Definition

Status epilepticus (SE) is one of the few illnesses named not by the physicians who treated it but by the patients who suffered from it. The expression état de mal was coined by the patients at the Salpêtrière, the world's first neurologic hospital, and reached the medical literature through the doctoral thesis of Louis Calmeil (6), then became latinized as status epilepticus in Bazire's English translation of the 1867 lectures of Armand Trousseau in London (44). Although Trousseau and Bourneville (5) recognized the existence of stages in SE, and Clark and Prout (9) identified SE as being different from single epileptic seizures, the first attempt at a formal definition came at the 1962 Marseilles Colloquium, where Gastaut defined it as “epileptic seizures which are so frequently repeated and so prolonged as to create a fixed and enduring epileptic condition” (15). The definition and classification of SE were further refined by Gastaut in the World Health Organization's 1973 publication, Dictionary of Epilepsy (18), in the 1974 Handbook of Clinical Neurology (35), and in the 1975 Handbook of Electroencephalography and Clinical Neurophysiology (20). Gastaut also proposed a formal classification of status epilepticus (Table 2.1) at the first Santa Monica meeting, in 1980 (12, 16).

Conceptual Basis for the Definition of SE

Is SE simply a cluster of severe seizures, or is it a separate condition with its own unique pathophysiology? Trousseau expressed the uniqueness of SE as early as 1867: “In that form of status epilepticus when the convulsions are practically continuous, something specific happens which demands an explanation” (44). Indeed, recent experimental evidence strongly suggests that SE is a separate phenomenon and not simply a series of seizures. Seizure-like stimulation of some excitatory pathways in the brain easily triggers SE. Once it is established, stimulation can be stopped and seizures continue. Moreover, these self-sustaining seizures can be suppressed for hours with a synaptic blocker, yet when that blocker's effects wane, the seizures return in the absence of any further stimulation, implying that self-sustaining SE is maintained by an underlying change in excitability (31). The transition from serial seizures to SE is modulated by neuropeptides, neurotransmitters, and receptor trafficking (29–31), and anticonvulsants that are effective against serial seizures are often ineffective against established, self-sustaining SE (32). At the second Santa Monica meeting (37) there was a vigorous debate over the basis for defining and classifying SE. Many participants suggested definitions based on seizure durations ranging from 5 to 30 minutes. Others defined SE by the subject's failure to recover consciousness before seizure recurrence. This criterion is retained in our definition, since it permits the inclusion of cases in which severe seizures are associated with long interictal intervals. Engel has proposed that SE is defined by the failure of normal mechanisms to terminate a seizure (14). A logical conclusion of that premise would be to define SE statistically, as a seizure duration that is clearly outside the range of “normal” seizure duration, for example 5 standard deviations (SD) removed from the mean. The definition of SE proposed later in the chapter is in accord with that line of thought.

The Concept of “Impending SE”

Ever since Gastaut defined SE as a fixed and enduring epileptic condition, the medical literature has struggled to determine the minimal duration and severity of seizures that constitute SE. Simple as it may be, such a determination is essential to clinical trials and therapeutic guidelines. Some of the dilemmas created by these attempts at a more precise definition are discussed in Shorvon's outstanding book (36) and are beyond the scope of this chapter. Yet we must recognize that seizures are rarely fixed, that the boundaries of what constitutes a seizure vary with the method of observation (e.g., electroencephalographic [EEG] or clinical), and that the most useful definition may vary with our main objective. For example, many epileptologists apply the treatment of status when faced with the admittedly unusual event of continuous generalized motor seizures lasting 5 minutes, which suggests that a seizure duration of 5 minutes meets their operational definition of SE. Reducing the definition of status to that extent, however, would enormously complicate epidemiologic studies by including in population studies a large number of patients who may not experience full-blown SE.

Clinical Basis for a Definition Based on Seizure Duration

The duration of what is accepted as SE has been shrinking progressively, from 30 minutes in the guidelines of the Epilepsy Foundation of America's Working Group on Status Epilepticus (51) to 20 minutes (2), and to 10 minutes in the Veterans Affairs (VA) cooperative trial on the treatment of SE (42). Most recently, Wasterlain (at the 1997 Santa Monica meeting), Lowenstein and Alldredge (25), Lowenstein, Bleck, and Macdonald (26), and Meldrum (33) have proposed an operational definition of SE that defines the time when severe seizures should be treated as SE. Those several authors have proposed that 5 minutes of continuous generalized convulsive seizures is sufficient to fulfill that criterion. Videotape-telemetry studies show that the mean duration of generalized convulsive seizures in adults ranges from 62.2 seconds (n = 120) to 52.9 ± 14 seconds (n = 50) (very close to the 1-minute estimate of Gastaut and Broughton [19]) for the behavioral manifestations and averages 59.9 ± 12 seconds for the EEG manifestations. None of those seizures lasted 2 minutes (21, 39). Therefore, the operational definition of SE as 5 minutes of continuous generalized convulsive activity is probably too conservative, since it defines status by a seizure duration that is 18–20 SD removed from the norm, restricting it to an extremely rare event. It might be more logical to treat with intravenous (IV) drugs after 2 minutes of continuous seizure activity (4–5 SD outside the norm), as proposed by Theodore and colleagues (39). However, in practice, there is essentially no difference between 2 and 5 minutes, since it takes more than 3 minutes to deliver the first IV injection. A definition of 5 minutes of continuous seizures has two advantages: first, it reconciles the definition of status with the almost universal emergency room practice of treating those patients as if they had SE, and second, it places the definition of status far outside the norm for seizure duration, clearly indicating that something distinctly unusual and severe is occurring. However, we propose to call it impending status epilepticus, since not all such patients are in SE, and a significant proportion will stop seizing spontaneously in the next few minutes. The Richmond data provide support for this concept: more than 40% of the seizures lasting from 10 to 29 minutes stopped spontaneously without treatment (32), and they had an overall mortality of 2.6% versus 19% for SE (P < 0.001). We define impending status epilepticus as “continuous seizures lasting more than 5 minutes, or intermittent clinical or electrographic seizures lasting over 15 minutes without full recovery of consciousness between seizures.” This recognizes the need to treat those patients intravenously with high-dose anticonvulsants, since their risk of developing frank SE is high, but it also acknowledges that not all of those patients are in frank SE. By adopting a new category of impending SE, these patients will receive proper urgent medical care but will not contaminate morbidity and mortality statistics, outcome measures, or clinical trials with a subpopulation that is not in frank SE. Frank SE continues to use the definition of the EFA's working group on SE. We also agree with Shorvon (36) and others that different definitions and classifications are needed for different age groups. The current definition should apply to adults and children over 5 years of age. Different criteria should be used for neonates, and yet other criteria for infants and young children. We will not discuss in this chapter the definition of SE in the neonate (for which we refer the reader to Shorvon's 1994 book [36] and to Wasterlain and Vert's book, Neonatal Seizures [48]), or the definition of SE in infants and children. In that age group, the high incidence of febrile seizures, which frequently last longer than 5 minutes (22), and their generally benign outcome may change the risk-benefit ratio of status versus treatment, so that most physicians have retained the traditional definition of SE as seizures lasting more than 30 minutes.

Experimental Basis for a Definition Based on Seizure Duration

The traditional argument that SE should be defined as the minimal duration of seizure that produces brain damage has collapsed with the demonstration (7) that even single seizures without a tonic component can produce neuronal loss in experimental animals. We know that generalized convulsive seizures in primates produce neuronal death much more quickly than nonconvulsive seizures (34), and therefore deserve prompt therapeutic attention, but we have no solid data on the minimum seizure duration sufficient to damage the human brain. On the other hand, the demonstration in animal experiments that repeated, brief, seizure-like discharges through excitatory pathways set in motion self-sustaining seizures (32, 45, 47), which rapidly become resistant to standard anticonvulsants (29), would suggest that treatment should be administered before self-sustaining seizures become established. In the rat, this can take place within 15 minutes (32) or 10 afterdischarges (45). Unfortunately, we have little information on the timing or even the existence of these phenomena in humans. Therefore, experimental data give us a sense of urgency without providing us with a precise time frame for defining or treating SE.

Is There a Risk of Overtreatment?

Intravenous medications undoubtedly entail significant risks: respiratory and even cardiovascular depression can result, some anticonvulsants impair cardiac conduction and can generate arrhythmias, allergic reactions often take the form of anaphylactic shock, idiosyncratic reactions may be more severe than with other forms of administration. Therefore, single generalized seizures that have not been documented to last more than 2 minutes and seizures that are benign or remain focal should not be treated with IV anticonvulsants; oral loading is usually feasible and much safer.

A Definition of Impending SE

We propose the following definition, which should be used by clinicians making diagnostic and therapeutic decisions:

An acute epileptic condition characterized by continuous generalized convulsive seizures for at least 5 minutes, or by continuous nonconvulsive seizures (clinical or electrographic) or focal seizures for at least 15 minutes, or by two seizures without full recovery of consciousness between them.

A Definition of SE

We propose the following definition of SE, which should be used by epidemiologists and other clinical investigators conducting studies of SE:

An acute epileptic condition characterized by continuous seizures (partial or generalized, convulsive or nonconvulsive) for at least 30 minutes, or by 30 minutes of intermittent seizures without full recovery of consciousness between seizures.