Abstract
Commentary
With the exception of oxcarbazepine and felbamate, all of the new antiepileptic drugs were approved initially as adjunctive therapy for partial epilepsy; approval was based on placebo-controlled, add-on trials involving patients with refractory epilepsy. In these trials, the main outcome measures were improvement in seizure frequency over baseline and the proportion of patients with 50% or greater reduction in seizure frequency. Superiority over placebo in add-on trials does not necessarily predict that an antiepileptic drug will be effective and well tolerated as an initial monotherapy. Confirmation for use of an antiepileptic drug as first-line treatment requires a sound monotherapy trial with newly diagnosed patients. In addition, the practicing physician would need to feel assured that the new antiepileptic drug is not less effective than established, standard therapy. Large, comparative antiepileptic drug trials are necessary to provide that assurance. The two large, cooperative VA comparative trials have played a major role in developing guidelines for the older antiepileptic drugs and propelled carbamazepine to the position of being the favored initial agent for the treatment of partial epilepsy (1,2). Consequently, carbamazepine has become the customary active control used for comparative, first-line, monotherapy trials of lamotrigine, oxcarbazepine, and gabapentin (36).
Levetiracetam is one of the most widely used add-on antiepileptic medications. Some of its advantages include rapid and almost complete absorption, initiation at an effective dose, absence of hepatic metabolism, absence of enzyme induction, absence of clinically significant interactions, and the availability of an intravenous formulation (7,8). Several published open-label reports of successful initial monotherapy administration of levetiracetam begged for a formal, blinded, and randomized trial to support this practice. The current study by Brodie et al. addresses this issue, is well powered, and also distinguishes itself from previous comparative trials by using a controlledrelease preparation of carbamazepine as well as flexible dosing. Controlled-release carbamazepine is better tolerated owing to less fluctuation in plasma levels. Levetiracetam and controlled-release carbamazepine were equally effective with respect to seizure freedom at 6 months and 1 year of treatment. With both antiepileptic drugs, most patients became seizure free at the lowest dose level. Overall, more patients discontinued therapy because of adverse experiences in the carbamazepine group, but the difference did not reach significance, and the two drugs showed a similar proportion of patients who had at least one adverse experience. The levetiracetam-treated group more often reported depression and insomnia, while the carbamazepine group more often reported back pain, which is hard to explain. There also was greater weight gain with carbamazepine than levetiracetam.
The results of this trial earned levetiracetam approval as a first-line monotherapy in the treatment of partial epilepsy in the European Union, but not in the United States. Approval of antiepileptic drugs by the U. S. Food and Drug Administration (FDA) requires demonstration of superiority over another treatment or over placebo and may not be based on equivalent efficacy. In adjunctive trials, demonstration of superiority is straightforward, with placebo used as a comparator, and since baseline antiepileptic medications are continued, there are no ethical issues involving the use of placebo. In contrast, the use of placebo as monotherapy for epilepsy does raise ethical concerns, and superior efficacy has never been demonstrated for a new antiepileptic drug in comparison with the old antiepileptic drugs. As a result of these difficulties, few drugs have earned initial monotherapy indication in the United States. One concern raised by the FDA in relation to equivalence trials is that it is possible that in a specific population studied, equivalence could be due to equal lack of efficacy (9). However, the proportion of seizure-free patients should help evaluate this possibility. One epidemiologic study suggested that the risk of seizure recurrence at 1 year is 57%, after two unprovoked seizures, and 61%, following three unprovoked seizures (10). Most patients in this epidemiologic study received treatment, and the chance of remaining seizure-free for 1 year without treatment must be less than 40%. The 1-year seizure-free rate of 56.6% to 58.5% in the study by Brodie and colleagues is significantly better than expected (p = 0.001) and represents evidence that both antiepileptic drugs were effective.
The Therapeutics and Technology Assessment Subcommittee and the Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society accepted large, blinded, and randomized comparative studies in its assessment of the new antiepileptic drugs (11). Its 2004 recommendation was that patients with recently diagnosed partial epilepsy who require treatment can be initiated on lamotrigine, gabapentin, oxcarbazepine, or topiramate (11). The current trial likely will add levetiracetam to that list. As a result of the growing list of antiepileptic drugs that can now be considered for first-line treatment, there is a need to develop guidelines. Such guidelines ideally should be determined by comparative trials that evaluate relative efficacy and tolerability of the new antiepileptic drugs. However, choice of first-line antiepileptic drugs also should take into consideration factors such as the acuteness of the epilepsy, the need for rapid titration, the need to avoid interactions, as well as associated comorbid conditions. In the future, an important criterion in the selection of the first antiepileptic drug for a patient may be its antiepileptogenic potential, which is the ability of a drug to arrest, delay, or reverse the development of epilepsy. Some antiepileptic drugs, such as valproate and levetiracetam, demonstrated an antiepileptogenic action in animal models of epilepsy, particularly suppressing the development of kindling. However, valproate failed to demonstrate an antiepileptogenic effect in patients with head trauma or brain tumors (12), and no drug has been demonstrated to have an antiepileptogenic effect in human epilepsy. In recent onset epilepsy, it is possible that epileptogenesis could still be active. An antiepileptogenic drug could potentially influence the course of the epilepsy, perhaps to the degree that seizures do not recur after antiepileptic drug withdrawal. If seizure-free patients in the trial of Brodie and colleagues continue to be followed after drug discontinuation, this trial may help provide data on whether levetiracetam has antiepileptogenic effects in recently diagnosed partial epilepsy. If treatment with levetiracetam is associated with less seizure recurrence on discontinuation, this finding could be evidence of an antiepileptogenic effect and potentially be a very important factor in choosing the first treatment for epilepsy.