Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: 25% of seizure-free patients relapse after stopping antiepileptic drugs

ACP J Club. 1994 Sept-Oct;121:53. doi:10.7326/ACPJC-1994-121-2-053

Source Citation

Berg AT, Shinnar S. Relapse following discontinuation of antiepileptic drugs: a meta-analysis. Neurology. 1994 Apr;44:601-8.



To determine the risk for relapse at 1 and 2 years after discontinuation of antiepileptic drugs, and the association of 3 clinical factors with the risk for relapse in patients with epilepsy.

Data sources

Studies were identified using MEDLINE and bibliographies of relevant papers and review articles.

Study selection

Studies were selected if they had adequate descriptions of methods and results and were designed to address the question of seizure relapse after discontinuation of antiepileptic drugs. Case series, published abstracts, review articles, and book chapters were excluded.

Data extraction

Data were extracted on study and analysis methods used, inclusion and exclusion criteria, sample size, risk for relapse at years 1 and 2, clinical features, and prognostic factors (age at onset, underlying brain abnormality, and abnormal electroencephalogram [EEG]). Authors were contacted to provide missing data.

Main results

42 studies were identified, of which 17 were excluded. Data on 5354 patients were included. The proportion of patients who relapsed ranged from 12% to 67%. The risk for relapse after withdrawal of antiepileptic drugs at 1 year was 0.25 (95% CI 0.21 to 0.30), and at 2 years it was 0.29 (CI 0.24 to 0.34), but the studies were heterogeneous (P < 0.001 for both). Compared with childhood onset-epilepsy, adolescent-onset epilepsy had a higher risk for relapse (relative risk [RR] 1.79, CI 1.46 to 2.19) as did adult-onset epilepsy (RR 1.34, CI 1.00 to 1.81). Remote symptomatic seizures had a higher risk for relapse than idiopathic seizures (RR 1.55, CI 1.21 to 1.98). Patients with abnormal EEGs had a higher risk for relapse (RR 1.45, CI 1.18 to 1.79).


In unselected groups of patients who are seizure-free, the rate of relapse after discontinuation of antiepileptic drugs is 25% at 1 year and 29% at 2 years.

Source of funding: National Institute of Neurological Disorders and Stroke.

For article reprint: Dr. A.T. Berg, School of Allied Health Professions, Northern Illinois University, DeKalb,IL 60115, USA. FAX 815-753-0720.


Epilepsy is a chronic disorder that often requires long-term antiepileptic drug treatment; unfortunately, these drugs can have long-term side effects. Until recently, many physicians were reluctant to discontinue antiepileptic drugs because the risk for seizure recurrence was thought to be dangerously high. The study by Berg and Shinnar suggests that patients may have a better chance of remaining seizure-free than previously thought.

By using meta-analysis, the authors evaluated a comprehensive group of primary studies that focused on seizure relapse after discontinuation of antiepileptic drugs. The review was carefully designed with no important methodologic problems. The risk for seizure relapse after antiepileptic drug withdrawal at 1 year was 25%, which represents the lowest risk for recurrence in patients with the most favorable clinical features.

Berg and Shinnar assessed 3 clinical features that modified the recurrence risk. The risk was increased for adult- and adolescent-onset compared with childhood-onset epilepsy in the presence of an underlying brain abnormality and an abnormal EEG. It seems reasonable, from review of several of the individual studies, that many other factors may be important, such as time seizure-free, number of medications used, duration of active epilepsy, type of EEG abnormality, type of seizure, and family history of seizure. Unfortunately, the effect of many potentially important clinical factors could not be derived from this meta-analysis. Without this information, it is difficult to predict the risk for relapse in patients who have a combination of risk factors.

The decision to withdraw antiepileptic drugs should be based on individual features. The physician should acknowledge and discuss with the patient that uncertainty exists about the prognosis and the consequences of recurrent seizure. All relevant factors, both medical and psychosocial, must be taken into account.

Barbara Scherokman, MD
Kaiser PermanenteSpringfield, Virginia, USA