Long‐term efficacy, tolerability, and retention of brivaracetam in epilepsy treatment: A longitudinal multicenter study with up to 5 years of follow‐up

Objective This study was undertaken to evaluate the long‐term efficacy, retention, and tolerability of add‐on brivaracetam (BRV) in clinical practice. Methods A multicenter, retrospective cohort study recruited all patients who initiated BRV between February and November 2016, with observation until February 2021. Results Long‐term data for 262 patients (mean age = 40 years, range = 5–81 years, 129 men) were analyzed, including 227 (87%) diagnosed with focal epilepsy, 19 (7%) with genetic generalized epilepsy, and 16 (6%) with other or unclassified epilepsy syndromes. Only 26 (10%) patients had never received levetiracetam (LEV), whereas 133 (50.8%) were switched from LEV. The length of BRV exposure ranged from 1 day to 5 years, with a median retention time of 1.6 years, resulting in a total BRV exposure time of 6829 months (569 years). The retention rate was 61.1% at 12 months, with a reported efficacy of 33.1% (79/239; 50% responder rate, 23 patients lost‐to‐follow‐up), including 10.9% reported as seizure‐free. The retention rate for the entire study period was 50.8%, and at last follow‐up, 133 patients were receiving BRV at a mean dose of 222 ± 104 mg (median = 200, range = 25–400), including 52 (39.1%) who exceeded the recommended upper dose of 200 mg. Fewer concomitant antiseizure medications and switching from LEV to BRV correlated with better short‐term responses, but no investigated parameters correlated with positive long‐term outcomes. BRV was discontinued in 63 (24%) patients due to insufficient efficacy, in 29 (11%) for psychobehavioral adverse events, in 25 (10%) for other adverse events, and in 24 (9%) for other reasons. Significance BRV showed a clinically useful 50% responder rate of 33% at 12 months and overall retention of >50%, despite 90% of included patients having previous LEV exposure. BRV was well tolerated; however, psychobehavioral adverse events occurred in one out of 10 patients. Although we identified short‐term response and retention predictors, we could not identify significant predictors for long‐term outcomes.