Back to Basics: The Right Anti-Epileptic Drug for the Right Diagnosis

Abstract & Commentary

Source: Benbadis SR, et al. Idiopathic generalized epilepsy and choice of antiepileptic drugs. Neurology. 2003;61: 1793-1795.

Benbadis and associates reviewed the medication history of patients with EEG-confirmed idiopathic generalized epilepsy (IGE). They classified patients according to the antiepileptic drugs (AEDs) used at the time of referral to their epilepsy center. They defined "adequate" or "broad spectrum" AEDs as valproic acid (VPA), ethosuximide, lamotrigine (LTG), topiramate (TPM), levetiracetam (LEV), and zonisamide. "Inadequate" AEDs for IGE included phenytoin (PHT), carbamazepine (CBZ), oxcarbazepine, gabapentin, and tiagabine. Of 58 patients, 17 (29%) were on adequate AEDs, 28 (48%) were on inadequate AEDs, and 13 (22%) were on both.


The fact that so many IGE patients (70%) were prescribed "ill-advised" AEDs (alone or in polypharmacy) for IGE at initial referral to a tertiary epilepsy center can be attributed to misdiagnosis of IGE or inappropriate AED choice despite an accurate diagnosis of IGE. In the patients in this study, Benbadis et al do not reveal how many patients had been misdiagnosed as having focal epilepsy; the criterion for referral was medically resistant epilepsy not necessarily intractable IGE (only 7 patients studied). The experience of our own epilepsy center indicates that, among patients referred specifically for presurgical evaluation of their epilepsy, approximately 10-15% are immediately ineligible because they in fact have generalized epilepsy rather than focal epilepsy.

While the issue of misdiagnosis is an important one, it can be relatively easily overcome with an open mind and appropriately scrupulous diagnostic evaluation. Inadequate knowledge of AED indications (FDA approved or "off-label") is a more difficult educational task. The problem can be divided into over-reliance and inappropriate use of older AEDs and/or lack of sufficient data to support use of newer AEDs for IGE.

We have personally been asked by a general neurologist whether we accepted the "brainwashing" by J. Kiffin Penry and others that VPA is really the agent of choice for absence and juvenile myoclonic epilepsy (JME). In that practitioner’s experience, PHT and CBZ were equally effective as VPA in treating absence and JME. His experience flew in the face of well-established data that absences are worsened by CBZ in both children and adults. The clinical experience continues to mount with reports that CBZ and vigabatrin administration produced de novo absences in 3 children who only had clinical or electrographic evidence of partial epilepsy.1 The general neurologist is, unfortunately, more comfortable with PHT, CBZ, and even phenobarbital than VPA, which helps to explain why 4 of 58 patients (7%) with IGE had never received a trial of VPA.

The flip side of the AED choice problem is that general neurologists are now presented with many AED choices (8 novel agents in the last decade). Overwhelmed by the newer AEDs and in the absence of strong evidence-based medicine, they are choosing to resort to the older agents vs nearly random usage of the newer drugs. The concept of a broad-spectrum AED (ie, one effective for both focal and generalized epilepsies) is one that has only arisen in the last 10 years. Few practitioners spoke of broad-spectrum AEDs when there was only one member of the class (ie, VPA).

We do not believe that there is a firm consensus on the drugs listed as adequate or broad spectrum. For example, there have been no published trials (controlled or not, blinded or not) to establish the efficacy of LEV for generalized epilepsy. Of the 6 references cited by Benbadis et al to support the use of LEV in IGE, one appeared in abstract form only and 3 were not peer-reviewed. As a result, at this time, we would not describe LEV as an adequate, broad-spectrum agent for IGE. By contrast, a paper by Prasad and associates2 provides useful data about the use of LTG and TPM in treating JME. Prasad et al also point out that while PHT and CBZ can reduce generalized tonic-clonic seizures (GTCS) associated with JME, these drugs are less useful for myoclonic seizures. This raises a further issue about the end points of AED therapy. Recently, a patient presented to us with a clear electroclinical history of JME. His previous neurologist felt that he had been adequately controlled for years on PHT since it had abolished his GTCS. The patient, though, had a further improvement in quality of life when he was changed to VPA monotherapy at our center. He remained free of GTCS and also began to enjoy tennis and golf, which he had never been able to play previously because of his constant jerking.

The answer to better pharmacologic seizure control in IGE involves no more than bread-and-butter medical practice: get the diagnosis right, know your drugs, and, most important, listen to the patient. — Andy Dean

Dr. Dean, Assistant Professor of Neurology and Neuroscience; Director of the Epilepsy Monitoring Unit, Department of Neurology, New York Presbyterian Hospital Cornell Campus, is Assistant Editor of Neurology Alert.


1. Yang MT, et al. Brain Dev. 2003;25:51-56.

2. Prasad A, et al. Arch Neurol. 2003;60:1100-1105.