The choice of antiepileptic drugs (AEDs) entails a number of considerations detailed later. All these considerations cannot be fully or adequately addressed by controlled studies alone, and are thus grounds for individualization of therapy. Although evidence-based data from randomized controlled trials (RCTs) on drug efficacy are a critical part of the choice, as they provide an objective method to compare efficacy, they often do not address many other aspects that are important in such choices such as comparative efficacy, comorbidity, and ease of use.

During the process of patient evaluation and medication decision, several considerations affect treatment choices. These include:

1. 
   

   Comparative efficacy: This refers to the ability of medication to control seizures and is usually best determined by RCTs. However, efficacy of drugs in specific epilepsy syndromes and long-term efficacy over years, even for common seizure types, are difficult to determine using RCTs, and data are limited. Furthermore, the phenomenon of seizure aggravation from AEDs has not been studied adequately but will affect the choice of an AED.

   
   
2. 
   

   Relative tolerability: This refers to the medication adverse effect profile, which varies for each patient. A prominent example is the increased risk of liver toxicity for valproate (VPA) therapy in children under the age of 2 years with metabolic disorders. Relative tolerability also applies to lifestyle side effects including weight gain, gingival hyperplasia, alopecia, hyperactivity, and others. Children with behavior problems and/or with attention deficit disorder are particularly hyperactive with GABA-ergic drugs such as benzodiazepines barbiturates or VPA.

   
   
3. 
   

   Cost and availability: The cost of the newer AEDs may prevent their use, particularly in developing countries. Furthermore, many drugs are available only in some countries either because they are too expensive or because, paradoxically, they are too inexpensive (with little financial incentive for their importation), or for regulatory reasons.

   
   
4. 
   

   Ease of initiation of the antiepileptic drug: Medications that are titrated gradually such as lamotrigine (LTG) and topiramate (TPM) may not be chosen if rapid therapeutic levels are required. In these situations, medications with intravenous preparations or rapid oral titration schedules such as VPA, phenytoin (PHT), or levetiracetam (LEV) are alternative choices.

   
   
5. 
   

   Preexisiting medications and the potential for pharmacokinetic drug interactions: An example is the reciprocal ability of VPA to increase the epoxide level of carbamazpine (CBZ) and CBZ lowering the level of VPA.

   
   
6. 
   

   Availability of syndrome-specific and age-specific efficacy and tolerability data: Many AEDs are only proven to be effective in adult symptomatic and cryptogenic partial epilepsy, but have not been fully studied in children. In addition, some unique epilepsy syndromes and seizure types in childhood including neonatal seizures, West syndrome, early myoclonic encephalopathy, Ohtahara syndrome, and several other syndromes do not occur in adults. Furthermore, many of these disorders are relatively uncommon or rare and controlled studies are not available. This is also the case in more common pediatric epilepsy syndromes such as benign occipital epilepsy of childhood.

   
   
7. 
   

   The presence of comorbid conditions: For example, the presence of migraine in a patient with epilepsy may suggest a medication effective against both disorders such as VPA or TPM. Obesity may dictate against medications such as VPA. Other comorbid conditions include malignancies like leukemia where enzyme inducers reduce the efficacy of chemotherapy or increase the risk of relapse. Thus, enzyme inducers should be avoided in patients with cancer. In adolescent females of childbearing potential, enzyme-inducing AEDs may interfere with birth control pills or increase the risk of fetal malformation.

   
   
8. 
   

   Coexisting seizures: The presence of multiple seizure types also influences AED choice. Some medications have a broad spectrum of antiseizure effects whereas others exhibit a more limited spectrum. The presence of both absence and generalized tonic–clonic seizures may favor LTG and VPA rather than ethosuximide (ESM) or CBZ alone or their combination.

   
   
9. 
   

   History of prior response to specific antiepileptic drugs: For example, if a patient had previously responded to a GABA-ergic drug, then introduction of another GABA-ergic drug may be beneficial.

   
   
10. 
    

    Mechanism of drug actions: At present, the understanding of the pathophysiology of epilepsy does not allow for specific choice of AEDs based on the assumed pathophysiology of the epilepsy, and the presumed AED mechanism of action. However, it is preferable to not combine medications with similar mechanisms of action such as PHT and CBZ whose major effects are on sodium channels. In contrast, several combinations including LTG and VPA and TPM and LTG are anecdotally reported to have synergistic effects, possibly through different mechanisms of action. When combining AEDs, the choice of the second may depend on the background drug (i.e., the first drug preceding the second) and on the combined mechanisms of action. Thus, pharmacodynamic interactions and not just pharmacokinetic interactions may be important.

    
    
11. 
    

    Ease of use: This is particularly important in children. Medications administered once or twice daily are easier to use than medications given three or four times daily. In addition, a pediatric liquid preparation rather than tablets or capsules is an important consideration in children under age 6 years.

    
    
12. 
    

    Ability to monitor the medication and adjust the dose: It is difficult to maintain stable plasma concentrations of some AEDs and frequent AED blood levels may be required. The prototype is PHT, but many of the older medications also require blood level monitoring.

    
    
13. 
    

    Patient and family preferences: All things being equal, the choice between two or more acceptable alternative AEDs also depends on patient or family preferences. For example, patients or their parents may want to avoid gingival hyperplasia and hirsutism but may tolerate weight loss, or vice versa.

Some, but not necessarily all of the previous considerations have been addressed through “expert opinion” surveys¹ or guidelines developed by professional societies such as the International League Against Epilepsy (ILAE),² National Institute for Clinical Excellence (NICE),³ Scottish Intercollegiate Guidelines Network (SIGN),⁴ or the American Academy of Neurology (AAN).^5,6 Some guidelines are evidence-based (AAN), while others (NICE, SIGN) incorporate other considerations. However, no guidelines incorporate all considerations relevant to each patient. Thus, the process of choosing an AED involves incorporating evidence from RCTs, guidelines, expert opinion surveys, and all the previous considerations for individualizing therapy.

Individualization is time-consuming and requires expertise that may not be present in all treating physicians outside tertiary referral centers. Physicians can rely on algorithms based on guidelines or expert opinion, as well as on RCTs. These algorithms are not designed to replace sound clinical judgment or subspecialty referral. In complex patients, decisions are typically made by the treating neurologist or epileptologist.

INITIATING THERAPY

The decision to initiate antiepileptic therapy following a first seizure is based on several considerations. These include patient age, gender, vocation, driving status, epilepsy or seizure type, EEG findings, physical examination, neuroimaging findings, as well as relevant data from randomized clinical trials.^7,8,9,10,11 Only one study (performed on adults after a single seizure) was placebo-controlled.⁸ In general, all studies reveal a reduced risk of seizure recurrence in patients randomized to treatment, but a corresponding increase in adverse drug effects. The decision to treat is therefore complex when considering the risk of a new seizure, an abnormal EEG, radiologic, or neurologic abnormalities. Social and personal patient preferences also influence the decision-making process. Three studies examining long-term remission found that earlier treatment did not affect its incidence.^7,9,11 Two studies^9,11 included both adults and children and one included only children.⁷

Choosing an Initial Drug

Choosing an AED for the initial treatment of a new-onset seizure disorder is challenging due to a lack of randomized, double-blind, placebo-controlled trial data. The SANAD study,^12,13 which was open label and included adults and children older than 4 years, attempted to address this issue. Newly diagnosed patients whose physicians thought CBZ was the appropriate initial drug were randomized to carbamazepine, oxcarbazepine (OXC), LTG, gabapentin (GBP), or TPM,¹² and those whose physicians thought VPA was the appropriate drug to VPA, LTG, or TPM.¹³ In the first group, LTG was more acceptable than CBZ due to differences in adverse effects, and both were more successful than GBP due to its decreased efficacy, or TPM-induced adverse effects. In the second group, VPA was more successful than LTG because of greater efficacy, and both were superior to TPM due to its side effects. This study has been criticized because AED doses and titration schedules could have substantially biased the results.

Currently, the choices of initial monotherapy for old and new AEDs exist in guidelines from the ILAE,² SIGN,⁴ NICE,³ and the AAN⁵ (Table 49–1).

Adding an AED

There are no RCTs that address the choice of a second AED when the first has failed. Furthermore, no trials address the choice of a first drug in febrile seizures or specific pediatric epilepsy syndromes with the exception of refractory partial seizures or resistant Lennox–Gastaut syndrome (LGS). Several studies of partial seizures patients achieved seizure control leading to approval for monotherapy use. Whereas the choice of AEDs is often influenced by the results of these studies, in the final decision is often individualized. Guidelines from NICE,³ SIGN,⁴ and AAN⁶ (Table 49–1) currently exist for add-on therapy in different seizure and epilepsy types.

EFFICACY AND TOLERABILITY

In 2004, the Quality Standards Subcommittee (QSS) and the Therapeutics and Technology Assessment (TTA) Subcommittee of the American Academy of Neurology and the American Epilepsy Society published two reports assessing the evidence on efficacy and tolerability of the new AEDs for new onset⁵ and refractory epilepsy.⁶ Studies were evaluated for their class of evidence based on study design and methodology into four classes (I–IV). Evidence was translated into recommendations, rating in strength from Level A to Level U.

In 2006, the ILAE published evidence-based treatment guidelines on antiepileptic drug efficacy and effectiveness as initial monotherapy for epileptic seizures and syndromes.² The criteria used to evaluate the studies for their class of evidence, as well as level-of-evidence of the recommendations were adapted from the US Agency for Health and Policy Research¹⁴ and the American Academy of Neurology¹⁵ scoring systems. The criteria used to rate the studies in the ILAE treatment guidelines share many similarities to those of the TAA and QSS subcommittees’ reports but include more stringent criteria, such as longer study duration and absence of forced-exit criteria. For instance, some studies that met a class I rating by the TAA and QSS criteria would be classified as class III by ILAE criteria. In reviewing the literature for this chapter, we adopted the ILAE study classification and rating of evidence for the sake of consistency with the guidelines.² In addition, and since the focus of this chapter is on pediatric and adolescent epilepsy, studies that did not include children below 16 years of age were excluded.

First Generation AEDs

First generation AEDs include drugs used broadly for the treatment of both partial-onset and primary generalized tonic–clonic seizures such as VPA, or predominantly partial seizures such as CBZ, PHT, and phenobarbital (PB). Other drugs like ESM) and adrenocorticotropic hormone (ACTH) are primarily used to treat very specific seizure types such as absence seizures or infantile spasms (IS).

With the advent of the second and third generation AEDs, antiepileptic treatment recommendations are changing in favor of newer agents, with some older agents like acetazolamide (ACZ) or bromides (Br) becoming obsolete. The use of older AEDs for pediatric epilepsy was mostly based on uncontrolled trials, extrapolation from studies done on adults and anecdotal reports. Most have not been rigorously tested in clinical trials for efficacy or tolerability. Table 49–2 summarizes the available evidence on the efficacy/effectiveness of the older AEDs for the treatment of pediatric epilepsies. The rating of evidence is based on the highest level available in the literature, consistent with ILAE classification of studies and rating of evidence.²

Acetazolamide

ACZ is used as adjunctive treatment for catamenial epilepsy based on class IV retrospective studies. Similarly, several class IV studies report the use of ACZ as add-on treatment for refractory absence, partial-onset, or myoclonic seizures with significant decrease in seizure frequency.^{16,17,18,19,20}

ACTH and Oral Corticosteroids

ACTH and oral corticosteroids (prednisone) are indicated for the treatment of IS. However, there are no clinical trials that compared the effectiveness of ACTH or oral corticosteroids versus placebo in controlling IS. Two class III RCTs revealed that ACTH was superior to prednisone in spasm cessation.^21,22 When compared to vigabatrin (VGB), ACTH results in a higher and faster response rate in spasm control, with similar cognitive development after 9–44 months of treatment²³ (class III). In another class III open-label RCT, ACTH had similar short-term efficacy and tolerability to nitrazepam for the treatment of cryptogenic IS.²⁴ ACTH may therefore be considered for initial monotherapy of IS, while oral corticosteroids may be considered as alternative first-line monotherapy. In addition, few class IV studies have reported the use of ACTH for the treatment of myoclonic seizures²⁵ and LGS^26,27 with variable response rates.

Clobazam

Evidence from one class III double-blind and one class III open-label trials suggest that clobazam is possibly equally effective to PHT or CBZ as initial monotherapy for partial or generalized-onset epilepsy.^28,29 Based on several class III double-blind, single-blind, and open-label trials, clobazam is possibly efficacious as adjunctive treatment or monotherapy for refractory epilepsy of mixed types, and for catamenial epilepsy. However, all studies suffered from small sample sizes.³⁰ A recent class III double-blind, placebo-controlled crossover study in 24 women with refractory catamenial epilepsy reported effectiveness in 78% of subjects.³¹

Clonazepam

Evidence from one class IV study found clonazepam to be effective for treatment of myoclonic, atonic and tonic seizures in patients with mental retardation.³² A class III double-blind RCT compared clonazepam to CBZ as initial monotherapy of psychomotor seizures and found similar efficacy and tolerability.³³ A class IV comparative study of 40 children with benign epilepsy of childhood with centrotemporal spikes (BECTS), aged 3–11 years, revealed that clonzazepam resulted in disappearance of rolandic discharges in 75% patients as compared to 10% of children receiving VPA and 0% of children treated with CBZ. However, seizure incidence, type, and clonazepam blood levels were similar between patients whose discharges disappeared and those who did not.³⁴ In a class III double-blind, placebo-controlled RCT, administration of a single adjunctive low dose of intramuscular clonazepam to 15 children with focal or generalized epilepsy reduced epileptiform discharges on long-term EEG monitoring as compared to placebo with a concomitant reduction of seizures.³⁵

Clorazepate

One class IV study reported excellent results with clorazepate when used in 59 patients (ages 7 months to 45 years) with intractable seizures of different types.³⁶

Midazolam

A class III open-label RCT compared intranasal midazolam to intravenous diazepam in 47 children (6 months to 5 years in age) with febrile seizures lasting more than 10 minutes. Intranasal midazolam was as safe and effective as intravenous diazepam. Seizures were controlled faster with diazepam but overall time to cessation of seizures after arrival to hospital was faster with midazolam,³⁷ due to the easy and fast intranasal drug administration.

Nitrazepam

A case series (class IV study) of 35 children (ages 7 months to 19 years) with different seizure types (petit mal, IS, myoclonic, psychomotor) reported the use of nitrazepam as add-on treatment with no placebo control. There was very good response in atonic seizures and IS, whereas there were poor and variable responses in myoclonic seizures, petit mal, “myoclonic type of absence, and complex partial seizures.³⁸ In IS, two case series (class IV) reported cessation of spasms in 30%–54% of patients with resolution of hypsarrythmia in 46%.³⁹

Another class III unblinded RCT compared the efficacy and safety of nitrazepam versus ACTH in 52 patients below 2 years of age with cryptogenic IS over 4 weeks. Both treatments were similarly effective in reducing spasms. Adverse events were similar in frequency for both drugs but more severe in the ACTH group.²⁴

Oxazepam

Oxazepam was introduced in 12 patients, ages 2–58 years, with complex partial seizures as monotherapy or as adjunctive treatment in a class IV nonrandomized, placebo-controlled, blind crossover study for 4 months followed by an open-label, 3–6-month phase. There was an excellent response in 10 patients; transient drowsiness or ataxia were observed.⁴⁰

Bromide (Br)

First used as an antiepileptic about 150 years ago, Br is currently restricted to the treatment of severely refractory seizures of different types.^{41,42,43,44,45} The use of Br is based on anecdotal reports as there are no RCTs that assessed efficacy as initial or adjunctive therapy for refractory seizures.

Carbamazepine

CBZ is an initial monotherapy for the treatment of partial-onset and generalized-onset seizures. However, CBZ has not been tested against placebo in double-blind RCTs. In several comparative RCTs, CBZ was tested against VPA, PB, GBP, LMG, TPM, and OXC. One systematic review of class III studies found that in patients 2–68 years of age with either partial or generalized tonic–clonic seizure, CBZ was equally effective to PB and better tolerated as monotherapy for either type of seizures.⁴⁶ In a meta-analysis of class III studies, CBZ was more effective than VPA as first-line monotherapy for partial-onset seizures and equally effective to VPA as first-line monotherapy for generalized-onset seizures.⁴⁷ Several class I and II studies found CBZ equally effective to GBP,⁴⁸ LTG,⁴⁹ TPM,⁵⁰ and OXC⁵¹ as monotherapy for new-onset epilepsy (partial or primary generalized). Based on the available evidence, CBZ is possibly effective as first-line drug for the treatment of partial-onset and primary generalized seizures.

Ethosuximide/Methsuximide

Available evidence from limited class III clinical trials suggest that ESM is equally effective to VPA in controlling absence seizures.⁵² However, the studies are of poor quality, and there are no placebo-controlled double-blind RCTs assessing efficacy of ESM in absence-seizures. Anecdotal reports (class IV studies) suggest that combining ESM with VPA may reduce seizures in epilepsy with myoclonic absence, eyelid myoclonia with absence and juvenile myoclonic epilepsy (JME).⁵³

Methsuximide (MSX) use in pediatric epilepsy is based on small retrospective studies and case series (class IV studies). These studies suggest that MSX is potentially efficacious as adjunctive therapy in refractory epilepsy.^54,55,56 MSX use for absence seizures is based on anecdotal reports. A single small series reported excellent control of JME with MSX monotherapy.⁵⁷ Currently, there is insufficient evidence to recommend MSX for absence seizures or JME.