Objective: To compare withdrawal of antiseizure medications (ASM) to continued treatment in newly diagnosed individuals achieving seizure freedom, and assess the risk of relapse and factors associated with relapse.
Methods: This is a multicenter retrospective cohort study with long-term follow-up. Patients with newly diagnosed epilepsy were identified from the medical records of 13 Italian epilepsy centers and followed up until the most recent visit or death. Seizure-free patients discontinuing treatment were compared to patients who maintained treatment for baseline characteristics. Treatment was stopped upon clinical judgment. The probability of relapse was calculated with the Kaplan-Meier method. Demographic, clinical, and instrumental variables associated with relapse were assessed with Cox proportional hazards models.
Results: One thousand and six patients aged 1 month to 72 years at diagnosis were enrolled and followed up for 17,892 person-years (median follow-up, 9.9 years). Three hundred and twenty patients (31.8%) underwent one or more treatment discontinuations. Factors associated with ASM withdrawal were younger age at remission and normal psychiatric examination. The probability of relapse after the first withdrawal was 16% at six months, 24% at 12 months, and 36%, 45%, and 53% at three, five, and ten years, respectively. The probability of remission after the first relapse was 59% at one month, 67%, 72, and 76% at three, six, and 12 months, respectively. Variables associated with relapse were age 14+ years, structural etiology, abnormal neuroimaging, ASM initiation after a single seizure, and symptomatic/cryptogenic epilepsy.
Conclusions: About one half of seizure-free patients stopping ASM relapse in 10 years. However, the possibility of remission after relapse is high, particularly in children and patients with idiopathic/cryptogenic epilepsy. Treatment deprescription might be encouraged at least in these patients.
Keywords: Antiseizure medications; Deprescribing; Epilepsy; Seizure recurrence.
Copyright © 2021 Elsevier Inc. All rights reserved.