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


Review: 6 newer drugs reduce seizures with no detectable differences among them

ACP J Club. 1997 May-Jun;126:72. doi:10.7326/ACPJC-1997-126-3-072

Source Citation

Marson AG, Kadir ZA, Chadwick DW. New antiepileptic drugs: a systematic review of their efficacy and tolerability. BMJ. 1996 Nov 9; 313:1169-74.



To determine the efficacy and tolerability of 6 different antiepileptic drugs in controlling seizures.

Data sources

Studies were identified by searching MEDLINE (1966 to 1995) and by hand-searching journals. Relevant drug companies were also contacted.

Study selection

Studies were selected if they were randomized controlled or cross-over trials that evaluated add-on treatment with gabapentin, lamotrigine, tiagabine, topiramate, vigabatrin, or zonisamide; if they included only patients with partial epilepsy; if treatment lasted ≥ 8 weeks; and if seizures were reported as an outcome. Trials that randomly assigned patients only after they showed a predetermined response to treatment were excluded.

Data extraction

Study design, length of treatment, treatment and dosage, proportion of patients with a 50% reduction in frequency of seizures compared with baseline, and patient withdrawal rate.

Main results

28 studies (3883 patients) met the selection criteria. 20 studies were published; 8 were unpublished. Each of the 6 drugs was superior to placebo in achieving a 50% reduction in the frequency of seizures. The pooled absolute benefit increase (ABI), number needed to treat (NNT) for 1 additional patient to achieve a 50% reduction in the frequency of seizures, and the relative benefit increase (RBI) for each drug compared with placebo are listed in the Table. For gabapentin, tiagabine, and vigabatrin, increasing efficacy occurred with increasing doses. For topiramate, the therapeutic effect reached a plateau at doses > 600 mg/d. Patients were more likely to stop taking tiagabine (odds ratio [OR] for withdrawal 1.8, CI 1.2 to 2.7), topiramate (OR 2.4, CI 1.4 to 4.1), vigabatrin (OR 2.6, CI 1.3 to 5.3), or zonisamide (OR 5.7, CI 1.8 to 18.5) than from taking placebo. A comparison of summary estimates of ORs for each drug failed to show differences in efficacy or withdrawal rates among drugs.


Each of 6 antiepileptic drugs is better than placebo in reducing seizures. No difference in efficacy among the drugs is detectable.

Source of funding: Wellcome Trust.

For article reprint: Dr. A.G. Marson, Department of Neurological Science, University of Liverpool, Walton Centre for Neurology and Neurosurgery, Liverpool L9 1AE, England, UK. FAX 44-151-525-3857.

Table. Effect of Newer Drugs vs Placebo on Seizure Frequency*

Drug RBI, % 95% CI ABI, % NNT CI
Gabapentin 121 48 to 232 11 9 7 to 16
Lamotrigine 114 39 to 229 9 11 7 to 25
Tiagabine 239 105 to 460 14 7 6 to 11
Topiramate 284 144 to 504 31 4 3 to 5
Vigabatrin 204 105 to 350 28 4 3 to 5
Zonisamide 120 28 to 278 14 8 5 to 21

*Abbreviations defined in the Glossary; RBI, ABI, NNT, and CI calculated from data in article.


Partial epilepsy is the most common form of epilepsy; roughly half of seizures in adults are complex partial seizures (1). This systematic review by Marson and colleagues compares 6 new anticonvulsant drugs as add-on therapy in partial epilepsy. The authors conclude that no evidence-based rationale exists for choosing one new drug over another. However, none of the source trials did direct drug-to-drug comparisons.

In most patients with epilepsy, seizures occur infrequently, are suppressed with monotherapy, and cease after a few years. In contrast, the patients in these trials had severe epilepsy and treatment with 1 or more drugs had failed.

Reduction of seizure frequency correlates poorly with improved quality of life (2). Other measures are more clinically relevant (3). An accompanying editorial (4) properly notes that these trials of add-on drugs, instead of assessing the efficacy or toxicity of individual drugs, test the overall effect of placebo or a new drug combined with drugs that were not completely effective by themselves. The long-term efficacy and toxicity is unknown because the trials were only 8 to 24 weeks in length. For unclear reasons, many patients eventually abandon lamotrigine or vigabatrin therapy (5).

Future trials should assess whether these drugs are better than the currently recommended choices of carbamazepine or phenytoin (6, 7) for partial seizures or, alternatively, valproate (8) for partial seizures that secondarily generalize to tonic-clonic seizures; and which drug is best as add-on therapy.

Steven Belknap, MD
University of Illinois College of Medicine at PeoriaPeoria, Illinois, USA


1. Cascino GD, Kelly PJ, Hirschorn KA, Marsh WR, Sharbrough FW. Mayo Clin Proc. 1990;65:1053-60.

2. Smith D, Chadwick D, Baker G, Davis G, Dewey M. Epilepsia. 1993;34(Suppl 5): 31-5.

3. O'Donoghue MF, Duncan JS, Sander JW. Epilepsia. 1996;37:563-71.

4. Mignot G. BMJ. 1996;313:1169-74.

5. Walker MC, Li LM, Sander JW. BMJ. 1996 Nov;313:1184-5.

6. Smith DB, Mattson RH, Cramer JA, et al. Epilepsia. 1987;28(Suppl 3):S50-8.

7. Mattson RH, Cramer JA, Collins JF, et al. N Engl J Med. 1985;313:145-51.

8. Mattson RH, Cramer JA, Collins JF. N Engl J Med. 1992;327:765-71.