After stroke

The mainstay of treating seizures associated with acute or chronic stroke, after identification and elimination of toxic or metabolic disturbances lowering the seizure threshold, is the use of antiepileptic drugs (AEDs). Whether other therapies, such as antioxidants or neuroprotective agents, can prevent the later development of epilepsy if they are given shortly after the insult is an area of active research, but no clear recommendations have yet emerged.

Of about 30 drugs approved in the United States for the treatment of epilepsy, fewer than a dozen are in widespread use, including several developed and approved in the 1990s.100 There are few comparative studies, and those do not suggest major differences in efficacy. There is wider variation with respect to common adverse effects, available routes of administration, recommended dosing interval, interactions, and cost. These differences may be of particular importance for stroke patients.

The first step in treatment is to verify whether the episode was a seizure, as opposed to another type of transient event. (SeeDifferential diagnosis of possible seizures in stroke patients.) Once that condition is satisfied, the decision to treat depends on the likelihood and potential morbidity of another seizure versus risks of treatment. In reported studies, relatively few patients have multiple acute seizures, but the fact that most such patients are given AEDs may account for this.

Although there are no data on the efficacy of treatment of acute seizures specifically in the setting of stroke, it may be reasonable to extrapolate from the widely cited study by Temkin and colleagues of patients with moderate and severe head trauma.101 In these patients, phenytoin was effective in preventing seizures within the first week, but it did not decrease the incidence of a first late seizure in patients followed for approximately 1 year.

Potential morbidity of acute post-stroke seizures includes increase in blood flow to an already edematous cerebrum, elevation of blood pressure, and danger of aspiration. Although seizures and status epilepticus are uncommon causes for deterioration,34,43 cases have been reported that involve permanent worsening after a seizure, despite lack of evidence of any additional stroke.102

On balance, it is reasonable to treat for a limited period any patient with a stroke (especially a large stroke) who has an acute seizure without an obvious toxic or metabolic precipitant. The duration of such treatment should be at least a week, and perhaps a month or longer.

Of patients with a single late seizure, the majority, although not all, go on to develop epilepsy. Treatment for these patients is justified for a longer period, perhaps for 1 to 2 years.

When epilepsy has developed (i.e., after two late seizures), treatment lasting at least 2 years, as in other cases of epilepsy, is indicated.

The use of EEG findings in deciding when to stop treatment has been questioned, although unequivocal epileptiform discharges and periodic lateralized epileptiform discharges (PLEDs) in the acute or subacute setting would argue for continued treatment. A history of status epilepticus of any type would also suggest longer treatment duration, because, in adults (unlike in children) the occurrence of status epilepticus as the first seizure carries a higher risk of seizure recurrence.6

One potential consideration concerns whether many AEDs, including phenytoin, phenobarbital, and benzodiazepines, perhaps by virtue of their ability to dampen sustained repetitive neuronal firing, may inhibit recovery from stroke.103 There is moderate laboratory and sparse clinical evidence that this is true. Because of the very real risk of morbidity posed by seizures in many situations, this possibility should not argue against treatment after a seizure has occurred, but it may be a consideration in cases in which prophylaxis is considered (e.g., large hemispheric lesions with edema) and may affect AED selection when there is more information available.

In subarachnoid hemorrhage, most seizures occur with rerupture of the aneurysm.53 Currently, no data are available to specifically justify prophylactic anticonvulsant agents in the presentation of a ruptured subarachnoid hemorrhage, but awareness of the morbidity risk of a seizure in this setting and the extrapolation of the data for head trauma101 would favor treatment.

In lobar hemorrhages, owing to the prevalence of generalized seizures, one can argue that prophylaxis is justified.19,35,49,50Also, the theoretical risk of impeding recovery from infarction would not apply.

In cerebral venous thrombosis, prophylactic administration of AEDs is not justified if it has not presented with a seizure.

The choice of AEDs depends on potential routes of administration, interactions with other medications, and specific metabolic conditions, such as renal or hepatic impairment. Please visit the pages to the right for more information.

Adapted from: Bromfield, EB, and Henderson GV. Seizures and cerebrovascular disease. In: Ettinger AB and Devinsky O, eds. Managing epilepsy and co-existing disorders. Boston: Butterworth-Heinemann; 2002;269–289.
With permission from Elsevier ( 

Reviewed By: 
Steven C. Schachter, MD
Thursday, April 1, 2004