Convulsive SE is the most common form of status, but its real frequency is difficult to document, precisely because the various types of status are not specified separately in most published series. The proportions quoted for convulsive status range from 1.3% to 16% of all patients with epilepsy (Hauser, 1983). Contrary to the situation that prevails in adults, status is often the first epileptic seizure in infants and children. This was the case for 184 (77%) of the 239 patients of Aicardi and Chevrie (1970). Similarly, 24% of all children with a first afebrile seizure before 10 years of age who were seen in Minneapolis (Hauser, 1983) presented with SE. Status is particularly common during the first 2 years of life, and approximately 75% to 85% of cases occur before the age of 5 years (Phillips and Shanahan, 1989; Aicardi and Chevrie, 1970). In children older than 3 years of age, the annual frequency remains stable at around 3% to 5% until the age of 15 years (Aicardi and Chevrie, 1970). The overall incidence of SE does not appear to have declined. A recent prospective population-based study (DeLorenzo et al., 1995) from Richmond, Virginia, evaluated the incidence of SE at 41 patients per year per 100,000 population. Total SE events and the incidence per 100,000 individuals per year showed a bimodal distribution, with the highest rates being observed during the first year of life and after 60 years of age (DeLorenzo et al., 1996).

However, the frequency of different causes of status appears to have changed during the past 30 years. In 1971, Aicardi and Chevrie noted 126 symptomatic cases and 113 cryptogenic ones, with half of these being associated with fever. They did not find any essential difference between the patients with febrile seizures and those with status, except for the duration of the attack. Thus, because of their frequency, febrile convulsions were probably a major cause of cryptogenic SE in children younger than 3 years of age. In this series, the proportion of cases due to changes in or interruption of treatment was not given. Although precise figures are not available, most investigators agree that drug withdrawal is a common cause of status in children with epilepsy (Delgado-Escueta et al., 1983b).

Symptomatic causes are more common in recent series, especially acute encephalopathies, whereas afebrile idiopathic cases and chronic encephalopathies appear less frequently; febrile cases remain the most common form. Lesional causes are particularly common in younger children (Chevrie and Aicardi, 1977). They were responsible for 69% of the cases in one series (Aubourg et al., 1985). Overall, approximately one-fourth of cases are afebrile idiopathic, one-fourth are febrile, and the rest are almost equally divided between chronic (remote symptomatic) and acute encephalopathies (Shinnar et al., 1992; Maytal and Shinnar, 1990; Dunn, 1988). Specific causes in
four large series are shown in Table 16.1. Brain tumors are the cause of status in children only on rare occasions. The preponderance of frontal lesions that are mentioned in most adult series is not evident in children. Toxic causes, especially theophylline toxicity, should always be considered (Dunn and Parekh, 1991). The recent study released by the Richmond group (DeLorenzo et al., 1996) highlights the marked differences in etiologies between the adult and pediatric populations. For example, SE related to remote etiologies was considered a cause in 27% of the pediatric patients and in only 16% of the adult patients. The percentages for noncompliance to antiepileptic drugs (AEDs) (15% in children and 22% in adults) and especially of infection (37% in children versus 5% of all etiologies in adults) also differed.

Precipitating factors in patients with previous epilepsy include intercurrent infections; sleep deprivation; and, especially, anticonvulsant drug withdrawal. In a Finnish study (Sillanpää and Shinnar, 2002), status occurred in 44% of patients with remote symptomatic epilepsy and in 20% of those with idiopathic epilepsy.

Convulsive SE is sometimes preceded by serial seizures separated by intervals of recovery of consciousness, especially in patients with known epilepsy. During this period of serial seizures, prompt treatment can prevent the emergence of true status (Shorvon, 1994).

Convulsive status proper may present either as a series of generalized tonic-clonic seizures without intervening recovery or as a more or less continuous seizure, which usually is purely clonic in nature. The latter type is especially common in children, and it was noted in 40% and 80%, respectively, of the individuals studied in two pediatric series (Congdon and Forsythe, 1980; Aicardi and Chevrie, 1970). In the cases presenting as a series of convulsions, generalized tonic-clonic seizures last 1 to 3 minutes. They tend to become briefer as time elapses, and the clonic phase may disappear or become inconspicuous (Roger et al., 1974). Interictally, the patients are in a coma, with vegetative disturbances, such as salivation, bradypnea, cyanosis, and arterial hypotension. Circulatory collapse may be the cause of 20% of deaths. The EEG displays typical tonic-clonic seizures (Gastaut and Tassinari, 1975a, 1975b). Interictally, the tracings show slow arrhythmic delta waves or, less commonly, faster rhythms in the alpha or beta range (Roger et al., 1974). The interictal neurologic signs may include unilateral or bilateral Babinski responses. The possibility of cerebrospinal fluid (CSF) pleocytosis as a consequence of status has been mentioned in adults (Edwards et al., 1983;
Schmidley and Simon, 1981) and, more recently, in infants and children (Woody et al., 1988b; Aicardi and Chevrie, 1970).

Clonic seizures are often predominantly unilateral, or they may shift from side to side (seesaw seizures) (Roger et al., 1974). Approximately 75% of patients with unilateral clonic seizures are 3 years of age or younger (Chevrie and Aicardi, 1975). Clonic seizures tend to be continuous and long-lasting rather than to consist of repeated attacks. They may last hours or even days, waxing and waning in intensity. They may involve an entire half of the body, but, at some periods, they may be restricted to one limb or segment; at other times, they may involve part of the contralateral side or they may transiently become generalized. The rhythm of the jerks is variable. Frequently, the jerks are at different rhythms in the various segments affected (Gastaut et al., 1974c). A postictal hemiplegia that may be persistent is usually observed. The first manifestation of the hemiconvulsion-hemiplegia-epilepsy syndrome (see Chapter 10) is a sudden prolonged hemiconvulsion in the form of status (Arzimanoglou and Dravet, 2001). The individual’s consciousness is variably affected, and it may be preserved during a part or the whole duration of an episode. The EEG during clonic status typically shows more or less rhythmic slow waves of higher amplitude over the hemisphere opposite the convulsing side that are associated with a predominantly posterior rapid rhythm at about 10 Hz (the epileptic recruiting rhythm). The elements of fast and slow rhythms can variously combine, and they may produce more or less typical spike-wave complexes (Gastaut et al., 1957, 1974c). Flattening of the tracing or arrhythmic slow waves are usual in the postictal phase (Gastaut et al., 1960). After prolonged convulsive activity, a suppression-burst tracing that has a serious prognostic significance may be observed.

Treiman et al. (1980) distinguished five successive phases in adults. After the establishment of continuous discharges (the first three phases), brief episodes of low-amplitude tracing appear, first sporadically and then more and more frequently, ending with periodic epileptic discharges on a “flat” background. Periodic epileptiform discharges (PEDs) are also observed in children. Autonomic involvement is generally less pronounced than with tonic-clonic seizures, but fever, respiratory embarrassment, and a decrease in blood pressure often occur with long-lasting status.

The outcome of convulsive SE has improved considerably over the past three decades. The death rate ranges between 6.6% and 17% in most adult series (Hauser, 1983; Aminoff and Simon, 1980). The mortality rate of the Richmond study was 22% overall (DeLorenzo et al., 1996). However, the mortality rate in children is only 3% in some series, and most pediatric deaths occur between the ages of 1 and 4 years (Leszczyszyn and Pellock, 2001). Aicardi and Chevrie (1970) found a mortality rate of 11% in a series of 239 children younger than 15 years. In other series, the mortality has varied between 3.6% and 7% (Maytal and Shinnar, 1990; Phillips and Shanahan, 1989; Dunn, 1988; Vigevano, 1986), except in two series (Aubourg et al., 1985; Chevrie and Aicardi, 1977) of children younger than 1 year, in which it reached 25%. Most deaths are caused by the underlying disorder. Death during status is associated with respiratory or cardiac arrest (Hauser, 1983; Tassinari et al., 1983).

Long-term mortality after a first episode of afebrile status was recently evaluated in a population-based retrospective cohort study in the Rochester Epidemiology Project Records (Logroscino et al., 2002). Cases surviving the first 30 days (145 out of 184) were followed until death or study termination. In this study, 19 (13%) were younger than 1 year, and 35 (24%) were between the ages of 1 and 19 years old. At 10 years, the cumulative mortality among 30-day survivors was 43% (62 deaths), 76% for patients 65 years or older, but only 5% and 2% in patients younger than 1 year and in the age group of 1 to 19 year olds, respectively. Using generalized SE as the reference group, only those with myoclonic SE had a significantly higher mortality. Within etiologic subgroups, the highest mortality rate was seen in the acute symptomatic group (27 of 66 patients) and the lowest in the idiopathic or cryptogenic group (8 of 28 patients). The mortality rate was 50% in the group that experienced SE for more than 24 hours, compared with 28% for those in whom SE lasted for less than 2 hours. These observations extend the findings of an earlier report by the same group on short-term mortality (Logroscino et al., 1997), which showed that the highest mortality was for acute symptomatic SE; similar studies (Towne et al., 1994; Lhatoo et al., 2001) demonstrate that the prognosis and outcome are mainly determined by the nature of the acute or progressive neurologic or systemic disorders that precipitate the status.

Neurologic sequelae were found in 88 of 239 children and infants with convulsive status (Aicardi and Chevrie, 1970). In 47 patients, the sequelae apparently were acquired at the time of status (20%). The neurologic syndromes encountered included diplegia,
extrapyramidal syndromes, cerebellar syndromes, and decorticate rigidity. Hemiplegia, which was found in 28 patients, was always acquired at the time of status, the so-called hemiconvulsion-hemiplegia syndrome (see Chapter 10). These figures are difficult to interpret because the sequelae may represent complications of the causal disease rather than of the status itself, or they may have antedated the convulsive episode. Of 59 patients with idiopathic status seen at the time of the initial episode and followed for at least 1 year, 12 (20%) had neurologic sequelae (Aicardi and Chevrie, 1983). In the same series, mental deficits were noted after the status in 114 (48%) of 239 patients; they apparently were acquired at the time of status in 79 (33%) of all patients and in 14 (24%) of those with cryptogenic status. Similarly, Fujiwara et al. (1979) found mental and/or neurologic sequelae in 40 of 79 children with status lasting 1 hour or more, 25 of whom had hemiplegia. Mental and neurologic deficits are often associated with each another and with the epilepsy. In the series of Aicardi and Chevrie (1970), epilepsy was present after the episode of status in 44% of patients, compared with an incidence of 23% with epilepsy before the status. Furthermore, seizures occurring after SE were mainly of a type that is usually associated with brain damage (Fujiwara et al., 1979; Aicardi and Chevrie, 1970).

The outcome in symptomatic cases is significantly poorer than that in cryptogenic cases (Aicardi and Chevrie, 1983). Other unfavorable factors include young age; long duration of status; and, possibly, female gender (Chevrie and Aicardi, 1978). Sequelae of cryptogenic status were found in 35% of children younger than 3 years and in only 9% of those older than 3 years. In infants younger than 1 year, Chevrie and Aicardi (1978) found neurologic abnormalities and severe mental retardation in 8 (40%) of 20 patients with symptomatic SE and in 4 (27%) of 15 infants with idiopathic SE. However, the significance of such figures is uncertain because most reports originate from specialized hospitals and referral centers. A lower proportion of severe cases would be expected in population studies.

The proportion of cases with neurologic or mental sequelae has been considerably less in later series (Shinnar et al., 1992; Maytal and Shinnar, 1990; Phillips and Shanahan, 1989; Dunn, 1988; Vigevano, 1986), and most residua now appear to be due to the underlying disorder. The incidence of hemiconvulsion-hemiplegia syndrome, which was once common, has decreased in industrialized countries (Arzimanoglou and Dravet, 2001; Roger et al., 1982).

Similarly, the risk of recurrent status following a first episode of status has decreased. According to Shinnar et al. (1992, 1997), new episodes occur mainly in symptomatic cases and in children with abnormal neurologic signs that predate the first status; these latter children account for 88% of the recurrences. Idiopathic or febrile SE in normal children is largely an isolated event. The recurrence risk for status in idiopathic cases was only 4% for initial idiopathic status and 3% for febrile status, but this rose to 11% for cases of acute symptomatic status and to 44% for remote symptomatic cases.

The improvement in prognosis of convulsive SE is at least partly related to more effective, faster treatment. Cases of severe sequelae of apparently idiopathic status continue to be observed in developing countries, and infants or children with prolonged seizures of whatever cause should receive immediate and appropriate therapy because status remains one of the most urgent emergencies in child neurology.

Several mechanisms may be operative in the genesis of sequelae, with the main issue being whether the convulsive activity itself, independent of the cause of the seizures, is able to produce brain damage. Strong arguments supporting the hypothesis that prolonged seizures, irrespective of their cause, are able to produce brain lesions exist both experimentally (Meldrum, 1978, 1983b; Ben-Ari et al., 1979; Blennow et al., 1978) and clinically (Sagar and Oxbury, 1987; Soffer et al., 1986). However, some authors have been unable to produce brain damage in experimental models (Brown and Babb, 1983), and cases of extremely prolonged SE without sequelae have been recorded in humans. Residual damage following febrile convulsions has been associated with long-lasting and lateralized seizures and with young age at onset (Aicardi and Chevrie, 1976), whereas short bilateral convulsions have not been followed by any neurologic or mental abnormalities. However, long-lasting unilateral febrile seizures in young infants may be the result of acquired acute encephalopathies and/or preexisting unrecognized brain damage (Nelson and Ellenberg, 1978; Chevrie and Aicardi, 1975).

Some of the brain damage incurred during status is at least partly caused by complications of the seizures (e.g., hypoxia or vascular collapse) (Glaser, 1983; Scheuer, 1992), and it is therefore preventable. Some of the lesions may be related to the epileptic discharge itself, regardless of the complicating factors (Aicardi and Chevrie, 1983). The significance of the neuropathologic lesions found in the brains of patients who have had SE remains controversial. Some authors (Roger et al., 1974; Radermecker et al., 1967) have noted that the lack of correlation among the duration, location, and intensity of seizures seen clinically with
the extent of pathologic damage argues against the causal role of status. Some of the discrepancy may result from the different ages of the patients studied. Although Corsellis and Bruton (1983) were only able to find acute lesions attributable to status (other than causal ones) in 1 of 12 brains of adults who died following status, they could demonstrate acute damage in the brains of all 6 children younger than 3 years who died shortly after an episode of status.

Tonic status is less common than tonic-clonic or clonic status. It occurs exclusively in children and adolescents with previous epilepsy, particularly in patients with Lennox-Gastaut syndrome (LGS) (Roger et al., 1974), most of whom have some degree of mental retardation. However, tonic status also may be observed in patients with epilepsy syndromes intermediate between a typical absence epilepsy and LGS. In some cases (Devinsky et al., 1991; Alvarez et al., 1981; Bittencourt and Richens, 1981), tonic status was precipitated by the intravenous injection of diazepam or clonazepam (Alvarez et al., 1981) for the treatment of absence status in children with LGS (Tassinari et al., 1972a). Oral clonazepam also can precipitate tonic status (Bourgeois and Wad, 1988).

Several episodes of tonic status commonly occur in the same patient. The duration of tonic status may be much longer than that of other convulsive types. In a series of 28 patients by Gastaut et al. (1967), the average duration of the episodes was 9 days. The serial tonic seizures are initially typical, although they tend to last longer (70 seconds) on average than do isolated tonic seizures (15 seconds). With the repetition of the attacks, they tend to become less conspicuous, with lessening of the tonic phenomena but aggravation of autonomic manifestation, especially bronchial secretion (Roger et al., 1974). After several hours, the attacks are often limited to slight eye deviation, with respiratory irregularities and marked tracheobronchial obstruction resulting from hypersecretion.

In some patients, the attacks may be attenuated from the start (Somerville and Bruni, 1983), or they may even be subclinical. In such cases, only polygraphic recordings can demonstrate the tachycardia, changes in respiratory rhythm, and occasional hypertonus of the trunk or neck muscles. In most patients, consciousness is moderately to severely impaired. The EEG demonstrates the successive tonic seizures. Interictal tracings may show diffuse slow waves or bursts of delta or theta activity (Lugaresi and Pazzaglia, 1975). Not uncommonly, tonic status is combined with absence status. In such cases, the fast activity characteristic of tonic attacks periodically interrupts the interictal spike-wave complexes (Fig. 16.1).

Tonic status is a serious condition; 4 of the original 28 patients collected by Roger et al. died (1974). Death now appears uncommon (Tassinari et al., 1983), but tonic status is often followed by a period of confusional state that may last several days. Neurologic sequelae have not been reported following tonic status.

Myoclonic status, which is characterized by the incessant repetition of massive myoclonic jerks, is an uncommon condition (Gastaut, 1983; Roger et al., 1974). It may occur in patients with myoclonic epilepsy of adolescence or in children with a mixture of generalized tonic-clonic seizures and absences or massive myoclonias (Janz, 1991; Asconapé and Penry, 1984). In such cases, consciousness may be preserved despite continuous jerking for hours. The EEG in primary types of myoclonic status shows bursts of multiple spikes that are synchronous and symmetric over homologous regions of the scalp (Gastaut and Tassinari, 1975b; Gastaut et al., 1967). Myoclonic status also may be observed in patients with degenerative myoclonus epilepsy (see Chapter 7). In such cases, the myoclonus is often less regular, and it often alternates with brief atonic seizures, either generalized or limited to a single segment or group of muscles.