Which is Better, the Older or the Newer Antiepileptic Drugs?

Abstract & Commentary

By Cynthia L. Harden, MD, Professor of Neurology and Neuroscience, Comprehensive Epilepsy Center, Weill Medical College of Cornell University. Dr. Harden reports no financial relationships relevant to this field of study.

Synopsis: These 2 clinical trials comparing Standard and New Antiepileptic Drugs (SANAD) showed that for partial epilepsy, lamotrigine was clinically better than carbamazepine, and that for generalized and unclassifiable epilepsy, valproate was better tolerated than topiramate and more efficacious than lamotrigine. Due to the potential adverse effects of valproate during pregnancy, the authors were cautious about the interpretation of their results regarding valproate.

Sources: Marson AG, et al; and the SANAD Study Group. The SANAD study of effectiveness of carbamazepine, gabapentin, lamotrigine, oxcarbazepine, or topiramate for treatment of partial epilepsy: an unblinded randomised controlled trial. Lancet. 2007;369:1000-1015. Marson AG, et al; and the SANAD Study Group. The SANAD study of effectiveness of valproate, lamotrigine, or topiramate for generalised and unclassifiable epilepsy: an unblinded randomised controlled trial. Lancet. 2007 ;369:1016-1026.

These 2 large clinical trials carried out in the United Kingdom aim to answer an important question for physicians treating epilepsy, that is, which is better, the older or the newer antiepileptic drugs (AEDs)? For that purpose, in these studies, the "standard" AED deemed to be appropriate therapy for partial epilepsy treatment was carbamazepine, and for generalized and presumably generalized epilepsy (unclassifiable), it was valproate. The investigators evaluated these standard treatments alongside several newer AEDS, including gabapentin, lamotrigine, oxcarbazepine, and topiramate for partial epilepsy, and lamotrigine and topiramate for generalized and unclassifiable epilepsy.

Patients at least 4 years of age and older were included, and had either newly diagnosed epilepsy, failure on AED monotherapy but not with one of the study AEDs, or were previously in remission but relapsed after withdrawal of AED treatment. The choice of AED was centrally randomized but was not blinded for either the physician or the patient. The AED starting dose, titration and dose adjustment after recurring seizures was performed by the treating physician in an open, unblinded manner. The aim of AED treatment was to control seizures using the lowest monotherapy dose.

The primary outcomes were 1) time to treatment failure, an outcome which included stopping the randomized AED due to inadequate seizure control, side effects, or the addition of another AED, which ever came first, and 2) the time from randomization to a one-year period of seizure remission. Secondary outcomes included quality of life and cost-effectiveness of the AED.

For the partial epilepsy portion of SANAD (arm A), 1721 patients were randomized, with the smallest treatment arm being oxcarbazepine since this arm was added later to the ongoing study. For time-to-treatment failure, lamotrigine was significantly better than carbamazepine, gabapentin and topiramate, and was slightly but not significantly better than oxcarbazepine. Lamotrigine provided better seizure control than gabapentin, and had fewer adverse effects than carbamazepine, topiramate or oxcarbazepine. Time to 12-month remission, however, was significantly better for carbamazepine over gabapentin, and carbamazepine had a non-significant advantage for this outcome over lamotrigine, topiramate and oxcarbazepine. Rash was more common with carbamazepine and oxcarbazepine than with lamotrigine. No differences between AEDs were found on the quality-of-life scale; patients with continued seizures had worse quality of life, including mood subscales. Lamotrigine was associated with a trend for reduced risk of depression compared to other AEDs. The lowest-cost-per-quality-adjusted-life year (QALY) was for lamotrigine and carbamazepine.

For the generalized and unclassifiable epilepsy portion of SANAD (arm B), 716 patients were randomized. For time-to-treatment failure, valproate was significantly better than topiramate, mostly accounted for by adverse effects from topiramate. For time to 12- month remission, valproate was significantly better than lamotrigine only. For the subset of patients with clear evidence of idiopathic primary generalized epilepsy, valproate was significantly better than lamotrigine and topiramate for time-to-treatment failure, but was only better than lamotrigine regarding time to 12-month remission. The authors interpret these results as indicating that valproate was better tolerated than topiramate and more efficacious than lamotrigine.


The impact of these large and ambitious studies is undeniable, and for those of us working intimately with epilepsy patients every day, there is a sense of validation in these results; they do reflect what is seen in clinical epilepsy practice. Without declaring AED "winners" or "losers" in these reports, since the response of patients must be considered on an individual basis, results emerge that lamotrigine is a well-tolerated and often effective AED for partial epilepsy, and valproate, despite its adverse metabolic and teratogenic potential, is a very effective AED for generalized epilepsy. Furthermore, this study confirms that persons with epilepsy who have continued seizures have a poor quality of life compared to those whose seizures are controlled.

Regarding the advantages of the "standard" vs the "newer" AEDs, one newer AED, lamotrigine, appears somewhat advantageous over carbamazepine, but there was, overall, little difference in treatment arms in the partial epilepsy study, with the exception that efficacy outcomes were most withering with gabapentin. As a "standard" AED in generalized epilepsy, valproate holds its own against the 2 newer AEDs. This finding, when extrapolated to clinical practice, goes a long way toward explaining why valproate, with its myriad of adverse effects, largely due to enzyme-related metabolic alterations, and its teratogenic risk, remains a widely prescribed AED even in the pediatric and adolescent populations where generalized epilepsy commonly occurs.

Drawbacks of these studies are obvious; they are unblinded and dosing and titration was not controlled. This can lead to bias in both directions: too cautious a titration can be associated with lack of efficacy and too high a starting dose or rapid titration can lead to adverse effects. Each AED has its own profile of risk regarding dose, titration, efficacy and adverse effects; some are more prone to one risk than another. Notably, the doses achieved in this study for lamotrigine were relatively low with a mean dose of 249 mg per day for long-term stable partial epilepsy patients and 203 mg per day for long-term stable generalized epilepsy patients. Doses reported for other AEDs seem very reasonable, with relatively small standard deviations, suggesting that dose and dose ranges in these studies are similar to standard practice in the United States. Also, it now seems like the lack of such information with levetiracetam is a gaping hole in our assessment of "standard" vs "newer" AEDs due to levetiracetam's wide use and broad-spectrum of indications.

The "real world" methodology of this study, where each physician starts and titrates the AED with a large degree of independence within the clinical trial, is in some way reassuring and suggests that these results can be readily extrapolated to clinical practice. Keeping in mind the lack of strict control on dosing in these studies, the results are a stronger endorsement of valproate as an effective and well-tolerated drug for generalized epilepsy than they are discouraging for carbamazepine or the other newer AEDs evaluated as effective treatments for partial epilepsy. Therefore the "standard" AEDs do not fail when compared to "newer" AEDs; however, lamotrigine has advantages for partial epilepsy. The use of valproate in women of child-bearing potential remains a difficult situation. The SANAD results support the effectiveness of valproate for generalized epilepsy, but the teratogenic and neurocognitive risk to the offspring with in utero valproate exposure are also becoming clear. The best choice of AED for women with generalized epilepsy remains a severe clinical challenge, not yet fully informed by known rates of teratogenic risk of valproate alternatives.