"Old" Versus "New" Antiepileptic Drugs

November 1, 2005

epilepsy, seizures

The new AEDs (Table 1) complement the conventional, or "old," AEDs (Table 2) that are well known to most physicians. In addition to providing more options for the adjunctive treatment of seizures, some of these new AEDs have been approved for monotherapy. Three of them have also received FDA indications for treatment of seizures associated with Lennox-Gastaut syndrome, providing therapeutic alternatives for difficult-to-treat patients. Pediatric-friendly forms of some new AEDs, such as a sprinkle form of topiramate (Topamax Sprinkles, Ortho-McNeil) and oral solutions of gabapentin (Neurontin, Pfizer), levetiracetam (Keppra, UCB Pharma), and oxcarbazepine (Trileptal, Novartis) also are available.

One of the advantages of using older AEDs is that their performance qualities, side-effect profiles, drug-drug interactions, and idiosyncrasies (such as lack of linear kinetics for phenytoin), are relatively well known. Although side effects of the old AEDs may be undesirable, they are more or less predictable, and physicians and patients face few surprises.

In addition, advances associated with the old AEDs make these drugs more user-friendly. These advances include an extended- release form of divalproex sodium (Depakote ER, Abbott) and a sprinkle form (Depakote Sprinkle Capsules, Abbott), 2 extended- release preparations of carbamazepine (Tegretol XR, Novartis, and Carbatrol, Shire), and intravenous forms of phenytoin (Cerebyx, Esai) and valproate (Depacon, Abbott). Generic forms of the commonly used older AEDs are available; their use may reduce drug purchase costs (whether generic AEDs are therapeutically equivalent is not discussed in this article).1

Idiosyncratic Adverse evEnts: A Numbers Game

The new AEDs are, well, new, and many physicians are unfamiliar with dosing, drug-drug interactions, and expected side effects. In addition, physicians need to be on the lookout for adverse events not described in the initial package insert. The FDA collects postmarketing reports of adverse events through Medwatch, a passive surveillance system.2 One of the reasons for the appearance of postapproval adverse events is the relatively small number of patients tested before FDA approval. When an AED goes through FDA- required clinical trials, the total number of patient exposures may range from 3500 to 10,000 or less, according to Jacqueline French, MD, professor of neurology and assistant dean of Clinical Research at the University of Pennsylvania, Philadelphia, and lead author of the epilepsy guidelines of the American Academy of Neurology (AAN). What if the incidence of an idiosyncratic adverse event is only 1 in 100,000? It will probably be missed until the AED is more broadly distributed.

This problem, which applies to all classes of drugs, was summarized in a letter written by Paul Leber, MD, director of the Division of Neuropharmacological Drug Products of the FDA, regarding the approval of one of the new AEDs3:

Finally, I note for the record that a regulatory conclusion that [this drug] has been shown to be "safe for use" is not a warrant that [this drug] is risk free. Rather it is an opinion, based largely on sentiment, that the benefits of [this drug's] use are sufficient to justify its use in the fact of the risks of use so far identified. Thus, the conclusion offered is a conditional one that may well change if there are serious risks associated with [this drug's] use that occur at an incidence too low to be detected in a drug development cohort comprised of 1700 or so subjects.

The best-known example of this phenomenon in the epilepsy community is felbamate (Felbatol, MedPointe), the first of the new AEDs to receive FDA approval in 1993 for add-on treatment of and monotherapy for partial and generalized seizures.4 After only 1 year of use, during which time the drug demonstrated marked effectiveness for many patients with epilepsy, reports of aplastic anemia and hepatic failure resulted in an FDA warning and calls from the company to discontinue use of felbamate in most patients.

The number of patients taking felbamate dropped from about 150,000 to less than 15,000.4 Subsequently, the risk of aplastic anemia induced by felbamate has been estimated to be 1 to 3 patients in 10,000, while the risk of aplastic anemia in the general population is only 2 per million.4 In addition, the risk of hepatic failure caused by felbamate is 1 in 18,000 to 25,000 cases.4 One can see why these idiosyncratic adverse events were not detected before FDA approval; the number of patients who had tried felbamate before its launch was only about 4000.4

There are several other examples in which idiosyncratic adverse events caused by new AEDs have been discovered or have been recognized as important only after approval. For example, although topiramate received FDA approval in 1996, it was not until 2001 that reported cases of topiramate-related acute myopia and secondary angle-closure glaucoma prompted changes in the package insert and an order from the FDA to issue a "Dear Health Care Professional Letter."5 Given that ocular side effects of topiramate are rare, it is not surprising that preapproval testing did not uncover them.

Other discoveries include the association between hyperchloremic metabolic acidosis6 and hypohidrosis and topiramate use7 and between hypohidrosis and zonisamide (Zonegran, Eisai) use.8 In addition, seizures have been reported in nonepileptic patients receiving tiagabine (Gabitril, Cephalon) for off-label use.9

Postmarketing discoveries of adverse events occur with other classes of drugs as well as the AEDs. The risks of heart attack and stroke related to the cyclooxygenase-2 inhibitors celecoxib (Celebrex, Pfizer), rofecoxib (Vioxx, Merck), and valdecoxib (Bextra, Pfizer) have been widely publicized.10 Other instances of postmarketing adverse events are less well known. For example, galantamine (Razadyne, Ortho-McNeil), a cholinesterase inhibitor approved for the treatment of mild to moderate Alzheimer disease, had a change in its labeling because of postmarketing discoveries of excess deaths during clinical trials for another indication, treatment of mild cognitive impairment. The change in the prescribing label was accompanied by a name change; what was Reminyl is now Razadyne.11

Impact on Patient Care

These postmarketing findings affect patient care. To use topiramate as an example, it should now be standard practice for physicians to query patients beginning a course of topiramate about visual changes and eye pain because of concerns about acute glaucoma. Particular attention needs to be paid to patients with cognitive impairment who may not be able to report their ocular symptoms. In addition, because of the possibility of metabolic acidosis, serum bicarbonate levels should be checked before and during treatment. Patients taking topiramate also should be monitored for decreased sweating and increased body temperature, particularly in hot weather and when taking other drugs with carbonic anhydrase inhibitor or anticholinergic activity.12

Similar issues have become apparent with the introduction of the "atypical" antipsychotics, which have been widely embraced as superior to traditional antipsychotic medications. However, as evidence of the "metabolic syndrome" has emerged and questions regarding superiority of efficacy have been posed, the advantage of these drugs has become less clear.13,14

AAN AED Guidelines

To help physicians respond to the advantages and disadvantages of new therapeutic options for the treatment of epilepsy, the AAN assembled an expert panel and published 2 sets of evidence-based guidelines regarding the efficacy and tolerability of the new AEDs.15,16 The first focuses on treatment of patients with new-onset seizures,15 and the second examines the data supporting treatment of patients with refractory epilepsy.16 These guidelines summarize the evidence from both preapproval and postmarketing clinical trials.

The AAN distributed the guidelines to 18,500 members in a special mailing and published them in Neurology, which has a circulation of 18,807. In addition, the guidelines have been downloaded from the AAN Web site more than 18,200 times, clinician summaries of the guidelines have been downloaded more than 6875 times, and PowerPoint presentations have been downloaded more than 7430 times. An AAN survey of 375 members (response rate, 43%) determined that 75.6% had read and incorporated some or all of the recommendations from Part I and 72.3% had read and incorporated some or all of the recommendations from Part II into their practices. In a separate survey, 88.4% of AAN members felt comfortable with their ability to determine appropriate drug treatment for epilepsy (personal communication, Becky Schierman, MPH, Guideline Dissemination Project Manager, AAN).

AEDs in the "Real World"

Applied Neurology interviewed a leading epileptologist and 2 clinicians to see how they approach the prescribing of new AEDs in their daily practices. Here are a few select comments:

Jacqueline French, MD, introduced earlier in this feature, observed, "There is a difference in risk/benefit, depending on the patient's situation. If a patient has refractory disease and has failed therapy with all other agents, it is appropriate to try a new drug early after its approval. Whether it should be tried in patients with less severe disease is a more complex question. It still may be appropriate to try it if the patient has special circumstances and the properties of the 'new' drug seem to offer a particular benefit--for example, lack of a particular side effect or drug interaction.

"In these cases, the physician should weigh the potential risks versus the apparent benefit. Ideally, as illustrated in the case of the antipsychotics in the recent New England Journal of Medicine article, a study will be done at some point that compares the risks and benefits of the new medication with other choices, to determine in an evidence-based fashion what the trade-offs are, if any."

With respect to comparative studies, French commented, "It is not necessarily the responsibility of the FDA to perform head-to-head trials of these medications. The role of the FDA is to prove that medications are safe and effective, not necessarily to prove their relative safety and efficacy compared to other options. Often, when medications are compared head to head, there are no efficacy differences, and any advantage of newer medications is realized by lack of drug interactions, improved side effects, [and other ways]. Head-to-head trials may need to be done by other entities, such as the National Institutes of Health. In epilepsy, these trials are often done by pharmaceutical companies, but this is the least preferable option." (A recent study compared carbamazepine with gabapentin and lamotrigine [Lamictal, GlaxoSmithKline] in the elderly.17)

Neil Busis, MD, a neurologist in private practice and clinical associate professor, Department of Neurology, University of Pittsburgh School of Medicine, considers himself an "early adopter" of new technology and medicines. "I want my patients to get the latest and greatest therapies as soon as possible. I assumed that the FDA approval process was bulletproof, so the risks were minimal. This stance was shaken a bit by Felbatol, but was really shaken by the Tysabri [natalizumab, Biogen Idec] debacle. I will be an early adopter for some new drugs, such as Lyrica [pregabalin, Pfizer], since it has been around in Europe for a long time and is related to a drug we all know and love, Neurontin. On the other hand, with brand-new drugs, especially those follow- ing new paradigms--Tysabri is a monoclonal antibody, for example--I am going to recommend that my patients wait at least 6 months to allow postmarketing surveillance results to become known and disseminated.

"We are fooling ourselves if we think all new drugs are safer than the old ones. We are also fooling our-selves if we think that all new drugs should replace older more established treatments for that condition."

Busis endorsed the AAN guidelines: "I am a big fan of AAN practice parameters. They set the bar very high--their evidentiary rules are more stringent than those of most other professional organizations, so their conclusions are quite compelling. However, there is a downside: it takes a long time for these guidelines to be developed and disseminated. The AED guideline, for example, only surveys literature to 2003. Since then, new publications may have suggested additional appropriate indications for some drugs. But I go with the guidelines, since that is what is published by my Academy."

Randy Kozel, MD, a neurologist in private practice in Newport, RI, revealed his approach to patients with seizures: "I don't really have a formula. I have to review what might best fit that particular individual. All of the AEDs are pretty much the same in efficacy, but how effective a particular drug is going to be for a particular patient has to be evaluated in terms of side-effect issues such as bone density, cognitive issues, pregnancy, and gender issues. It is somewhat easier now because all of the older drugs that have enzyme-inducing properties are less likely to be drugs of choice. Now with the newer ones, it's a matter of finding one that fits their pocketbook, their clinical situation, and their lifestyle."

In addition to randomized, double-blind clinical trials necessary for the approval of AEDs, more comparative studies of old versus new and new versus new AEDs are needed. Physicians prescribing new drugs can avail themselves of published guidelines for prescribing and can remain alert to the possibility of previously undescribed adverse events. Formalized medical education dedicated to periodic review of drug safety could be considered. *


1. Wilner AN. Therapeutic equivalency of generic antiepileptic drugs: results of a survey. Epilepsy Behav. 2004;5:995-998.

2. Medwatch. The FDA Safety Information and Adverse Event Reporting Program. Available at: http://www.fda.gov/medwatch/safety.htm. Accessed October 20, 2005.

3. Leber P. Memorandum. Available at: www.


correspondence.pdf. Accessed October 20, 2005.

4. Pellock JM. Felbamate. Epilepsia. 1999;40(suppl 5):S57-S62.

5. Temple R. New drug application approval letter for topiramate (NDA 20-505/S-002 and NDS 20-884/S-101). Available at: www.fda.gov/cder/ foi/appletter/2001/20505S002ltr.pdf. Accessed October 20, 2005.

6. Warning on metabolic acidosis with Topamax. FDA Patient Safety News: Show #25, March 2004. Available at: www.accessdata.fda.gov/psn/ printer.cfm?id=205. Accessed October 20, 2005.

7. Galicia SC, Lewis SL, Metman LV. Severe topiramate-associated hyperthermia resulting in persistent neurological dysfunction. Clin Neuropharmacol. 2005;28:94-95.

8. Knudsen JF, Thambi LR, Kapcala LP, Racoosin

JA. Oligohydrosis and fever in pediatric patients treated with zonisamide. Pediatr Neurol. 2003;28: 184-189.

9. Seizures in patients without epilepsy. FDA Alert (February 28, 2005). Available at: www. fda.gov/cder/drug/infopage/tiagabine/default.htm. Accessed October 20, 2005.

10. The experts' verdict on painkillers [editorial]. New York Times. February 19, 2005:A14.

11. Hand L. A new warning and a name change. App Neurol. 2005;1(5):9.

12. Topamax. In: Physicians' Desk Reference. 59th ed. Montvale, NJ: Thomson PDR; 2005:2541-2549.

13. Freedman R. The choice of antipsychotic drugs for schizophrenia. N Engl J Med. 2005;353: 1286-1288.

14. Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med. 2005; 353:1209-1223.

15. French JA, Kanner AM, Bautista J, et al. Efficacy and tolerability of the new antiepileptic drugs, I: treatment of new onset epilepsy: report of the Therapeutics and Technology Assessment Subcommittee and Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology. 2004;62:1252-1260.

16. French JA, Kanner AM, Bautista J, et al. Efficacy and tolerability of the new antiepileptic drugs, II: treatment of refractory epilepsy: report of the Therapeutics and Technology Assessment Subcommittee and Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology. 2004;62:1261-1273.

17. Rowan AJ, Ramsay RE, Collins JF, et al. New onset geriatric epilepsy: a randomized study of gabapentin, lamotrigine, and carbamazepine. Neurology. 2005;64:1868-1873.