Choice of Antiepileptic Drugs

Elinor Ben-Menachem

Jacqueline A. French

Introduction

When the diagnosis of epilepsy has been made and extraneous factors that can provoke seizures are eliminated, then there is usually the decision to treat the patient with an antiepileptic drug (AED). The patient (or parents, when it concerns a child) must be given as complete information as possible about the diagnosis and prognosis and the different treatment alternatives. Without this understanding between the patient and doctor the risk for noncompliance is significant. The goal of treatment is to achieve seizure freedom without side effects, especially those that cause more negative consequences than the seizures themselves and reduce the quality of life.

With the advent of the new era of antiepileptic drugs, heralded by the approval of ten new compounds, all developed in the last two decades, selection of the most optimal therapy is more complex than ever. It often difficult to predict which drug will be the best tolerated and most likely to produce the best seizure control in a given individual. However, it is perhaps possible to narrow the choices among the available drugs to a few optimal selections based on characteristics that are known about each patient when he or she presents. Some patient characteristics can be seen to fall on a continuum, such as that from the newly diagnosed patient, who has a high likelihood of responding to antiepileptic drugs, to the drug-resistant patient, who is unlikely to remit. Another such continuum is that of age, from infancy, through adulthood, and finally advanced age. Different drug choices might be needed at different points on these continua. In addition, within the epilepsy population there are other characteristics that divide patients into subpopulations, including gender, epilepsy syndrome, and comorbidities.

Thus, the choice of AED is based on not only which drug has been shown to be effective for a particular seizure type, but also the patient’s age and gender, comorbidities, and whether he or she is newly diagnosed or refractory. Much is known about likelihood of seizure control, drug tolerability, and side effect profile that is specific to each of these characteristics. These characteristics can also inform choices based on potential interactions and even cost. For example, a new AED that is costly but has been shown to be potent as add-on therapy in refractory epilepsy may be warranted in a difficult-to-control patient but not in a patient who is newly diagnosed.

Stages of Disease

Initial Therapy

Studies in patients with newly diagnosed epilepsy have suggested that approximately 40% of patients with newly diagnosed epilepsy will be rendered seizure free on the first therapy tried.⁵ Once seizure free, a patient might be unwilling to attempt a change in therapy, even in the presence of side effects. Patient and physician may not choose to risk a medication conversion due to the possibility of destabilization. The stakes become even higher when the patient has returned to previous activities, including driving. The result is that many patients remain on their initial therapy for years, if not a lifetime. For this reason, the choice of initial therapy is a critical juncture in the care of epilepsy.

Efficacy

Although it would be ideal to have efficacy data for all drugs specifically obtained from well-designed trials in the newly diagnosed population, unfortunately there is usually a lag between the time that a new AED is approved and the availability of data on the newly diagnosed. It is unclear whether trials need to be performed prior to using the new drugs in this population. Although it is likely that drugs that work as add-on therapy also work in monotherapy and that drugs that work in refractory patients are also effective in newly diagnosed patients, who are typically considered to be easier to control, it is also reasonable to assume that the pathophysiology underlying refractory epilepsy differs from that of easy-to-control epilepsy, and therefore drugs might be more suitable for one population or the other. For example, vigabatrin was highly efficacious in placebo-controlled add-on trials in refractory patients,¹⁸ whereas it was less effective (but better tolerated) than carbamazepine in a trial of newly diagnosed patients.⁸ It does not appear that relative drug efficacy in add-on studies is predictive of better effect in newly diagnosed patients. Despite great hopes, no new AED has been proven to be more efficacious than a “standard” AED in the newly diagnosed population. Trials in newly diagnosed patients have been successfully completed for vigabatrin, lamotrigine, oxcarbazepine, gabapentin, and topiramate, and soon will be for evetiracetam.¹^,³^,⁴^,⁵^,⁸^,⁹^,¹¹^,¹⁴^,²²^,³¹^,³⁷

Safety and Tolerability

The absence of demonstrated efficacy differences between AEDs when tested in newly diagnosed populations puts emphasis on other drug characteristics as a reasonable basis for AED selection in this population. One important issue is drug safety and tolerability. In the presence of an abundance of options, drugs with few or no safety or side effect concerns should be high on the list. The older AEDs, although generally safe, all carry a small risk of potentially life-threatening adverse events such as hepatic failure and aplastic anemia. A number of the newer antiepileptic drugs, although they have not been used for the same duration, or in as many patients as the older drugs, appear to be safer. As experience and duration of follow-up grows, we can hope that this apparent safety will be confirmed. Even now, the newest antiepileptic drug, pregabalin, has reportedly been used in well more than
100,000 unique patients (Pfizer Inc., personal communication). Several older AEDs have also been associated with risk of long-term side effects, such as gum hyperplasia, neuropathy, connective tissue disorders, and osteopenia.²⁹ Of the newer drugs, vigabatrin has been associated with permanent visual field defects,²⁵ and topiramate and zonisamide carry a risk of renal nephrolithiasis.³⁶ So far, no other long-term side effects have been identified for the new AEDs, although it is probably naïve and definitely premature to hypothesize that none exist. It is important to maintain vigilance for such potential effects because they take unexpected forms and may not be identified before hundreds of thousands of patients have been exposed. On the other hand, in the head-to-head comparison trials the new AEDs were almost always as well or better tolerated than the older drugs. Examples are the comparative trials between lamotrigine and carbamazepine⁵; oxcarbazepine and carbamazepine,¹⁴ phenytoin,³^,²² and valproate¹¹; topiramate and carbamazepine and valproate³¹; and gabapentin and carbamazepine.⁹ Of note, the studies with carbamazepine as a comparator used the immediate-release rather than the currently favored extended-release formulation, and titration rates were not always optimized. Nevertheless, the new drugs probably have better tolerability especially when given in appropriate doses after careful titration. These outcomes might indicate that these drugs should be considered as initial monotherapy in newly diagnosed patients.

Initiation/Dose

Newly diagnosed patients remain at risk for additional seizures until the chosen drug reaches a therapeutic dose. This gives an advantage to drugs that can be initiated at a therapeutic dose or titrated rapidly. Some AEDs, such as lamotrigine, topiramate, and zonisamide, typically require prolonged titration, and this can limit their use in certain patients. Carbamazepine can typically be titrated to an effective dose within 1 to 2 weeks. Other drugs, such as oxcarbazepine, gabapentin, and pregabalin, have been initiated at a therapeutic dose in clinical trials without safety concerns but have been demonstrated to be better tolerated when titrated slowly.²^,¹⁶^,³⁴ Titration schedules are outlined in Table 1.

Table 1 Initiation schedule for common antiepileptic drugs

Drug	Initiation
Phenytoin	Intravenous/oral load
Carbamazepine	Gradual initiation/1–2 wk
Valproate	Immediate therapeutic dose
Topiramate	Gradual initiation/4–6 wk
Gabapentin	Immediate therapeutic dose
Lamotrigine	Gradual initiation/4–6 wk
Levetiracetam	Immediate therapeutic dose
Zonisamide	Gradual initiation/4–6 wk
Oxcabazepine	Gradual initiation/4–6 wk

It may be difficult to choose a dose to titrate to as initial therapy in a newly diagnosed patient. Table 2 shows the doses that were used in randomized, placebo-controlled trials in newly diagnosed patients. In trials of lamotrigine and oxcarbazepine, variable dosing was permitted in the initial portion of the trial.⁴^,⁵^,¹¹^,¹⁴^,²² In these cases, it is likely that the average dose used in the trial is a reasonable starting point for newly diagnosed patients. In trials of topiramate and gabapentin, fixed doses were employed. Topiramate 100 mg was better tolerated, and equally as effective, as 200 mg, and gabapentin 900 mg was more effective and better tolerated than 300 or 1,800 mg. However, other doses were not tested.³¹

Table 2 Doses used in randomized, placebo-controlled trials of four new antiepileptic drugs in newly diagnosed patients

Drug	Dose
Gabapentin	Fixed, 300, 900, 1,800 mg⁵
	(900 mg superior)
Lamotrigine	∼150 mg¹⁸
Oxcarbazepine	∼1,000 mg (adults)⁸^,³⁷
	∼18 mg/kg (children)³
Topiramate	Fixed, 100 or 200 mg³¹