Status Epilepticus

Published on 20/04/2015 by admin

Filed under Neurology

Last modified 20/04/2015

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 2960 times

Chapter 58 Status Epilepticus

The spectrum of seizure events extends from isolated, brief seizures to status epilepticus (SE), incorporating a full range of recurrent unprovoked seizures and prolonged or acute repetitive seizures. Typically, seizures are brief and self-limited. There has been considerable research directed at efforts to explain the mechanisms underlying seizure termination, as well as perpetuation, in SE. SE is a pediatric and neurologic emergency associated with significant morbidity and mortality [Dodson et al., 1993; Mitchell, 1996; Pellock, 1993a, 1994; Towne et al., 1994; Weise and Bleck, 1997; Wilson and Reynolds, 1990]. Its prompt recognition and management lead to the best chance of successful outcome. SE may represent the brain’s reaction to an acute insult, or it may be a manifestation of already existing epilepsy, either as the initial symptom or as a prolonged exacerbation of seizures [DeLorenzo et al., 1992, 1995, 1996].

Although the definition and classification of SE have been changed numerous times, the term refers to seizures that continue for a prolonged period. Studies suggest that SE frequently goes unrecognized, and that its occurrence has been underestimated in the general population [Treiman, 1993]. One study [O’Dell et al., 2005a] documented that 34 percent of cases of febrile SE presenting in emergency departments across five centers were not recognized as such by the hospital personnel assigned to treat them. Based on this finding, febrile SE has only a 2 out of 3 chance of being recognized in an emergency treatment setting.

This chapter reviews the pathophysiology, definition, classification, epidemiology, etiology, treatment, and prognosis of SE as it occurs in children. The mortality associated with SE is greater in adults than in children, but morbidity and mortality in children are considerable without treatment. Accordingly, current thinking about optimal management uses a more aggressive clinical approach to this neurological emergency, including prompt recognition and initiation of therapy and accelerating the progression of treatment for more rapid termination of the episode [Millikan et al., 2009].


A distinguishing feature of SE is the self-sustaining seizure condition. The pathophysiology and biochemical changes underlying the evolution from discrete seizure to SE remain unclear [Lowenstein and Alldredge, 1998; Pellock, 1994; Wasterlain et al., 1993]. Pathophysiological changes that accompany SE can be divided into neuronal (cerebral) and systemic effects. The mechanisms involved in the initiation and maintenance of SE may be different. Ultimately, SE results from a failure of inhibitory mechanisms.

Gamma-aminobutyric acid (GABA) is the most prevalent inhibitory neurotransmitter in the brain. GABAA receptors are postsynaptic ionotropic receptors that bind directly to chloride channels, producing a fast inhibitory postsynaptic potential (IPSP). GABAA receptors are the binding sites for the benzodiazepines and it is the activation of this receptor that accounts for its antiseizure effect. A unique feature of SE compared to brief seizures is the time-dependent development of pharmacoresistance to benzodiazepines [Mazarati et al., 1998]. In animal models, investigators studied GABAA receptor currents by whole-cell patch-clamp techniques in CA1 pyramidal neurons acutely dissociated from rats undergoing lithium/pilocarpine-induced limbic SE and from naive rats [Kapur and Coulter, 1995]. The GABAA receptor current was absent in 47 percent of SE neurons and reduced in 55 percent of the remainder, compared with naive neurons, thus aiding in seizure perpetuation. Kapur et al., using a paired-pulse technique in an electrogenic model of experimental SE, showed that a marked deterioration of GABA-mediated inhibition occurs during continuous hippocampal stimulation [Kapur et al., 1989].

Additionally, more recent work has demonstrated the development of benzodiazepine pharmacoresistance shortly after the onset of ictal spike wave activity [Jones et al., 2002] and in young naive rats [Goodkin et al., 2003]. Treiman et al. reported similar loss of inhibition in hippocampal slices obtained during various electroencephalography (EEG) stages in lithium-/pilocarpine-induced SE [Treiman et al., 2006]. Altered receptor function and changes in representation affect both seizure representation and consequences in the neonatal brain. GABAA receptors are heteromeric protein complexes that mediate most fast synaptic inhibition in the forebrain and have many distinct subtypes. Adult rats that develop epilepsy following pilocarpine-induced SE and adult patients with refractory temporal lobe epilepsy demonstrate significant alterations in GABAA receptor properties in hippocampal dentate granule neurons [Gibbs et al., 1997; Brooks-Kayal et al., 1998, 1999]. In rat pups exposed to pilocarpine-induced SE, different changes are seen in α-1 subunit expression and augmentation compared with those in adult rats. This produced the opposite effect and may serve to enhance inhibition; the rat pups did become epileptic, unlike the adult rats who developed recurrent spontaneous seizures [Zhang et al., 2004a]. Nevertheless, changes in the function of the immature GABAA receptor, from the dual role of excitatory-inhibitory to inhibitory as the rat matures in infancy, may contribute to increased excitability and hence more seizure susceptibility in the neonate [(Khazipov et al., 2004]. Additionally, alteration in glutamate receptor representation may alter seizure susceptibility. Alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors are over-represented in the premature brain, initially in the oligodendrocytes; later they shift their main representation to neurons in the cortex and hippocampus. This condition may partly explain increased risk of periventricular leukomalacia in the preterm infant and increased seizure susceptibility in the term infant [Jensen, 2002]. In one study the effect of SE induced by lithium/pilocarpine in 10-day-old rats demonstrated that pilocarpine-induced prolonged SE caused long-term changes in both glutamate receptors and transporters in hippocampal dentate gyrus. These include a decrease in glutamate receptor 2 mRNA expression and protein levels, as well as an increase in protein levels of the excitatory amino acid carrier 1[Zhang et al., 2004b].

Different studies provide evidence suggesting that endocytosis of GABAA receptors takes place as a seizure transitions to SE. This leads to a decrease in the number of GABAA receptors on the synaptic membrane and therefore a decreased response to benzodiazepines [Naylor et al., 2005; Goodkin et al., 2008]. Resistance to other antiseizure medications, such as phenytoin, has also been demonstrated. In addition to GABAA receptor trafficking, changes in receptor expression and phosphorylation have been shown [Brooks-Kayal et al., 1998, 1999; Terunuma et al., 2008]. Functional changes in voltage-dependent sodium channels have also been described after pilocarpine-induced SE in rats [Remy and Beck, 2006].

Glutamate is the primary excitatory amino acid neurotransmitter and binds to several neuronal receptors, including the N-methyl-d-aspartate (NMDA) receptor, which is activated by depolarization, as well as the AMPA receptor, which mediates fast synaptic transmission in the central nervous system. Using a paired-pulse method in a continuous hippocampal stimulation-induced SE model, NMDA receptors become activated during continuous hippocampal stimulation, and NMDA antagonists block the deterioration of GABA-mediated inhibition [Kapur and Lothman, 1990]. During SE, both NMDA and AMPA receptor subunits increase at the synaptic surface [Mazarati et al., 1998; Wasterlain et al., 2002a]. This increase in glutamate receptors further enhances excitability and is proconvulsant in the midst of uninhibited seizures. In a model of self-sustaining SE, seizures are abolished by NMDA receptor antagonists [Wasterlain et al., 2000].

Neuropeptides have emerged as having a role in SE. Substance P agonists facilitate the initiation of SE in animal models. During the course of SE, the de novo expression of substance P occurs in the cells, such as dentate granule cells, that do not usually express it, possibly playing a role in maintenance of SE [Wasterlain et al., 2002b]. In addition, galanin, a 29–30 amino acid peptide, suppresses hippocampal excitability by post- and presynaptic actions [Mazarati et al., 2000]. However, galanin appears to be depleted by SE.

Laminar necrosis and neuronal damage after prolonged seizures are similar to those after cerebral hypoxia. Neuronal injury and cell death from SE are most prominent in areas rich in NMDA glutamate receptors, including the limbic region. The increase in intracellular calcium is critical to cell death, and calcium activates proteases and lipases that degrade intracellular elements, leading to mitochondrial dysfunction and fatal cellular necrosis. Although young animals may be less likely to develop brain damage from SE [Holmes, 1997; Moshè, 1987], studies using alternative models demonstrate hippocampal cellular injury, even in immature rodents [Sankar et al., 1997; Thompson and Wasterlain, 1994]. It is believed that the glutamate-initiated calcium-dependent cascade is mechanistically similar to the NMDA receptor-mediated cell death occurring during cerebral ischemia. In SE, the degree of neuronal injury is related to seizure duration. In a study of limbic SE induced in adult rats using perforant path stimulation, when animals were allowed to recover, there was evidence of mitochondrial injury and dysfunction demonstrated possibly through a free radical mechanism of injury [Cock et al., 2002]. Further evidence suggests that acute and long-term changes in gene expression may occur after prolonged seizures. These changes in the expression of messenger RNA (mRNA) may lead directly to some of the observed hyperexcitability [Rice and DeLorenzo, 1998].

In animal studies using adolescent baboons, induced SE lasting 1.5–5 hours produced neuronal loss in the hippocampus, cerebellum, and neocortex. Significant cell loss continued to occur, although to a lesser extent if the animals were paralyzed and ventilated with maintenance of oxygen, carbon dioxide, serum glucose, body temperature, and blood pressure [Meldrum, 1974, 1983]. Similar changes have been produced in rat models, as well [Cavalheiro et al., 1987; Sperber et al., 1989].

There are limited data in humans. However, evidence from neuropathology, as well as imaging, has been presented in a pediatric case with direct excitotoxic injury in the absence of hypoxia-ischemia [Tsuchida et al., 2007].

Generalized convulsive SE is associated with serious systemic effects resulting from the metabolic demands of prolonged seizures and the autonomic changes that accompany them: alterations in blood pressure and heart rate, incontinence, emesis, acidosis, hypoxia, changes in respiratory function and body temperature, leukocytosis, rhabdomyolysis, and extreme demands on cerebral oxygen and glucose use [Simon et al., 1997]. Circulating catecholamines increase during the initial 30 minutes of SE, resulting in a hypersympathetic state. Tachycardia, sometimes associated with more severe cardiac dysrhythmias, occurs and may be fatal, but this seems more common in adults [Boggs et al., 1993]. Cardiac output also diminishes, and total peripheral resistance increases, along with mean arterial blood pressure, perhaps because of the sympathetic overload. Hyperpyrexia may become significant during the course of SE, even without prior febrile illness in both children and adults, and may persist for some time. Fever may influence the process of neuronal injury [Liu et al., 1993].

In addition, serum pH and glucose levels frequently are abnormal because lactic acidosis increases from increased anaerobic metabolism. Associated respiratory acidosis also may occur as a result of hypoventilation, hypoxia, and pulmonary edema. An increase in the peripheral white blood cell count frequently occurs in the absence of infection. Rhabdomyolysis may compromise renal function. Thus, the cerebral physiologic changes linked to SE are accompanied by increased metabolic demand for oxygen and glucose, and are further complicated by a variety of systemic changes, along with the pathology responsible for SE. As the cascade of neurophysiologic changes occurs, increased lactate from anaerobic metabolism continues, and excitatory neurotransmitters bombard cells, which accelerates metabolic activity and leads to overall neuronal failure. Recovery from this complicated derangement of metabolism is time-dependent; more prolonged seizures produce further neuronal injury and death.

Unlike in adults, in whom biology is more static, the effect of growth and development influences the impact of seizures, both in clinical presentation and in the biologic consequences for the developing brain. Most of our knowledge derives from animal models, so its applicability to the human situation is unclear. Experimental models of seizures in immature animals suggest comparatively less vulnerability to seizure-induced brain injury than in mature animals [Wong and Yamada, 2001]. Repetitive or prolonged neonatal seizures may increase the susceptibility of the developing brain to experience subsequent seizure-induced brain injury later in life. This susceptibility appears to be more closely related to alterations in neuronal connectivity and network properties, rather than to increased cell death during the neonatal period [Holmes et al., 1998; Koh et al., 1999; Schmid et al., 1999].

Rats exposed to early-life seizures demonstrate persistent changes in CA1 hippocampal pyramidal cells, possibly leading to long-term changes in behavior and learning and in epileptogenicity [Villeneuve et al., 2000].


SE is best considered as a state produced by continuous or repetitive seizures, which has the potential to produce significant systemic or neuronal injury if not aborted. This definition, however, does not provide a clinically useful treatment guideline.

SE is internationally classified as a seizure lasting more than 30 minutes or recurrent seizures producing more than 30 minutes during which the patient does not regain consciousness [ILAE, 1981]. The World Health Organization previously defined SE as “a condition characterized by epileptic seizures that are so frequently repeated or so prolonged as to create a fixed and lasting condition” [Gastaut, 1982]. Lack of recovery for a fixed period, possible frequent repetition, prolongation, and possible propagation of further seizures are inherent in this definition. In the past, the definition of SE required 1 hour of continuous seizures, but more recent studies have used a 30-minute duration of continuous or recurrent seizures without full recovery as the standard clinical and electrographic definition of SE. The recognition that longer duration of seizures increases risk for long-term injury and the risk for fracture during seizures and their treatment has required a definition implying need for expediency in stopping prolonged seizures. Lowenstein and associates proposed an “operational” definition of 5 minutes or more of continuous seizures or “two discrete seizures between which there is incomplete recovery of consciousness” in adults and children older than 5 years of age [Lowenstein et al., 1999]. This definition applies primarily to generalized convulsive SE and may be used to direct treatment to avoid refractory SE, as well as its sequelae. Aggressive early treatment is justified by recent work demonstrating a 10-fold lower rate of mortality for seizures of 10–29 minutes’ duration versus those lasting longer than 30 minutes [DeLorenzo et al., 1999].

Because of the difficulty in diagnosing and quantifying seizures in the neonate, no broadly accepted definition of SE in the neonate exists. A proposed definition is either 30 minutes of continuous electroencephalographic seizures or presence of seizure activity for 50 percent of the EEG recording time, with or without the expression of coincident clinical signs (Scher et al., 1993b). Debate continues regarding what constitutes a neonatal seizure. Neonatal seizures can be broken down into three categories: electroclinical, electrographic, and clinical only. Controversy still exists about whether episodic abnormal movements seen in some infants, not accompanied by simultaneous ictal discharges on the EEG, are true seizures. Many neonatal paroxysmal events classified as “subtle seizures” have no EEG correlate [Mizrahi and Kellaway, 1998]. Typical subtle seizures include movements of progression, such as bicycling, oral-buccal-lingual movements, such as chewing and tongue thrusting, and other movements, such as random eye movements. Other movements that typically have no EEG correlate include generalized tonic posturing. Movements typically demonstrating simultaneous ictal discharges on EEG include focal clonic, multifocal clonic, and focal tonic; myoclonic movements may or may not have an EEG correlate [Mizrahi and Kellaway, 1998].


Any type of seizure may become prolonged and thus develop into SE (Box 58-1). Classification of SE should be performed by observing the clinical events and combining electrographic information when possible. The fundamental distinction between seizures is that some are generalized from onset, whereas others are partial in onset. The latter type may or may not then secondarily generalize. From a management standpoint, however, it may be more useful to consider whether the event is convulsive or nonconvulsive, as this may impact more directly on ready recognition and intervention.

Generalized tonic-clonic SE is the most dramatic and life-threatening form of SE. Myoclonic, generalized clonic, and generalized tonic SE occur primarily in children. These children usually have encephalopathic epilepsies [Lockman, 1990; Pellock, 1994; Treiman, 1993], and their consciousness may be preserved throughout the attacks. About one-half of the cases of generalized clonic SE occur in normal children, in whom it is associated with prolonged febrile seizures; the remaining half are distributed among those children with acute and chronic encephalopathies [DeLorenzo et al., 1992]. Generalized tonic SE appears most frequently in children, particularly those with the Lennox–Gastaut syndrome. Prolonged generalized tonic convulsions have been precipitated by benzodiazepine administration.

Nonconvulsive SE also may include complex partial, simple partial, and absence seizures that continue for more than 30 minutes [Kaplan, 1996; Scholtes et al., 1996; Stores et al., 1995]. Complex partial SE may be manifested as an epileptic twilight state marked by a cyclic variation between periods of partial responsiveness and episodes of seemingly motionless staring and complete unresponsiveness accompanied, at times, by automatic behavior [Delgado-Escueta and Treiman, 1987; Privitera, 1997; Scher et al., 1993a; Treiman, 1993]. Simple partial SE is characterized by focal seizures that may persist or be repetitive for at least 30 minutes without impairment of consciousness. When this condition lasts for hours or days, it is termed epilepsia partialis continua [Cockerell et al., 1996; Takahashi et al., 1997]. Absence, or petit mal, status also has been referred to as spike-wave stupor. This type of nonconvulsive SE may be extremely difficult to differentiate from complex partial SE without the aid of an EEG. Classically, features of absence status include a continuous alteration of consciousness without the cyclic variations seen with complex partial SE [Grin and DiMario, 1998]. The EEG recording exhibits prolonged, sometimes continuous, generalized synchronous 3-Hz spike-and-wave complexes, rather than focal ictal discharges, which characterize partial SE [Porter and Penry, 1983; Treiman, 1993]. Absence status does not appear to cause permanent neurological damage [Drislane, 1999]. The child presenting with a prolonged confused state, with a fluctuating level of consciousness or with prolonged unconsciousness, may require both clinical and EEG evaluations in addition to other studies.

Nonconvulsive SE is most likely not rare but simply underdiagnosed. A special category of nonconvulsive SE is subtle SE [Treiman, 1990, 1993]. These patients have severe encephalopathies stemming from a variety of intracranial processes or prolonged uncontrolled convulsive seizures. This type of SE is manifested clinically by the occurrence of mild motor movements, such as nystagmus, or by clonic twitches, which may be unilateral and are intermittent, brief, and without a true sequential pattern. These subtle movements are associated with marked impairment of consciousness, usually with continuous bilateral EEG ictal patterns. These continuing electrographic seizures that are not accompanied by clinical manifestations demonstrate a true “electroclinical dissociation,” and may be seen not only in neonates but also in severely ill children and adults [Mizrahi and Kellaway, 1987; Scher et al., 1993a]. The EEG progressively becomes uniform to produce a pattern of continuous ictal discharges, which then becomes interrupted by periods of relative flatness and then severe cortical depression. In a study of children, nonconvulsive SE most commonly followed a bout of convulsive SE or briefer convulsive seizure but with prolonged alteration in mental status [Tay et al., 2006], and is more likely to occur in children with a prior history of epilepsy [Abend et al., 2007] and remote risk factors for seizures [Classen et al., 2004]. In general, these patients respond poorly to traditional treatment with antiepileptics.

Febrile SE, which is unique to children, represents the extreme end of the complex febrile seizure spectrum. Febrile SE has long been suspected of having a relationship to the development of mesial temporal sclerosis. Patients with intractable temporal lobe epilepsy and mesial temporal sclerosis often have histories of severe febrile convulsions as infants. Diagnostic advances made possible by magnetic resonance imaging (MRI) have shown that very prolonged febrile convulsions may produce hippocampal injury [Lewis, 1999]. More recent studies support the link to the development of mesial temporal sclerosis [Provenzale et al., 2008]. Neuroimaging studies generally show hippocampal swelling during the acute stage [Scott et al., 2002, 2006]. A prospective study investigating long-term outcome in febrile SE found that most were partial (67 percent), and SE was unrecognized in the emergency department about a third of the time [Shinnar et al., 2008]. Human herpesvirus-6B appears to be the most common cause of febrile SE and may play an important pathogenic role in the etiology of mesial temporal lobe epilepsy [Theodore et al., 2008].


It is projected that between 102,000 and 152,000 events occur in the United States annually, an incidence 2–2.5 times greater than that previously proposed by Hauser [1990], who reported that SE occurs annually in 50,000–60,000 persons in the United States [DeLorenzo et al., 1996; Hauser et al., 1990]. More recent work continues to demonstrate a high incidence of SE varying between 20 and 41 patients per year per 100,000 population [Govoni et al., 2008; Chin et al., 2006; Coeytaux et al., 2000].

Approximately one-third of cases manifest as the initial seizure of a developing epilepsy, one-third occur in patients with previously established epilepsy, and one-third occur as the result of an acute isolated brain insult. Among those previously diagnosed as having epilepsy, estimates of SE occurrence range from 0.5 to 6.6 percent [DeLorenzo et al., 1996].

Hauser reported that up to 70 percent of children who have epilepsy that begins before the age of 1 year will experience an episode of SE [Hauser, 1990]. Also, within 5 years of the initial diagnosis of epilepsy, 20 percent of all patients will experience an episode of SE. Although the subsequent development of epilepsy is likely in adults with SE as their first unprovoked seizure [Hauser et al., 1990], a prospective study of children with SE found only a 0.3 probability plotted on a Kaplan–Meier curve that epilepsy will develop after ≥9 months in those who initially presented with SE [Maytal et al., 1989].

Among children, SE is most common in infants and young toddlers, with more than 50 percent of cases of SE occurring in children younger than 3 years of age [Shinnar et al., 1995]. In a study in Richmond, Virginia, total SE events and incidence per 100,000 population per year demonstrated a bimodal distribution, with the highest values during the first year of life and during the decades beyond 60 years of age [DeLorenzo et al., 1992, 1995, 1996]. Infants younger than 1 year of age represent a subgroup of children with the highest incidence of SE, whether events, total incidents, or recurrences are counted. The recurrence rate for SE in the Richmond study was 10.8 percent [DeLorenzo et al., 1996], but 38 percent of patients younger than 4 years of age had repeat episodes. Children have a much lower mortality rate than adults after adequate treatment [Dunn, 1988; Maytal et al., 1989; Pellock, 1993b; Phillips and Shanahan, 1989; Shinnar et al., 1995]. Age, etiology, and duration directly correlate with mortality [DeLorenzo et al., 1996; Towne et al., 1994].


SE usually is a manifestation of an acute precipitating event that affects the central nervous system (CNS) or is an exacerbation of symptomatic epilepsy. Less than 10 percent of cases of SE in adults and children are truly idiopathic in that no precipitating or associated cause can be identified [DeLorenzo et al., 1995, 1996]. Acute symptomatic causes are those most commonly associated with prolonged SE lasting for longer than 1 hour [DeLorenzo et al., 1996]. Thus, a full evaluation for etiology must be undertaken in every case of SE [Dodson et al., 1993; Pellock, 1994]. In patients with pre-existing epilepsy, a precipitating or associated factor may be clearly identified. Identification of this factor may help in treating the episode of SE, preventing further consequences of SE, and perhaps preventing future recurrences.

Although, typically, a precipitant to SE can be identified, a genetic susceptibility may predispose certain persons to develop prolonged seizures in response to an acute insult. Recent work in twins demonstrated a higher incidence in monozygotic twins than in dizygotic twins, providing evidence for a genetic contribution to the risk for SE [Corey et al., 2004]. Seizure type and specific epilepsy syndrome may differ between monozygotic twins, however, and SE may not be a function of the seizure type or syndrome experienced by each person.

A clear difference between causative disorders of SE in adults (Table 58-1) and those in children (Table 58-2) has been identified [Pellock and DeLorenzo, 1997; Riviello et al., 2006]. A major cause of SE in children is that associated with fever secondary to non-CNS infections, an etiology that essentially does not exist in adults. Inadequate antiepileptic drug levels and remote causes, including congenital malformations, also account for a significant number of episodes of SE in children, although some studies have found that patients may have had reasonable drug levels when SE occurred [Maytal et al., 1996]. Of note, many patients with subtherapeutic levels of antiepileptic drugs closely followed the instructions of their physicians and recently had drug-dosage alterations.

Table 58-1 Cause and Mortality Data for Status Epilepticus in Adults

Etiologic Disorder/Condition % of Cases Mortality Rate (%)
Anoxia 5 71
Hypoxia 13 53
Cerebrovascular accident 22 33
Hemorrhage 1 0
Tumor 7 30
Infection 7 10
Central nervous system infection 3 0
Metabolic 15 30
Low antiepileptic drug level 34 4
Drug overdose 3 25
Alcohol withdrawal 13 20
Trauma 3 25
Remote 25 14
Idiopathic 3 25

(From Pellock JM, DeLorenzo RJ. SE. In: Porter RJ, Chadwick D, eds. The epilepsies 2. Boston: Butterworth–Heinemann, 1997;267.)

Table 58-2 Etiologies of Status Epilepticus in Children

Etiology Percentage
Remote symptomatic epilepsy 33%
Acute symptomatic seizures 26%
Febrile 22%
Cryptogenic 15%
Central nervous system infection 13%
Acute metabolic disorders 6%

(From Riviello JJ et al. Practice Parameter: Diagnostic assessment of the child with status epielpticus [an evidence based review]. Neurology 2006;67:1542.)

The distribution of causes associated with SE in children is highly age-dependent [Shinnar et al., 1995]. More than 80 percent of children younger than 2 years of age have SE from a febrile or acute symptomatic cause, whereas cryptogenic or remote symptomatic causes were more common in older children. By contrast, in adults, subtherapeutic levels of antiepileptic drugs, remote causes, and cerebrovascular accidents represent the three most common causes of SE [Pellock and DeLorenzo, 1997]. In adults, SE resulting from remote causes occurred primarily in relation to stroke, so that both acute and previously occurring strokes account for a significant proportion of adult episodes of SE. Stroke is the etiologic disorder in approximately 10 percent of childhood SE episodes, either as the primary acute cause or as a remote event.

Recurrence rates for SE are age-specific, as illustrated in Figure 58-1. Repeat occurrences are much more common in children younger than 1 year of age. In the prospective Richmond study of SE, pediatric, adult, and elderly recurrence rates were 35 percent, 7 percent, and 10 percent, respectively [DeLorenzo et al., 1996]. Recurring SE is more frequent in children with remote symptomatic encephalopathy or progressive degenerative disease [DeLorenzo et al., 1995, 1996]. The extent of clinical and laboratory evaluation that should be performed in each child with recurrent SE will depend on the presentation, signs and symptoms, and underlying medical condition of the patient. For example, a child with recurring SE who has an indwelling shunt for hydrocephalus will almost always need to be evaluated for the possibility of shunt malfunction or infection.

Recently, CNS or systemic autoimmune disorders have been recognized as causes of SE [Shorvon and Tan, 2009]. In some cases, the diseases may be paraneoplastic, but for many, no cause has been determined. Since patients with these etiologies usually are described in case reports or small series, the contribution of the disorders to the epidemiology of SE is difficult to assess, but they appear to represent a fraction of the cases previously diagnosed as infectious, since the patients often have a syndrome qualifying as an encephalitis.

A recently described entity with seizures and persistently altered mentation, including a fluctuating level of consciousness or with prolonged unconsciousness, is anti-NMDA-receptor encephalitis [Dalmau et al., 2008]. Anti-NMDA-receptor encephalitis is a disorder with antibodies against the NR1 subunit of the receptor. This disorder largely affects young people, and its diagnosis is facilitated by the characteristic clinical picture that develops in association with cerebrospinal fluid pleocytosis. Only about half the patients had MRI abnormalities. Recovery from this disorder is typically slow, and symptoms may relapse. The mainstay of treatment is immunotherapy with steroids, plasma exchange, or intravenous immunoglobulin. These immunotherapies may be given individually or in combination [Dalmau et al., 2008]. Other possible autoimmune illnesses include systemic lupus erythematosus, antineuronal antibody syndromes with limbic encephalitis, limbic encephalitis following various systemic viral infections, and Hashimoto’s encephalopathy (autoimmune thyroid encephalopathy).

Management and Therapy


SE is a neurologic and medical emergency. Therapy includes maintenance of respiration, general medical support, and specific treatment aimed at stopping both electrographic and clinical seizures while the cause of the event is investigated [Epilepsy Foundation of America, 1993]. Prompt diagnosis and management provide the best outcome [DeLorenzo, 1990; Pellock, 1993a, 1993b, 1994; Pellock and DeLorenzo, 1997; Rider and Thapa, 1995; Treiman, 1993].

A single generalized convulsion in a child with a prolonged period of impaired consciousness is much more difficult to diagnose, and an EEG should be obtained urgently. If on-going ictal discharges or electrographic seizures are noted, the patient should be considered to be in electrographic SE, and prompt treatment is indicated. The goals of SE emergency management are listed in Box 58-2. Clinical and electrographic seizure activity must be rapidly terminated while ensuring optimal oxygenation and metabolic balance. Both clinical and electrical seizure activity should be terminated as soon as possible [Treiman, 1990]. The longer an episode of SE continues, the more likely it is to result in permanent neurologic damage and to become refractory to treatment [DeLorenzo, 1990; DeLorenzo et al., 1996; Hauser, 1990; Pellock and DeLorenzo, 1997; Towne et al., 1994; Treiman, 1990; Treiman et al., 1992, 1994].

Rapid initiation of care is essential to ensure the best possible outcome. Increased seizure duration commonly is regarded as an important factor contributing to increased morbidity and mortality. Prolonged seizures of any type are associated with an increased risk of complications [Lowenstein et al., 1999]. In a study in adults and children, the mortality rate in SE was 34.8 percent when seizures lasted longer than 1 hour, compared with 3.7 percent when seizure duration was less than 1 hour [DeLorenzo et al., 1992]. Therefore, limiting the time from seizure onset to initial treatment and attainment of control is essential to minimize the complications of prolonged seizures. In a retrospective study, patients with SE may not arrive at the emergency department for 30 minutes or longer, and treatment initiation may not occur for 40–263 minutes, thereby increasing risk for prolonged seizures [Jordan, 1994]. The Veterans Affairs Cooperative Study reported that the mean delay before treatment of SE was 2.8 hours in generalized convulsive SE and 5.8 hours in subtle SE; this study involved exclusively adult patients [Treiman et al., 1998]. In a prospective study of 889 patients (625 adults, 264 children), 41.5 percent received treatment within 30 minutes and 70.9 percent within 60 minutes of seizure onset; 18.1 percent did not have treatment initiated until after 90 minutes of seizures. This latter group of patients, however, included patients in coma diagnosed with subtle or nonconvulsive SE [Pellock et al., 2004].