Following the diagnosis of epilepsy and making the decision to start a treatment, as discussed in the prior chapter, the clinician must choose the most appropriate initial medication. This decision requires knowledge of the clinical pharmacology of medications used for the treatment of epilepsy and an ability to individualize treatment to the child in question.¹ The ultimate goal is complete control of seizures without producing adverse effects. Although the perfect drug does not exist, the current therapeutic armamentarium offers a broader range of choices than available ever before with each drug having some advantages and disadvantages. A single drug should be chosen so that the child is treated with monotherapy, which has the likelihood of producing the fewest number of side effects.^2,3 The first issue in deciding upon initial therapy depends on seizure type. An appreciation of epilepsy syndrome is also extremely important, as it may select for or against an agent and may exacerbate seizure types, which have not been yet appreciated in the child.

Besides the expected efficacy of the medication for the seizure type(s), other important aspects that must be considered in children with epilepsy include the spectrum of toxicity, side effects, likelihood of compliance, ease of administration, and the potential for additional side effects following long-term therapy, if prolonged treatment is thought to be probable. Besides possible immediate age-specific organ toxicities, potential side effects affecting cognition and behavior must be considered. Effects upon daily activity such as sleeping, eating, and behavior are of most concern to caregivers, but the potential for developing rare but life-threatening adverse effects must be carefully reviewed and appropriately discussed.

Following the establishment of the need to treat and having a clear description of seizure type, the clinician and caregiver should together decide upon initial therapy. The clinician advising the family regarding antiepileptic drug (AED) use will frequently recommend an agent that is comfortable.

This comfort comes from prior experience in prescribing the medication, recommendations of experts and peers derived from multiple types of studies. For some, randomized clinical trials (RCTs) offer the strongest evidence concerning which AEDs should be selected as first line.

THE ROLE OF CLINICAL TRIALS IN CHOOSING A DRUG

RCTs of AEDs applicable to treatment of children with epilepsy are performed only following studies in adults with refractory partial seizures. Pediatric partial seizure trials are initiated and then recommendations follow. Specialized studies in children with the Lennox–Gastaut syndrome and childhood absence epilepsy have only recently generated specific recommendations for the use of AEDs in these disorders.¹ Thus, the timing of the performance and reporting of the most rigorous type of studies applicable to the treatment of children with epilepsy may be substantially delayed following the approval and introduction of a newer AED. A discussion of clinical trials versus antedoctal reports is presented in subsequent chapters. Nevertheless, many AEDs are never formally studied in specific childhood epilepsy syndromes because of significant procedural difficulties in conducting these studies, particularly in infants and young children with relatively uncommon syndromes.

The clinician must therefore balance the importance of RCTs, comparative studies, and expert opinion with their own experience when individualizing to selecting the drug of choice for a specific child. Thus, effectiveness rather than pure efficacy is a guiding principle of AED treatment.

Classic studies demonstrated a near equipoise in the treatment of convulsive pediatric epilepsy with phenobarbital, carbamazepine, phenytoin, and valproate; however, the intolerability of barbiturates was quickly identified as other agents became available; phenobarbital became a second-line medication because of its lack of tolerability, rather than its efficacy.⁴

Various authoritative agencies recommended approval for the use of newer AEDs, such as lamotrigine, oxcarbazepine, topiramate, and levetiracetam for partial onset seizures and lamotrigine has been added to prior approvals of ethosuximide and valproate for absence epilepsy.¹ Topiramate and levetiracetam have recently been approved to treat primary generalized epilepsy, primarily based on trials involving patients with juvenile myoclonic epilepsy. Felbamate, topiramate, and lamotrigine have proven efficacious for seizures associated with Lennox–Gastaut syndrome in well-controlled RCTs.¹

The case of infantile spasms demonstrates the difficulties in doing large trials in a relatively rare disorder. Following extensive review, the American Academy of Neurology has published a guideline establishing ACTH as probably effective and vigabatrin as possibly effective for the treatment of this encephalopathic epilepsy despite significant side effects of both agents. Although other AEDs including valproate, topiramate, and zonisamide may be efficacious in some children with infantile spasms, the published studies are frequently too small to allow establishing definitive guidelines.⁵

Recommendations from the National Institutes of Clinical Excellence⁶ state that the newer AEDs including gabapentin, lamotrigine, oxcarbazepine, tiagabine, topiramate, and vigabatrin are recommended for epilepsy management in children when used within their licensed indications and in children who have not benefited from treatment with older agents such as carbamazepine or valproate, or in those for whom the more established AEDs were inappropriate. Within the NICE Guidelines, vigabatrin is recommended as first-line therapy for infantile spasms. Other recommendations are given in Table 47–1. Expert opinion surveys performed using similar methodology have recently been published.^7,8,9 Through these clinically based expert surveys, it becomes obvious that certain AEDs such as carbamazepine, oxcarbazepine, and valproate are well-established agents and are utilized widely despite their potential adverse effects. These consensus surveys also demonstrate that the newer agents such as lamotrigine, topiramate, and levetiracetam have become somewhat more popular over time.

Valproate remains the treatment of choice for Lennox–Gastaut syndrome with topiramate and lamotrigine also first line. For the acute treatment of prolonged seizures (with or without fever) and clusters of seizures rectal diazepam was most preferred. For benign childhood epilepsy with centrotemporal spikes (BECTS), oxcarbazepine and carbamazepine were treatments of choice with gabapentin, lamotrigine, and levetiracetam also first line. Ethosuximide was rated as the first choice for childhood absence epilepsy, with valproate and lamotrigine also first line. For juvenile absence epilepsy, valproate and lamotrigine were treatments of choice. For juvenile myoclonic epilepsy in adolescent males, valproate and lamotrigine were preferred with topiramate also being first line, whereas in adolescent females lamotrigine was favored with topiramate and valproate other first-line options. First-line treatment for neonatal status epilepticus was intravenous phenobarbital, with lorazepam, and fosphenytoin also first line. Valproate was more commonly recommended by experts internationally as first-line therapy than in the United States.¹⁰