According to the World Health Organization, status epilepticus (SE) is defined as seizures that are sufficiently frequent or prolonged to produce an “enduring epileptic condition.”1 Implicit in this definition is the concept that the event’s duration renders it physiologically distinct from other seizure types and, therefore, it requires specialized workup and management.
Although there is no well-determined absolute duration beyond which a prolonged seizure is definitively distinct from shorter ones, in practice SE is defined as continuous seizures, or a cluster of seizures without return to consciousness, lasting 30 minutes or more. With this definition, the incidence of SE in children approximates 18 to 38 per 100,000 persons/year.2 More than 40% of SE cases occur in children younger than 2 years of age.3
A reasonable and more inclusive definition of SE takes into account observations that individual seizures rarely last longer than 5 minutes, and that in clinical practice seizures longer than 5 minutes are often treated acutely. Accordingly, there is a trend toward reducing the duration required for diagnosis of SE, with some reports defining minimal SE duration as 5, 10, or 20 minutes; incidence of SE by these definitions has not been extensively studied.4
The recent trend toward reducing seizure duration in the definition of SE is supported by clinical and experimental data. For instance, Shinnar and colleagues found a bimodal distribution of new-onset seizure durations in children clustered around 2 means: 3.6 minutes (76% of cases), and 31 minutes (24% of cases).5 This corroborates the observation that seizures lasting longer than 5 minutes are likely to continue for at least several minutes longer, and fits well with the widely accepted practice to administer anticonvulsants acutely for seizures lasting 5 minutes or more. The rationale for earlier treatment of SE is also supported by data from animal seizure models that show loss and dysfunction of GABAA receptors and reduced sensitivity to benzodiazepines after prolonged seizures in rats.6-9
Numerous seizure types, if prolonged, can comprise SE. With some overlap, these can be divided into convulsive and nonconvulsive subgroups. Prolonged primary or secondary generalized tonic-clonic seizures, including febrile seizures, are the most common form of convulsive SE in children.
Nonconvulsive SE (NCSE) is a relatively heterogenous group of disorders.11 The clinical picture is often that of depressed consciousness ranging from confusion to coma. Particularly in the intensive care unit (ICU), NCSE is now becoming more widely recognized, in part due to increased use of electroencephalogram (EEG) monitoring. In a recent report, EEG ordered to evaluate unresponsiveness in the pediatric ICU showed a pattern consistent with NCSE in 33% of 178 cases.12 Prolonged absence seizures (absence status) are also a form of NCSE characterized by altered consciousness, at times lasting hours until the diagnosis is made by EEG. Another form of NCSE, one in which consciousness is typically preserved, is characterized by prolonged simple partial seizures such as focal motor seizures involving only a small muscle group or sensory seizures affecting a relatively small region. Rarely, simple partial SE lasts hours to days and is then termed epilepsia partialis continua (EPC), an epileptic state seen in some conditions such as Rasmussen’s encephalitis, malformation of cortical development, and mitochondrial disease.
SE may be provoked and symptomatic of an acute condition such as sepsis, meningitis, encephalitis, intracranial intoxication, or other metabolic disturbance. Therefore, especially in children without known epilepsy or a neurologic condition that predisposes them to seizures, the diagnosis of SE requires emergent workup for an underlying medical cause. Alternatively, SE may be remote symptomatic, or the manifestation of an idiopathic epilepsy. In such cases, a common scenario is SE in a patient with epilepsy who has missed a medication dose, or in whom medications were changed or withdrawn. Febrile seizures (febrile status) are another common cause of pediatric SE and are the leading cause of convulsive SE in children ages 1 to 3 years.10 The management problems in such instances may be restricted to aborting the prolonged seizures, and an immediate diagnostic workup for an underlying cause may be less relevant than in cases of acute symptomatic SE.
DIAGNOSIS AND TREATMENT
Convulsive SE is a medical emergency, in part because it can reflect a severe underlying medical condition, and also because prolonged seizures can be injurious to the brain. Among the mechanisms for SE-related brain injury are seizure-related excitotoxicity, increased cerebral metabolic demand, and compromise of cerebral oxygenation and perfusion due to hypoventilation and disturbances in cardiac output.
Treatment of convulsive SE generally requires simultaneous anticonvulsive and systemic management. Preservation of cerebral oxygenation and perfusion is the primary systemic concern. The airway should be cleared and protected, and high-flow oxygen should be provided. Continuous cardiorespiratory monitoring is essential. Intravenous access should be established, and blood samples should be drawn to assess causes such as metabolic disturbances and infection. Metabolic abnormalities such as hypoglycemia, hyponatremia, and hyperthermia should be corrected promptly.
Benzodiazepines are the most common first-line anticonvulsants given in SE. Typically, a single intravenous dose of lorazepam (0.1 mg/kg, not exceeding 4 mg) is given once and can be repeated after 2 to 4 minutes if seizures persist. Diazepam (0.2 mg/kg, not exceeding 10 mg) is an alternative option. If intravenous access is not available, diazepam (0.5 mg/kg, not exceeding 20 mg) can be given rectally.
If SE persists beyond initial administration of benzodiazepines, phenytoin or fosphenytoin should be administered. Fosphenytoin, a water-soluble prodrug that is rapidly converted to phenytoin after administration, is the preferred drug in most pediatric centers. It is dosed in phenytoin equivalents (PE) and should be given as 20 mg PE/kg over 7 minutes, not to exceed 1000 mg PE. If SE is not terminated with benzodiazepines and fosphenytoin, phenobarbital (20 mg/kg over 20 minutes) can be used.
Beyond the initial administrations of benzodiazepines, fosphenytoin, and phenobarbital, practice preferences vary widely between centers and are often individually adjusted to the SE case. Fosphenytoin administration (10 mg PE/kg) can be repeated. Intravenous valproate, propofol, or midazolam can also be used. Muscle paralysis and inhaled anesthetic agents are also used for refractory convulsive SE.
Treatment for nonconvulsive status epilepticus (NCSE) is similar to that for convulsive SE. As with convulsive SE, respiratory support is at times necessary, although generally less difficult in NCSE, where ventilation and oxygenation are often preserved and the respiratory management requires only clearing the patient’s oral secretions. Intravenous benzodiazepines (diazepam or lorazepam) are good initial drug choices for NCSE. However, in instances of absence SE, intravenous valproate is more beneficial than fosphenytoin or phenobarbital and should follow the initial benzodiazepine dose.
Death from SE is relatively uncommon in children and is often associated with a severe acute or degenerative brain disorder. Two studies show mortality associated with convulsive SE is 3% to 4% in children, in contrast to overall convulsive SE mortality of 22% in the general population.10,13
The risk of neurologic morbidity after SE is appreciable, and it is related to the underlying cause. Maytal and colleagues found that 17 (9%) of 193 children had residual neurologic deficits after recovery from SE. The majority of these patients had acute symptomatic SE, and none of 67 children with unprovoked or febrile SE studied prospectively had new neurologic findings after SE.10 However, recent reports suggest that more subtle forms of brain injury follow SE. Among these are epileptogenic changes in the mesial temporal lobe that increase the probability of recurrent seizures following SE. At the time of this writing, a large multicenter trial is underway to investigate the relationship between mesial temporal lobe injury and febrile SE.14,15