Because antiepileptic drugs (AEDs) target brain structures for their effect, they are being used to treat several neurologic conditions other than epilepsy. While control of seizures is the primary therapeutic use for such drugs, as specified in most submissions to governmental regulatory agencies, there is growing interest in AEDs for disease modification or for neuroprotection. Prophylaxis is defined as a process of guarding against the development of a specific disease by a treatment or action that affects pathogenesis. In contrast, to prevent means to render impossible by an advanced provision, or to keep from happening. A preventive intervention does not imply an effect other than alterations of symptoms, signs, or manifestations of a disease. Prevention may alter some component of a disease, but specificity is not implied, and effect on pathogenesis is not required.

Drugs that control epilepsy or change the intensity of seizures have been used to attempt disease modification. Within years of the introduction of phenytoin, investigators reported various open and uncontrolled trials purporting to show that drug’s efficacy as an agent effective for antiepileptogenesis. However, randomized, controlled trials of both phenytoin and valproate failed to show that use of drugs with effects on seizures had any impact whatsoever on chronic epileptogenesis, a marker for brain injury.^68,69,70 In fact, animal studies of the effect of corticosteroids and phenytoin on brain lipid peroxidation initiated by a free radical generating mechanism showed phenytoin associated with unabated brain injury while seizures were controlled.⁷⁹ Of interest, high dose corticosteroids in this model caused abatement of both seizures and lipid peroxidation. There is continued interest in disease modification that might accrue should antiepileptic drugs control seizures and somehow protect neural tissue as well.³ Mechanisms of action of AEDs that are critical to anticonvulsant activity have been considered as having a role in neuroprotection. If this were the case, the therapeutic value of AEDs would include not only control of seizure activity (anticonvulsant), but also the prevention of injury responses, resulting in disease modification.⁷⁶

Neuroprotective agents have been sought among various molecular designs, including calcium channel blockers, anti-inflammatory drugs, and various sterol compounds. While animal trials have suggested robust efficacy, clinical studies have failed. Whether the end-point is inappropriate or the nature of the injury to neural tissue of such intensity that a pharmacologic intervention could not possibly succeed has yet to be determined.⁵⁰

Better understanding of the mechanisms of brain injury has had impact on drug design, with introduction of several new antiepileptic drugs for clinical use that have unique actions. Mechanisms of injury in animals and in some human studies have included cerebral infarction, ischemia, status epilepticus, and traumatic brain injury.⁶⁶ Common cellular responses to injury include marked increase in the extracellular concentration of glutamate followed by increase in intracellular calcium and cell death.⁶ Marked increase in glutamate occurs in humans having seizures as measured with microdialysis.¹² Similar changes occur in animals with status epilepticus.⁷⁵ In stroke and hypoxia, failure of high-energy substrates cause failure of transporters with loss of membrane potential and neurotransmitter release.³⁶ Glutamate binding to specific receptor molecules initiates signaling that is coupled to channels permeable to ions or to G-protein second messenger systems.¹⁰ Inotropic glutamate receptors are categorized by selective responsiveness to N-methyl-D-aspartate (NMDA), kainate, or AMPA. Calcium entry into cells by the NMDA receptor complex in the various patterns of brain injury is critical to cell injury,⁷ while antagonists are protective.^5,17

One example of preventive treatment is the administration of anticonvulsants to patients with head trauma of such severity that hypertension and hypoxia associated with a convulsive seizure would complicate management. Antiepileptic medications are administered to patients who are thought to be at risk to have tonic–clonic seizures with the intent to prevent the occurrence of convulsive seizures. Patients with absence seizures may be treated with a broad-spectrum antiepileptic drug prior to occurrence of a tonic–clonic seizure, rather than using a syndromic-specific agent such as ethosuximide. Preventive treatment in this narrow context may be successful. However, some patients are given anticonvulsants with an apparent attempt to interfere with the process of epileptogenesis. Examples of this prophylactic use include the routine administration of antiepileptic drugs to patients with head trauma,⁵⁴ or to patients undergoing neurosurgical procedures requiring incision of the neocortex.⁷⁷ Although prevention of seizures is a worthy goal and may be effective, prophylaxis of epilepsy in the strict sense may not be effective, since no data are available to suggest that antiepileptic drug administration has any impact on the process of epileptogenesis.⁷⁰

Trauma dose as estimated by factors correlated with severity of head injury allow a crude prediction of the liability to develop posttraumatic epilepsy (PTE).¹⁴ Classification of head injury based on clinical evidence of trauma dose reveals a correlation with epilepsy risk.⁴ Patients with severe head trauma with cortical injury and neurologic sequelae, but with intact dura mater, have an incidence of epilepsy from 7% to 39%.⁴ However, if dural penetration occurs in association with neurologic deficits, then the range of epilepsy incidence is 20% to 57%.⁴ Application of weighted risk factors
allowing mathematical estimation of liability for development of PTE at the time of injury suggests correlation between severity of injury and subsequent epileptogenesis.⁷⁸ Brain volume loss also reflects trauma intensity and provides a correlation with PTE.⁵⁹

Fifty-seven percent of head-injured patients developing PTE will have their first seizure within 1 year of injury.⁵⁹ Although specific mechanisms remain unknown, the latency between injury and occurrence of convulsive seizures must represent the process of epileptogenesis. Whether a seizure occurs immediately after injury, within the first week, or beyond the first week may have prognostic significance for development of epilepsy.²⁵ The occurrence of an immediate seizure, within hours after trauma, and posttraumatic status epilepticus may complicate management of an injured patient by adding hypoxia and hypertension to the primary processes initiated by trauma. Although an immediate seizure may be a nonspecific reaction, such a symptom may herald the presence of an intracranial hematoma.²³ However, if seizures occur within the first week after injury (an early seizure), then an increase in the incidence of late epilepsy has been observed.²⁶ Other predictors of risk of late epilepsy include 24 hours of posttraumatic amnesia, the presence of a depressed skull fracture, or an intracranial hematoma.²⁴

Patients with absence epilepsy commonly develop generalized convulsive seizures as an additional manifestation of this epilepsy syndrome. Patients with the best prognosis for control of absence seizures have normal intelligence and a negative family history of epilepsy. These patients may have a 90% chance of remission. In contrast, the overall remission rate for all patients with absence ranges from 37% to 57%. Patients at high risk for the development of concurrent or subsequent generalized tonic–clonic seizures have late onset of absence epilepsy.⁵⁷

Drugs used for treatment of absence epilepsy include ethosuximide, clonazepam, and valproic acid. Ethosuximide has specific efficacy for nonconvulsive seizures, while valproate is effective in controlling both absence and generalized convulsive seizures.⁶¹ Ethosuximide used as initial therapy for this seizure syndrome may leave a patient vulnerable to the occurrence of convulsive seizures. Preventive treatment using a broad-spectrum anticonvulsant such as valproic acid would be appropriate for a patient with late childhood or early adolescent onset of absence seizures. There is no published report addressing the specific efficacy of this course of treatment as preventive for convulsive seizures, particularly with comparison of ethosuximide and valproate. Clinicians should be aware of the possibility of anticipation and prevention in this context by discussing such problems with parents of patients.

Six to nine percent of patients with stroke may develop the complication of epilepsy.³⁴ Precise prediction of development of postinfarction seizures may depend on etiology.³⁹ Not unlike head trauma, a differential liability to develop seizures may depend on acute versus late seizures relative to onset of the infarction.⁶³ The role of seizures as a so-called precursor to stroke remains controversial.³⁴ Patients with cerebral infarctions involving the cerebral cortex with persistent paresis have 20% liability to develop seizures.⁹ Although the risk of epilepsy following stroke has been recognized, administration of anticonvulsant as a prophylaxis is not a common practice and has not been reported. If medications were to be administered to this group, then specific knowledge of the risks and benefits would be most important before such an action were recommended.