The classification of seizures according to whether consciousness is impaired is a major component of the commonly used convention outlined by the International League Against Epilepsy (ILAE) in 1981.¹ According to this system, partial seizures have initial clinical and electroencephalographic (EEG) evidence for localization to a neuronal system within one cerebral hemisphere. Partial seizures are then subcategorized as simple partial seizures (SPSs) and complex partial seizures (CPSs) based on whether the seizure produces an impairment of consciousness. If consciousness is not impaired, the seizure is considered simple partial. However, the matter of consciousness introduces a nuance to the classification that appears straightforward on a superficial level but actually may be difficult to implement. In the ILAE’s 1981 system, consciousness is defined as the integrating activity by which a totality is grasped and incorporated into experience. This differs from vigilance, and consciousness is operationally defined as awareness and responsiveness to external stimuli and then recollection of the experience. Because consciousness is a philosophically and scientifically complicated matter, an operational definition is reasonable, but even this definition is problematic because impairment of consciousness is not binary. Impairment may involve components of consciousness and also may occur in gradations.²

With greater recognition of the problems inherent in classifying seizures according to consciousness, the ILAE proposed in 2001 that simple partial and complex partial cease to be classification terms.^3,4 In the new classification system, impairment of consciousness is one of many seizure descriptions but not the basis of seizure classification. Nevertheless, the simple partial classification endures and is commonly used, perhaps reflecting both tradition and the importance of consciousness as a key factor in assessing the clinical impact of a seizure. As such, this chapter will use the term simple partial seizure while still recognizing the term’s limitations and proposed eventual elimination from the classification lexicon.

The distinction between SPS and CPS is further complicated by the common progression of an SPS to a CPS during one relatively brief episode. This possibility adds the diagnostic challenge of determining whether a seizure that is clearly an SPS at its onset becomes a CPS toward its resolution. However, the occurrence of an SPS at the onset of a CPS or a secondarily generalized seizure is also important for reasons beyond classification. Such seizures can provide a critically important warning for the patient. As has been known for millennia and described by Hippocrates, seizures may have early signs that allow patients to “know beforehand when they are about to be seized and flee from men, either to their homes or to a deserted place, and cover themselves up.”⁵ This integrity of awareness and responsiveness (consciousness) at the time of the seizure’s symptomatic onset allows the patient an opportunity to minimize the impact of the seizure. During the SPS, the patient also may alert others of the seizure. This warning historically has been called an aura, originally referring to a particular sensation that rises through the body.⁶ Aura is ancient Greek for “breeze,” but the term has lost its original specificity for this sensation. The ILAE’s 2001 glossary defines an aura as a subjective epileptic phenomenon that may precede an observable seizure and constitutes a sensory seizure if it remains subjective.⁷ Therefore, auras are a subcategory of SPSs that excludes simple partial motor seizures.

Prodromes

Auras should be distinguished from prodromes, which also precede seizures, but are a common but unrelated phenomenon in epilepsy. Unlike auras and other SPSs, which actually are ictal, prodromes are a preictal phenomenon. The ILAE’s 2001 glossary defines prodromes as subjective or objective clinical alterations that herald a seizure but are not part of the seizure.⁷ The neurologic basis of prodromes is not known, including whether early brain changes related to seizure onset cause the experience or are caused by it.

Prodromes do not directly lead to a seizure in about 50% of occurrences and may occur several hours earlier. About 70% of prodromes occur more than 6 hours before seizure onset, and about 5% occur more than 48 hours before seizure onset.⁸ Unlike auras, prodromes typically have an insidious onset, a long duration, and a more multifaceted abnormality that may include a combination of manifestations. The most common behavioral changes are irritability, emotionality, and general bad temper. Cognitive changes are most commonly psychomotor slowing and inattention; however, other common manifestations include uneasiness, apathy, fatigue, insomnia, and somnolence. The total duration most often is 30 minutes to several hours.

The common occurrence of prodromes adds to the importance of distinguishing them from auras. In one series of 100 patients (including 77 with focal epilepsies and 23 with generalized epilepsies), 51 had SPSs, and 39 had at least one prodrome.⁸ In another series, which included 148 patients (128 with focal epilepsies and 20 with generalized epilepsies), 29% of patients reported prodromes.⁹ The prevalence difference between focal epilepsies (42%) and generalized epilepsies (26%) was not significant (p = .2) in the first series and highly significant (34% of focal, 0% of generalized, p < .01) in the second series. This difference between studies underscores the relatively limited understanding of prodromes.

Auras

Auras occur in a broad range of patients and in a myriad of experiences. Lennox reported that among 750 of his patients with auras, 327 different descriptions were used.¹⁰ Of course, the report of an aura depends on the patient’s verbal ability. Among 155 patients ages 10 months to 49 years, 605 seizures were recorded, and the likelihood of an aura occurring was the same for all patient age groups with the language ability to report an aura.¹¹ However, the complexity of the aura’s manifestation depends on the level of language ability, as verbal intelligence is a better predictor of this complexity than any other clinical feature. Within this variety of aura experiences, the common experience is unpleasant. Lennox and Cobb’s early series found unpleasant auras (5.5% of auras) were more than 4 times more common than pleasant auras (1.2%).¹² A more recent review of 549 patients found only 11 (2%) with pleasant auras.¹³ Perhaps this commonality reflects the intrinsic qualities of the nature of perception.

Case series have provided characterizing statistics for auras in general and, with the use of video-EEG monitoring (VEM), for particular seizure types and epilepsy syndromes. An early review by Gowers included 2013 patients and reported that 57% had auras.¹⁴ This is remarkably similar to a series published by Lennox and Cobb 32 years later that included 1359 patients and reported that 56% had auras.¹² More recent series based on comprehensive epilepsy evaluations have correlated auras to seizure and epilepsy syndrome anatomy. Auras were reported by 83% (224/270) of patients with ictal EEG evidence of temporal lobe epilepsy (TLE) who were evaluated during the era before neuroimaging.¹⁵ A later, more modern TLE series reported 183 of the 260 patients (70%) had auras.¹⁶ However, patients commonly have seizures with and without auras. In a series of 19 patients with auras and TLE, auras occurred with 67% (36/54) of seizures.¹⁷ A higher rate was found in a collection of patients with TLE and epilepsy of the temporal lobe plus neighboring structures (TLE+).¹⁸ Of 80 seizures from 80 patients, 89% included an aura.

The aura rate for frontal lobe epilepsy (FLE) is similar to TLE. Of 26 patients undergoing an epilepsy surgery evaluation for FLE, 69% had a history of auras.¹⁹ However, the aura rate was associated with the type of ictal behavior. Auras occurred before all 9 of the complex partial frontal lobe seizures and just 9 of the 17 other types of frontal lobe seizures. This association with CPSs, regardless of lobar localization, is consistent with the observation that auras are more common in seizures including spread into the limbic system from a neocortical origin than in those that remain neocortical. They also are more likely to occur without progression to CPSs or generalized seizures (termed auras in isolation) when the origin is within the limbic system. In one series, frequent auras in isolation were present in 50% (29/58) of patients with TLE and 23% (14/61) of patients with FLE, which is a highly significant difference (p = .001).²⁰ A similar rate for FLE was observed in another series (23%, 6/26).¹⁹ However, this association with a limbic origin has not always been observed. An intracerebral EEG series of 55 TLE patients who had 187 seizures divided the patients according to whether the EEG ictal onset was mesial (24 patients), lateral (13 patients), or combined mesial and lateral (18 patients).²¹ The aura rate was similar across the three groups, with 79% of mesial, 84% of lateral, and 83% of combined.

As subjective episodes, auras are inherently sensory, but they may assume a broad variety of sensory experiences. Elementary sensory auras encompass one type of sensory experience and are defined as manifestations of one primary sensory modality. Their subdivision according to the sensory modality is clinically useful for localization.⁷

Somatosensory

All tactile sensations across the body’s surface are included as somatosensory auras and seizures. The sensations may be either negative (loss of sensation) or positive (illusion or hallucination of sensation), and positive auras may be any type of sensation. The sensation may be as familiar as tingling or an ache or as unusual as a sense of movement or a distal sense of need for movement. Overall, the paroxysmal and stereotyped nature and the association with other neurologic abnormalities are more useful in distinguishing the experience as epileptic than the quality of the sensory hallucination. The reported frequency of somatosensory auras within epilepsy varies greatly, possibly because of differences in selectivity during ascertainment. The frequency has been reported across the range of 1 to 60%.²² In one case series of 600 patients, 12% had somatosensory auras. Of these, 46% of the auras were contralateral to the cerebral lateralization, 6% were ipsilateral, 25% were not lateralized, and 24% had an unknown epilepsy lateralization.²² The consistency in this difference between contralateral and ipsilateral localizations has led to the interpretation of any unilateral somatosensory auras as indicating a contralateral cerebral localization.²³

The cerebral localization of somatosensory auras is not as reliable as the lateralization. Although the parietal lobe is highly developed for somatosensory function, other regions commonly produce somatosensory auras. This may be explained by the relatively low frequency of parietal lobe epilepsy; however, false localization of auras to other lobes because of an underrecognition of parietal lobe epilepsy is another possible explanation. Somatosensory auras are moderately common in TLE with a frequency of 17% for true TLE and 26% for TLE+. This frequency difference is not significant. Somatosensory auras are much more commonly associated with seizures arising from supplementary sensorimotor areas with a frequency of 43% (6/14).²⁴ The aura may be in either the upper or lower body and usually is contralateral, but a bilateral sensory experience is not uncommon, and ipsilateral also occurs.^24,25

Painful somatosensory auras are uncommon, with frequencies in two case series of 3% (24/856) and 4% (25/604).^26,27 Although they may arise from any cerebral lobe, they may have some localization value. However, this value is complicated by the greater prevalence of TLE. Within the series of 604 patients, 425 had TLE, and 3% of patients with TLE had painful auras. In contrast, painful auras were reported by 7% (6/85) of patients with FLE, and 10% (3/30) of patients with perirolandic epilepsy.²⁷ As such, the largest number of patients was those with TLE, but the other forms had higher percentages of patients with this aura. Therefore, aura type is more predictable from localization than is localization from aura type. Nevertheless, the presence of a painful aura decreases the overall likelihood of TLE. This is further demonstrated with painful abdominal auras, which have been reported to occur in 5% (6/119) of TLE patients with abdominal auras and in 50% (4/8) of FLE patients with abdominal auras. Therefore, the presence of a painful abdominal aura markedly raises the likelihood of FLE.

Visual

Most epileptic visual hallucinations are intermittent lights or spots, simple patterns, and sometimes amaurosis (Figure 8-1). Durations vary considerably, ranging from seconds to 150 minutes, with shorter durations occurring more commonly.²⁸ For both the positive and negative visual phenomena, several features are helpful in distinguishing visual auras due to epilepsy from visual auras due to migraine. The common use of the term aura in both epilepsy and migraine is not indicative of identical experiences or even similar pathophysiology. In a comparison of 67 visual hallucinations from 28 patients with epilepsy, 23 with migraine, and 3 with a combination of migraine and epilepsy, the visual field location of the positive and negative phenomena was found to have specificity. The positive visual phenomena of epilepsy were usually at the center of vision (83%, 20/24) and the positive visual phenomena of migraine were usually in the peripheral vision (77%, 10/13) (p < .0001).²⁹ Negative visual phenomena (amaurosis) in epilepsy were usually diffuse (71%, 10/14), and the negative visual phenomena of migraine were peripheral (75%, 9/12) (p < .05). Epileptic visual auras also are more commonly colored and circular than migraine auras, which are classically achromatic and angular.²⁸

Elementary visual auras suggest occipital lobe epilepsy (OLE) and are uncommon in TLE.³⁰ In a case series of 80 patients with TLE, elementary visual auras were experienced by 7% of patients with TLE and 13% of patients with TLE+.¹⁸ Therefore, elementary visual auras are not specific to OLE. Of 878 patients undergoing epilepsy surgery evaluations, 20 were found to have visual auras, and a total of 27 types of visual auras were identified.³¹ Of these, 8 were elementary, 3 were complex, 5 were illusions, and 11 were forms of visual loss. Of the 8 elementary visual auras, four were attributable to OLE, two to the temporal occipital junction, and two to the anterior mesial temporal lobe. Of the three complex visual auras, which were described as people or animals, one was attributable to the temporal occipital junction and two to the anterior mesial temporal lobe. This is consistent with the common wisdom that OLE is not the usual localization for complex visual auras.

Auditory

Auditory seizures are most commonly unformed sounds, such as buzzing or repeating tones (Figure 8-2). Less commonly, they are voices without words, animal sounds, or speech without language content. Their frequency is estimated to be between 2 and 16% of patients. In a series of 18 patients, the auditory experience was bilateral for 15 and contralateral to the epileptic localization for 3.³² Rarely, the auditory seizure is a negative phenomenon, which may be experienced as ear plugging. This was found to be contralateral to the epileptic localization for all three patients in one series.³³ The localization of auditory seizures is temporal, but they arise from the temporal neocortex so are not common in purely limbic TLE. Their frequency is 3% in TLE and 9% in TLE+.¹⁸ A comparison of nine patients with TLE and auditory auras who underwent intracranial ictal EEG found that six had a lateral temporal ictal onset, two had a mesial-lateral onset, and one had a mesial onset.²¹ This is highly supportive of auditory seizures usually having neocortical temporal sources (p = .005).

Vestibular

True vestibular seizures are manifested as vertigo and are distinct from less specific dizziness or wooziness. The vertigo is much more frequently experienced as an illusion of rotation than any other illusion of movement, such as translational movement that sometimes is categorized as a form of vertigo. The direction of movement usually is stereotyped for the individual, but the direction often is not recalled and does not have clinical significance regarding the lateralization of the seizure. The experience of true vertigo has localization value, as it is a sign of TLE, and much more likely to be neocortical TLE than mesial (limbic) TLE. Overall, vertigo is uncommon. About 2% of patients with TLE have vertigo.¹⁸ The rate is greater for TLE+ at about 13%, but still is not common.

Olfactory

The scent experienced during a seizure-related olfactory hallucination often is indescribable, usually is disagreeable, and essentially always is stereotyped for the individual experiencing it (Figure 8-3). Olfactory seizures occur in about 2% of TLE and 4% of TLE+ and are more likely to occur when hippocampal sclerosis is present.¹⁸ Of 202 seizures recorded from 35 patients with hippocampal sclerosis, 6 seizures (3%) included an olfactory hallucination.³⁴ In contrast, none of the 80 seizures recorded from 13 patients with other causes of TLE included an olfactory hallucination.³⁴ This approximate rate for olfactory hallucinations in TLE has been replicated in a series of 217 patients, of whom 5.5% had olfactory hallucinations. The specific hallucinations in this series were unpleasant for 11 of the 12 patients and indescribable for the remaining patient. The specific hallucinations were rotting for three patients; fetid, burning, or medicine for two patients each; and stink or alcohol for one patient each.³⁵

Gustatory

Analogous to the olfactory seizure, the taste experienced during a seizure-related gustatory hallucination often is indescribable, usually disagreeable, and essentially always stereotyped. The most common gustatory hallucinations are metallic, bitter, and acidic.

Commonly, patients have difficulty determining whether the experience was a taste or a smell, and a combination of both senses occurs occasionally. The frequency of gustatory hallucinations is similar to olfactory hallucinations. One series reports their occurrence in 4% of patients with epilepsy.³⁶ The localization also is similar, but with a greater rate of occurrence in extra-TLE. About 50% are TLE, 20% are TLE+, including the parietal lobe, and 30% are parietal lobe epilepsy.³⁶ This occurrence of greater likelihood in TLE compared with TLE+ was not observed in Barba et al.’s case series, which reported 5% in TLE and 22% in TLE+ and a significant difference between these groups (p = .02).¹⁸