during the first 12 months of treatment. Patients in the ESL retention group had higher age at epilepsy onset (p<0.001), shorter mean duration of epilepsy (p<0.001) and lower number of monthly seizure frequency at baseline (p<0.001), than patients who have withdrawn from ESL (Table 1). In univariate logistic regression model, <2 previous AEDs (OR=2.57; p<0.001), absence of seizures at baseline (OR=1.87; p=0.014), <5 seizures/month (OR=1.73; p<0.001), age of epilepsy onset (OR=1.02; p<0.001), epilepsy duration (OR=0.98, p<0.001) and the number of concomitant AEDs at baseline (OR=0.81; p<0.001) were associated with higher ESL retention at 12 months. In multivariate regression model adjusted for age and epilepsy duration, the association of age of epilepsy onset with retention at 12 months was modified (OR=1.01; p=0.724), while the association of <2 previous AEDs, absence of seizures at baseline, <5 seizures/month, epilepsy duration, number of concomitant AEDs at baseline remained statistically significant (Table 2).  Conclusions: To our knowledge, this is the first study identifying predictors of ESL retention in the real-world setting. In this post hoc analysis of data pooled from a large cohort of patients, shorter epilepsy duration, absence of seizures at baseline, <5 seizures/months, <2 previous AEDs, and lower number of concomitant AEDs were found to be associated with increased odds of ESL retention at 12 months. Predictive factors of retention may help physicians choose appropriate AED treatment for patients. Funding: Study supported by Eisai; Study sub Analysis supported by Bial