Long-term retention rates for antiepileptic drugs: A review of long-term extension studies and comparison with brivaracetam

Antiepileptic drug (AED) retention rates are frequently reported in the literature and used to inform clinical decision-making, but methodological differences in the determination of retention rates make comparisons between trials difficult. Open-label extension (OLE) studies of AEDs in patients with focal epilepsy were identified from the literature. Retention calculation methods were reviewed, and published AED retention rates qualitatively compared with corresponding data for brivaracetam (BRV), a synaptic vesicle protein 2A ligand. The search identified 40 publications (corresponding to 17 studies of nine AEDs: eslicarbazepine, gabapentin, lacosamide, levetiracetam, oxcarbazepine, perampanel, pregabalin, topiramate and zonisamide) meeting eligibility criteria for inclusion in the review. Three methodologies to estimate retention rate were identified, which differed in whether patients randomised to placebo in the preceding randomised controlled trials (RCTs) were included or analysed separately, and whether retention was measured from the start of the OLE or of active treatment exposure. The most robust, conservative approach included all patients and measured retention from start of active treatment exposure, whether during the blinded RCT or at the start of the OLE (placebo RCT patients). Data using this method was available for five AEDs in this review, including BRV. The corresponding BRV 52 week retention rate (modal doses 50-200 mg/day; therapeutic range) was 69.8% (63.3-66.7% for other AEDs at this time point). No statistical indirect comparison was performed, as study populations were clinically heterogeneous. To avoid inconsistencies in methodologies, and allow comparison between AEDs when OLE data are the only long-term data available, retention rate analyses would benefit from the development of consistent reporting standards and guidelines.