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


Immediate antiepileptic drugs reduced seizure recurrence at 2 years after a first unprovoked tonic-clonic seizure

ACP J Club. 1993 Sept-Oct;119:34. doi:10.7326/ACPJC-1993-119-2-034

Source Citation

First Seizure Trial Group. Randomized clinical trial on the efficacy of antiepileptic drugs in reducing the risk for relapse after a first unprovoked tonic-clonic seizure. Neurology. 1993 Mar;43:478-83.



To compare the effectiveness of immediate antiepileptic drug (AED) treatment with treatment only after seizure recurrence in patients with a first unprovoked tonic-clonic seizure.


An interim report of a 2-year, randomized controlled trial.


35 clinical centers in Italy.


397 of 807 patients identified as having a first-time seizure were available for interim analysis. All were ≥ 2 years of age and had been seen within 7 days of a first witnessed, unprovoked seizure (72% > 16 y, 168 women). Exclusion criteria were isolated partial seizures, > 1 seizure within 24 hours, seizure recurrence before randomization, acute symptomatic seizures (associated with stroke or trauma), progressive neurologic disorders, metabolic disorders, alcohol or drug abuse, use of epileptogenic drugs, or overt psychological disorders.


Patients were randomized to immediate treatment (n = 204) or treatment only after seizure recurrence (n = 193). Physicians chose carbamazepine, phenytoin, phenobarbital, or sodium valproate at their discretion and modified doses to bring plasma levels to the therapeutic range within 1 month.

Main outcome measure

Seizure recurrence.

Main results

103 patients took phenobarbital, 63 took carbamazepine, 33 took sodium valproate, and 5 took phenytoin. Intention-to-treat analysis was used. 41 patients (20%) discontinued their medication; none stopped on the advice of their physician. 14 patients discontinued their medication because of side effects. Patients who were allocated to treatment had fewer relapses during follow-up than did patients in the control group (Table). The probability of relapse at 2 years was 25% in the treatment group and 51% in the control group. Using the Cox proportional hazards model adjusted for all prognostic factors, the hazards ratio for relapse was 2.8 for patients who were not treated (95% CI 1.9 to 4.2). Independent of the treatment effect, patients < 16 years had a 2-fold increase in recurrence; patients showing epileptiform abnormalities on electroencephalography (EEG) had a 1.7-fold increased risk for relapse.


Treatment with antiepileptic drugs after a first unprovoked tonic-clonic seizure reduced the 2-year risk for recurrence of seizures by approximately 50%. Electroencephalographic abnormalities and young age increased the risk for relapse, independent of treatment status.

Source of funding: CIBA-GEIGY.

For article reprint: Dr. M. Musicco, Instituto di Tecnologie Biomediche Avanzate, I.T.B.A., CNR, Via Ampere 56, 20131 Milan, Italy. FAX 39-2-236-3788.

Table. Immediate vs delayed (after seizure recurrence) antiepileptic drugs for a first unprovoked seizure*

Outcome at 2 years Immediate antiepileptic drugs Delayed antiepileptic drugs RRR (95% CI) NNT (CI)
Seizure recurrence 18% 39% 55% (36 to 68) 5 ( 3 to 8)

*Abbreviations defined in Glossary; RRR, NNT, and CI calculated from data in article.


The patient who has had a single seizure presents a difficult problem for clinical epileptologists. The workup is reasonably well established: history and physical examination, an EEG, and brain imaging using computed tomography or preferably a magnetic resonance scan (1) (except for a few clinical syndromes, such as absence epilepsy, in which imaging is not useful). The clinical circumstances (e.g., ethanol withdrawal) may mitigate against therapy. At this point, however, consensus is lost. Treatment of first unprovoked tonic-clonic seizures has been a matter of opinion with few data to support conclusions. In the United States, AED treatment is frequently initiated, whereas in Great Britain it is not. Prospective studies suggest a recurrence rate after a first unprovoked seizure of 23% to 52%; the cause of the seizure and the EEG abnormalities are the strongest predictors of outcome (2). One randomized trial of 31 children showed that carbamazepine decreased the risk for a recurrent seizure when compared with no medication (3).

The conclusions of this study are limited for 3 reasons. First, although it addresses the clinical conundrum of generalized seizures, it does not instruct us about partial seizures without generalization. Second, although the sample size was sufficient to show that AEDs reduced the risk for relapse during the study, whether this translates into long-term prevention of epilepsy remains to be shown. Third, because the physicians chose which of 4 AEDs to prescribe, the study does not show which drugs are most effective. Nevertheless, this study argues strongly for initiating AED treatment after a first unprovoked seizure.

Thomas P. Bleck, MD
University of VirginiaCharlottesville, Virginia, USA


1. Bergen D, Bleck T, Ramsey R, et al. Magnetic resonance imaging as a sensitive and specific predictor of neoplasms removed for intractable epilepsy. Epilepsia. 1989;30:318-21.

2. Berg AT, Shinnar S. The risk of seizure recurrence following a first unprovoked seizure: a quantitative review. Neurology. 1991;41:965-72.

3. Camfield P, Camfield C, Dooley J, et al. A randomized study of carbamazepine versus no medication after a first unprovoked seizure in childhood. Neurology. 1989;39:851-2.