Those disorders that are characterized by episodic attacks involving brain functions are not epilepsy. The realizations that have been slowly reached are that all seizures marked by loss of consciousness, involuntary movements, or disturbances of sensorium are not of the same origin, that they may be the result of different mechanisms, and that they have widely different outcomes and implications. Although such paroxysmal events as syncopes, hysterical fits, sleepwalking, and others were undoubtedly confused with epileptic attacks in the past, they are now considered completely separate from epileptic seizures. The latter term applies only to a particular type of attack that can be attributed to a specific neurophysiologic mechanism. According to Jackson’s classic definition (Jackson, 1931), “epilepsy is the name for occasional, sudden, excessive, rapid, and local discharge of gray matter.” That definition has remained the basic foundation to the modern understanding of epileptic phenomena even though the clinical differentiation may be difficult.

Epileptic seizures or attacks are transient clinical events that result from the abnormal, excessive activity of a more or less extensive population of cerebral neurons. As used by Jackson, the term local may not be appropriate for those many cases in which instantaneous or very rapid involvement of large areas of brain is seen without any demonstrable evidence of a local discharge. Epileptic seizures are usually brief, lasting from seconds to minutes, and they are marked by the sudden appearance of behavioral manifestations that may be purely motor or that may affect other brain functions (Gastaut and Broughton, 1972; Penfield and Jasper, 1954). The abnormal and excessive neuronal activity at the origin of the epileptic seizures is inferred from both the clinical events that constitute the attacks and the electroencephalogram (EEG) concomitants of the clinical seizures. The clinical events are not always obviously excitatory. Many seizures are characterized to a greater or lesser degree by negative phenomena, such as the loss of awareness, muscle tone, or language. However, in all cases, an epileptic seizure results in a paroxysmal disorganization of one or several brain functions (Bancaud et al., 1965, 1973).

The EEG events are essential for the characterization of attacks as epileptic seizures as opposed to one of the other types of paroxysmal attacks. These events constitute the epileptic discharge (Gastaut, 1973), which is defined as a temporary paroxysmal change in EEG activity, whether diffuse or localized. The epileptic discharge usually can be recorded on the scalp, but, on occasion, recording may occur only from the surface of the brain or even from the buried convolutions and other deep structures (Bancaud et al., 1973). This EEG recording is the most direct evidence of the abnormal, excessive neuronal activity postulated by Jackson. The main characteristics of the epileptic discharge are its high amplitude and its rhythmicity, both of which are caused by the excessive synchronization of an abnormal number of potentials in a neuronal aggregate; these therefore imply dysfunction in a population of neurons (Aird et al., 1989; Fenwick, 1983). The epileptic discharge may be related to the abnormal behavior of individual neurons, known as the
paroxysmal depolarization shift (Prince and Connors, 1986), especially in focal seizures. In some models, a temporal and spatial correlation has been shown between the paroxysmal depolarization shift and the occurrence of EEG spikes on the cortex or the scalp (Aird et al., 1984). However, neither the paroxysmal depolarization shift nor the “epileptic spike” on the EEG can be equated with clinical epilepsy. The epileptic discharge is a complex phenomenon resulting from the interaction of excitatory and inhibitory influxes from a network formed by multiple diverse neuronal sources. In its most classic form, the epileptic discharge comprises an initial rapid activity of 10 to 20 Hz that increases progressively in amplitude, which is known as the epileptic recruiting rhythm (Gastaut et al., 1974a, 1974b; Gastaut and Broughton, 1972). As the seizure proceeds, the rapid rhythm is fragmented by the admixture of a slower rhythm—usually about 3 Hz in frequency—that progressively becomes more prominent until the discharge ends (Gastaut and Tassinari, 1975a). This latter rhythm is thought to reflect the activity of inhibitory elements that are within, around, or distant from the site of origin of the discharge. Many other patterns of epileptic discharges are possible (Gastaut and Tassinari, 1975a), depending on the type of epilepsy and the recording techniques. No pattern is general enough to apply to all of the types that are observed. These include fast rhythms of low amplitude or flattening of the EEG activity (desynchronization), rather than typical hypersynchronous discharges. In practice, the repertoire of EEG discharges that electroencephalographers and clinicians recognize as epileptic is limited (Blume, 1982). Theoretically, such discharges are a constant concomitant of epileptic seizures even though they are not necessarily detectable by conventional techniques of EEG recording in every case. Experience with depth electrodes and stereo-EEG has shown that discharges recorded from the depth of the brain do not necessarily appear on the overlying scalp or even on the nearby cortex (Bancaud and Talairach, 1975). Not all epileptic discharges, however, correspond to epileptic seizures (i.e., paroxysmal clinical events). Subjects who do not demonstrate any clinical evidence of paroxysmal brain dysfunction yet who have generalized (Eeg-Olofsson et al., 1971) or partial (Cavazzuti et al., 1980) paroxysmal abnormalities showing the typical EEG characteristics are not uncommon. In some such patients, minimal behavioral or cognitive changes can be demonstrated with the use of special tests performed at the time of the discharges, even if they are brief, despite the absence of an overt clinical seizure (Kasteleijn-Nolst Trenité et al., 1988; Aarts et al., 1984; Fenwick, 1983; Gloor, 1979). Some authors refer to such cases as subclinical seizures (Gastaut, 1973). For those cases in which the same techniques of observation fail to detect any clinical change, the authors of this text sometimes use the terms infraclinical or electrical seizure. Such terms should be avoided because a seizure, by definition, is a clinical phenomenon (Gastaut, 1973).

Epileptic seizures are often difficult to distinguish from other paroxysmal events such as syncope and hysteria attacks. In theory, the distinction implies the presence or absence of an epileptic discharge. The epileptic discharge, however, is not recorded in most cases because seizures rarely occur at the time of the EEG recording. In addition, the discharge may be difficult to recognize, it may be obscured by muscular artifacts, or it may be altogether absent on the scalp. Such difficulties limit the practical usefulness of the epileptic discharge as a diagnostic tool. Even from a theoretical standpoint, the differentiation of an epileptic discharge from other paroxysmal activities may be difficult. For example, tonic seizures occurring with acute anoxia usually are interpreted as a release phenomenon that results from the interruption of the inhibitory influences from the cortex, which is more sensitive than the brainstem reticular formation to the lack of oxygen (Stephenson, 1990; Gastaut, 1974; Lombroso and Lerman, 1967). In such a situation, one must ask what the nature of the excessive activity of the brainstem neurons responsible for tonic contraction is. Could this activity be considered an excessive discharge in the gray matter? Other such examples could be given (e.g., the paroxysmal dyskinesias), but debate on that problem is beyond the scope of this book. However, the reader should understand that reasonable agreement does exist on which paroxysmal events should be regarded as epileptic seizures, even though borderline cases are undoubtedly encountered.