Anti-Epileptic Drug Selection

By | August 10, 2013

Selection of anti-epileptic drugs can be very confusing. Each anti-epileptic drug has unique characteristics, including spectrum of activity, cost, pharmacokinetic and pharmacodynamic properties, likelihood for dose-related side-effects and risk of serious health risks. Therefore, to the frustration of many prescribers, there is not a ‘first choice’ selection in specific treatment situations such as initiation in newly diagnosed patients, first add-on, women anticipating pregnancy and so on. For each of these scenarios, there may be more optimal and less optimal choices, but the final selection will depend on many patient characteristics. Table Suggestions for initiations of anti-epileptic drugs provides pragmatic information about initiating anti-epileptic drugs. What follows is a rational sequence of issues that can be used to narrow down drug choices.

Table Suggestions for initiations of anti-epileptic drugs () *

Anti-epileptic drug Suggested titration Can drug be loaded? How is initial target dose usually selected? Typical initial daily target dose
Carbamazepine Add 200 mg every 2-3 days No By attaining minimal serum concentration (4-6μg / ml) 400-800 mg
Ethosuximide 500 mg / week No By attaining minimal serum concentration (40μg / ml) 500-1000 mg
Felbamate 300-600 mg / week No By dose 1200-1800 mg
Gabapentin 300-600 mg / week No By dose 900-1800 mg
Lamotrigine 25 mg; initial monotherapy: 25 mg / day for 2 weeks, then 50 mg / day for 2 weeks, followed by 50- mg increases every week No By dose, and by attaining minimal serum concentration (4μg / ml) 100-200 mg
Levetiracetam 250-500 mg; increase by 250-500 mg / week Yes, intravenous By dose 500-1500 mg
Oxcarbazepine 300-600 mg / week No By dose 900-1800 mg
Phenobarbital 30 mg / week Yes, intravenous in emergency By attaining minimal serum concentration (15μg / ml) 60-120 mg
Phenytoin No Yes, 13-15 mg / kg By attaining minimal serum concentration (10μg / ml) 3-5 mg / kg
Primidone 125-250 mg / week No By attaining minimal serum concentration (8μg / ml) 500-1000 mg
Pregabalin 75-150 mg; increase 75-150 mg / week No By dose 150-300 mg
Tiagabine 4 mg / week No By dose 16-36 mg
Topi ram ate 25-50 mg / week No By dose 100-200 mg
Valproate, valproic acid, divalproex sodium 250-500 mg / week, or 10-15 mg / kg orally daily, increasing by 5-10 mg / kg every week Yes, intravenous By attaining minimal serum concentration (40-50μg / ml) 750-1500 mg
Vigabatrin 500 mg / week No By dose 1000-2000 mg
Zonisamide 50 mg / week No By dose 100-200 mg

*This table presumes that the patient is not on a hepatic enzyme inducer such as phenytoin, phenobarbital or carbamazepine, or an enzyme inhibitor such as valproate.

Epilepsy Syndrome

Classification of seizure type and epilepsy syndrome will be a critical issue in selecting a new anti-epileptic drug, and should be done as a first step in selection. As noted above, some anti-epileptic drugs are considered ‘narrow spectrum’ and do not treat all seizure types. In addition, they may worsen some seizure types. For example, carbamazepine, oxcarbazepine, gabapentin, tiagabine, pregabalin and vigabatrin are most appropriate for seizures associated with partial epilepsy, and may worsen generalized seizure types such as myoclonus and absence. Felbamate, lamotrigine, topiramate, zonisamide and levetiracetam and valproic acid are considered to be broad spectrum. They may be appropriate for most seizure types. However, as noted above, in a recent study of patients with generalized or unclassified seizures randomized to valproate, lamotrigine or topiramate, lamotrigine was found to be inferior to valproate in efficacy, leading the authors to speculate as to how broad spectrum it actually was. Ethosuximide has the narrowest spectrum, and is only effective in treating absence seizures.

Often, seizures cannot be definitively classified at the time of diagnosis. In the randomized study described above, 27% of patients who were entered into the study were described as unclassified. If a patient cannot be classified, it is wise to choose a broad-spectrum anti-epileptic drug.


It is very important to keep cost in mind, particularly for patients who have limited or no coverage for drug expenditure. Such patients may ask about use of generic substitution. These patients can be told that at present, the risks vs. benefits of using generic drugs in patients with epilepsy are only known for the older drugs. In the case of phenytoin and carbamazepine, evidence suggests that generic substitution is unwise. However, for the more expensive newer anti-epileptic drugs, the data do not exist, and are not likely to be available any time soon.

Selection In Newly Diagnosed Patients

Because most patients will become seizure free on initial therapy, if selected appropriately for epilepsy syndrome, and because the newly diagnosed patients are considered ‘easy to treat’, drugs should be selected that are most likely to be safe and well tolerated. This is particularly true because once stabilized on an effective medication, patients may be reluctant to switch for fear of further seizures, even in the presence of side-effects. For these reasons, the first drug selected often becomes a long-term choice. Issues in selection of drug therapy in newly diagnosed patients are discussed in site.

Selection Of An Add-On Drug

When selecting a drug that will be added on to an existing regimen, a number of factors should be considered, including whether added benefit is foreseen, whether pharmacokinetic interactions will occur, whether compounded side-effects are likely, whether there is a risk of additional safety issues, and finally whether the drug can eventually replace the background drugs. It has not been proven that combining drugs with different mechanisms will actually increase the likelihood of seizure control. A second issue is whether the addition of a drug will lead to pharmacokinetic interactions. Interactions between old and new anti-epileptic drugs are listed in Tables Pharmacokinetic impact of old anti-epileptic drugs () on new anti-epileptic drugs and Pharmacokinetic impact of new anti-epileptic drugs (anti-epileptic drugs) on old anti-epileptic drugs. These potential interactions should be recognized, and doses of background drugs should be adjusted accordingly. Another type of interaction is called ‘pharmacodynamic’. In this type of interaction, levels do not change, but side-effects may be greater than expected if either drug was given by itself. For example, more side-effects may be seen when combining two of the the sodium channel blockers lamotrigine, carbamazepine and phenytoin. Lamotrigine may also cause more side-effects when given with valproate, but this combination has also been noted to be favourable in terms of efficacy in some patients. In contrast, some of the newer drugs, such as levetiracetam, gabapentin and pregabalin, seem to be easier to add on, with fewer additive side-effects.

Anti-Epileptic Drugs For Patients With Other Health Concerns

Some patients may be poor candidates for specific anti-epileptic drugs because of their past history. For example, patients who have demonstrated hypersensitivity to anti-epileptic drugs may have problems with more allergenic compounds. Some anti-epileptic drugs have a higher likelihood of cross-reacting with each other, including phenytoin, carbamazepine, phenobarbital and lamotrigine. In principle, patients who are overweight should avoid carbamazepine, gabapentin, pregabalin, valproic acid and vigabatrin as these drugs tend to increase weight, although they may be warranted in some patients. Similarly, patients who are overly thin or have eating disorders may not be ideal candidates for topiramate, felbamate or zonisamide, which suppress appetite. Drugs with greater potential for psychiatric problems such as irritability, mood disorders and psychosis include levetiracetam, topiramate and vigabatrin, whereas lamotrigine, carbamazepine and valproate may stabilize mood. Use of drugs in patients with concomitant medical conditions is discussed in site.