!DOCTYPE html PUBLIC “-//W3C//DTD XHTML 1.1//EN” “http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd”>
CHAPTER 22 EPILEPSY WITH REFLEX SEIZURES
DEFINITION AND CLASSIFICATION
Although seizures occur usually unexpectedly, at least 50% of patients recognize a general or more specific precipitating factor (1). Emotional stress and physical stress are by far the most recognized factors. Physical stress includes fever, sleep deprivation, hyperventilation, drug or alcohol withdrawal, menstruation, and physical exercise. Mental or emotional stress is less clearly demonstrable and can be due to either happy occasions (birthday, Santa Claus, etc.) or more negative experiences (problems at work, and major life events like death and divorce) (2).
Seizures provoked by the above-mentioned general factors alone are, however, not considered to be reflex seizures. The provoking factor needs to be specific. Usually, the triggers are brief and sudden like flashing lights, sudden noises, and tapping, provoking especially myoclonic jerks, although they can be more complex and gradual like reading, thinking, eating, bathing, drawing, gaming, and listening to music, with less sudden seizure expressions (temporal lobe type of seizures).
In most hospital-based studies, about 6% of epilepsy patients recognize specific factors as their only or predominant precipitant of seizures. In a survey of seizure-provoking factors in a community-based cohort of adolescent (>12 years) and adult epilepsy patients, a total of 47% reported at least one factor that could provoke their seizures. Of these subjects, seizures provoked by specific factors were mentioned to be evoked predominantly by flickering lights (sunlight, stroboscope light, TV, or video games) and sounds in 17% and 5%, respectively (3); thus, more than 8% of epilepsy patients mention reflex seizures.
Reflex seizures may be classified as occurring in generalized or in focal epilepsy syndromes (4). Reflex seizures that occur in patients who also have spontaneous seizures are generally listed as seizure types, for example, photosensitive seizures in patients with juvenile myoclonic epilepsy (JME). Reflex seizures can also be classified according to the seizure trigger. If the trigger is exceptionally specific or exotic, the epilepsy is even named after the provocative factor like mahjong epilepsy, telephone epilepsy, vacuum cleaner epilepsy, and tooth-brushing epilepsy (5–8). However, it is questionable whether it is useful to make such detailed subdivisions (9).
The use of the term reflex has been and still is controversial. Marshall Hall (10) first applied it to epilepsy in 1850; he differentiated seizures precipitated by peripheral stimuli (eccentric or reflex) from central causes. About a century later, in the 1960s, Servit, a neo-Pavlovian, stated that the genesis of epilepsy can be considered as a reflex mechanism (11). Arguing that no reflex arc is involved in reflex epilepsy, others proposed terms such as sensory precipitation (12,13) or stimulus-sensitive epilepsies with simple and complex forms (14).
The International League Against Epilepsy (ILAE) (15) describes in 1989 reflex epilepsies as “epilepsies characterized by specific modes of seizure precipitation.” The classification proposal of Engel from 2001 (16) redefines reflex epilepsy syndromes as “syndromes in which all epileptic seizures are precipitated by sensory stimuli.” Apart from the fact that reflex epilepsies are characterized by seizures, specific syndromes among the reflex epilepsies have also been recognized: idiopathic photosensitive occipital lobe epilepsy (IPOE), other visual-sensitive epilepsies, primary reading epilepsy, language- induced epilepsy, seizures induced by thinking, eating epilepsy, and musicogenic epilepsy (16). At the same time, the ILAE task force published a glossary of descriptive terminology for ictal semiology (17). In this glossary, a provocative factor is described as “a transient and sporadic endogenous or exogenous element capable of augmenting seizure incidence in persons with chronic epilepsy and evoking seizures in susceptible individuals without epilepsy.” This article therefore distinguishes between reactive seizures (due to illness, sleep loss, or emotional stress) and reflex seizures (objectively and consistently demonstrated to be evoked by a specific afferent stimulus or by an activity of the patient). Afferent stimuli can be elementary (i.e., unstructured [light flashes, startle, a monotone]) or elaborate (i.e., structured, [a symphony]). Activity may be elementary (e.g., motor [a movement]), elaborate (e.g., cognitive function [reading, chess playing]), or both (reading aloud).
The more recent ILAE report on classification and terminology (18) has mentioned only reflex epilepsies as an electroclinical syndrome or other epilepsies with a less specific age relationship; the fact that reflex seizures were not included at all has been criticized by Panayiotopoulos (19). As an advocate of the multifactorial Lennox theory of the epileptic threshold, Shorvon introduces in his etiologic classification (2011) a new, broader, category of epilepsy, namely “provoked epilepsy” that includes the reflex epilepsies (20). This provoked epilepsy is defined as an epilepsy in which a specific systemic or environmental factor is the predominant cause of the seizures and in which there are no gross causative neuroanatomic or neuropathologic changes. Thus, until today, the definition and classification of reflex seizures and reflex epilepsies remain a point of discussion and debate: is there really a dichotomy between reflex and spontaneous seizures and—if so—to what extent? (9,20,21).
Most epileptologists refer to reflex epilepsy when a certain stimulus regularly elicits a response in the form of abnormal, paroxysmal, electroencephalographic (EEG) activity associated with or without a clinical seizure. Although some investigators restrict the term reflex epilepsy to cases in which a certain stimulus always induces seizures (22), it may include cases in which typical spontaneous seizures occur without the trigger and in which triggers do not always induce a seizure (23).
In a recent inventory of definitions used in the literature between 2002 and 2012, it was demonstrated that few definitions were identical, and definitions of a particular type varied considerably with a wide range of components being referred to as part of reflex seizures (21). Consensus on terminology thus does not exist at this moment. Reflex seizures may not differ in semiology from those encountered in other forms of epilepsy, but understanding of the seizure trigger is important in the management of the patient and helps study mechanisms of epileptogenicity. Interesting in that respect is the documentation of a 57-year-old man with temporal lobe epilepsy, who had 40% of his seizures provoked by tooth brushing, looking at the toothbrush, or even thinking about the brush (24). In this case, a somewhat wider epileptogenic network seems to be involved than usually based on the broader range of triggers.
The term “epilepsy with reflex seizures,” or perhaps provoked epilepsy as suggested by Shorvon (20), to underline the unmistakable interaction between the various general and specific factors, would better reflect clinical reality and more accurately describe cases with both reflex and spontaneous attacks.
Reflex seizures have long fascinated epileptologists. Photosensitivity or better visual sensitivity, thus comprising all types of visual stimuli like TV screens, video games, and striped patterns, is mostly recognized and studied, while making up for 90% of all reflex epilepsy patients (25). After introducing the rather simple provocation method of intermittent photic stimulation (IPS) during an EEG in the early fifties, visual sensitivity has over the years been studied intensively both clinically and electroencephalographically. Apart from epileptic photosensitivity to flickering light and other visual stimuli, other cases of reflex epilepsy such as thinking- and reading-evoked seizures are rare and permit glimpses into the organization of cognitive function.
The identification of a patient with reflex epilepsy depends on the physician’s awareness and on the observations of the patient and witnesses. The epileptogenic trigger must occur often enough in everyday life so that the patient suspects its relation to the resulting seizures. If the trigger is ubiquitous, however, the seizures appear to occur by chance or with no obvious antecedent. If longer stimulation is necessary like in reading and musicogenic epilepsy, detecting a relationship between the stimulus and seizure is much more difficult. Many different triggers have been recognized and studied (5–8).
This chapter reviews the neurophysiology of reflex epilepsy from available human and animal studies. We also discuss clinical features of reflex seizures classified by the triggering stimulus.
BASIC MECHANISMS OF REFLEX EPILEPSY
There are two types of reflex epilepsy animal models. In the first, irritative cortical lesions are created, and their activation by specific stimuli is studied. The second model type involves naturally occurring reflex epilepsies or seizures induced by specific sensory stimulation in genetically predisposed animals.
The first approach has been used since 1929, when Clementi (26) induced convulsions with IPS after applying strychnine to the visual cortex. This technique also demonstrated that strychninization of auditory (27), gustatory (28), and olfactory cortex (29) produced focal irritative lesions that may produce seizures with the appropriate afferent stimulus. EEG studies showed that the clinical seizures (chewing movements), which were induced by photic stimulation in rabbits with strychnine lesions of the visual cortex, resulted from rapid transmission of the epileptic discharge from the visual cortex to masticatory areas (30). The EEG spread of paroxysmal discharges from the visual cortex may also occur in fronto-rolandic areas during seizures (31). The ictal EEG spread was thought to represent corticocortical conduction (17,31), although later work with pentylenetetrazol also implicated thalamic relays (32) and demonstrated spread of the visual evoked potential to the brain stem reticular formation (33). Hunter and Ingvar (34) identified a subcortical pathway involving the thalamus and reticular system and an independent corticocortical system for spread of visual-evoked responses to the frontal lobe. In cats and monkeys, the fronto-rolandic region was also shown to receive spreading-evoked paroxysmal activity from auditory and other stimuli (35,36).
The second approach, the study of naturally occurring or induced reflex seizures in genetically susceptible animals, has been pursued in domestic fowl and chickens with photosensitivity (37,38), rodents susceptible to sound-induced convulsions (39), the E1 mouse sensitive to vestibular stimulation (40), and the Mongolian gerbil sensitive to a variety of stimuli (41,42).
The species in which the reflex seizures and EEG findings are quite similar to those in humans is the baboon Papio hamadryas papio, better known as Papio papio from Senegal (43), except that the light-induced epileptic discharges in baboons occur in the fronto-rolandic area rather than in the occipital lobe (44). Additionally, there is a statistically significant association of spontaneous epileptiform discharges and seizures with photosensitivity (45,46). Like in humans, photosensitivity is maximal when the animals had their eyes closed during IPS or with chemically induced pupillary dilation (47). When the light source is closer (8 cm vs. 25 cm) and its surface area is larger (128 cm2 vs. 8 cm2), IPS is more likely to trigger epileptic discharges. However, unlike in humans, colors of shorter wavelengths, such as blue-green and dark green, are more effective in producing EEG afterdischarges than are colors of longer wavelengths, such as red.
EEG, visual evoked potentials, intracerebral recordings, lesions, and pharmacologic studies show that visual afferents are necessary to trigger fronto-rolandic light-induced epileptic discharges. The occipital lobe does not generate this abnormal activity but sends corticocortical visual afferents to hyperexcitable frontal cortex, which is responsible for the epileptiform activity (48). The interhemispheric synchronization of the light-induced paroxysmal EEG activity and seizures depends mainly on the corpus callosum and not on the brain stem. Brain stem reticular activation depends initially on frontal cortical mechanisms until a seizure is about to begin, at which point the cortex can no longer control reticular activation. The genetically determined hyperexcitability may be related to cortical biochemical abnormalities, involving regulation of extracellular calcium concentration (49,50) or an imbalance between excitatory and inhibitory neurotransmitter amino acids (51) similar to those described in feline generalized penicillin epilepsy and in human epilepsy (52).
In recent studies in members of the baboon subspecies Papio hamadryas anubis or Papio hamadryas cynocephalus or hybrids housed in the Southwest National Primate Research Center at the SFBR in San Antonio, photosensitive baboons demonstrated increased cerebral blood flow using functional PET in the right orbitofrontal and anterior cingulate region compared to nonepileptic nonhuman primate controls. There were significant activations in the parietal and motor cortices, but no activation of the occipital lobes. In the control animals, there was an expected activation of the striate and peristriate cortices and posterior cingulate gyrus. While the activations of the motor cortices were expected due to activation of discrete myoclonic seizures, the role of the frontal and parietal lobe activations in the photosensitive baboons was unclear. The absence of occipital activation and deactivation of the posterior cingulate gyrus suggested inhibition in these areas (53).
In human epileptic photosensitivity, generalized epileptiform activity and clinical seizures can be activated by the occipital trigger. Studies in photosensitive patients who are also sensitive to black and white striped patterns suggest that generalized seizures and EEG paroxysmal activity can occur in these subjects if normal excitation of the visual cortex involves a certain “critical mass” of cortical area with synchronization and subsequent spreading of excitation (54–57). Wilkins et al. (58) suggested that a similar mechanism involving recruitment of a critical mass of parietal rather than visual cortex is responsible for generalized seizures induced by thinking or by spatial tasks. Studies of reading epilepsy also suggest that increased task difficulty, complexity, or duration increase the chance of EEG or clinical activation (59,60).
Wieser (61) proposed a neurophysiologic model for critical mass, referring to the group 1 and group 2 epileptic neurons of the chronic experimental epileptic focus described by Wyler and Ward (62). Group 1 neurons produce abundant, spontaneous, high-frequency bursts of action potentials. Group 2 neurons have a variable interspike interval, and their spontaneous epileptic activity is less marked. Moreover, these properties are influenced by external stimuli that can promote or inhibit the incorporation of group 2 neurons into the effective quantity of epileptic tissue and thus trigger or inhibit a seizure. The stimuli effective in eliciting reflex seizures would act on this population of neurons, recruiting them into the highly epileptic group 1 neuron pool to form the critical mass needed to produce epileptogenic EEG activity or clinical seizures. This mechanism can also explain conditioning (63) and deconditioning (64) of reflex epileptic responses. A further generalizing system must also be postulated to account for the seizures observed with photic or cognitive stimulation, analogous to the corticocortical pathways linking occipital cortex with fronto-rolandic cortex in Papio papio. A role for reticulothalamic structures has been suggested but seems unnecessary, at least in certain animal models in which corticocortical spread of evoked epileptic activity persists after mesencephalic and diencephalic ablation (34).
Patients and parents may report that emotion plays a fundamental role in seizure induction, and such an observation has been also described recently in a 9-year-old girl by Gilboa (65).
Gras et al. (66) emphasized the influence of emotional content in activating EEG spikes in a patient with reading epilepsy. An emotional component was also obvious in several cases of musicogenic and eating epilepsy. Similarly, Scollo-Lavizzari and Hess (67) showed that the mere sight, memory, or anticipation of releasing stimuli could elicit the same epileptiform EEG abnormalities and seizures as did the stimulus itself.
Fenwick (68) described psychogenic seizures as epileptic seizures generated by an action of mind, self-induced attacks (e.g., by thinking sad thoughts), and those unintentionally triggered by specific mental activity such as thinking. This use of the term psychogenic seizures, common in European epileptology, does not refer to nonepileptic events. Fenwick related seizure induction and inhibition in some individuals with or without typical reflex seizures to the neuronal excitation and inhibition accompanying mental activity. He also referred to the alumina cream model, with recruitment of group 2 neurons and evoked change in neuronal activity surrounding the seizure focus as factors in seizure occurrence, spread, and inhibition.
Wolf (69) believed that two pathophysiologic theories have arisen in the discussion of reflex epilepsies. For primary reading epilepsy, he observed that seizure evocation would depend on involvement of the multiple processes used for reading, an activity involving both hemispheres, with a functional rather than a topographic anatomy. “Maximal interactive neuronal performance is at least a facilitating factor,” (69) he wrote and suggested that the functional complexity of the epileptogenic tasks leads to seizure precipitation. He contrasted this with the suggestion described previously that the latency, dependence on task duration and complexity, and influence of nonspecific factors such as attention and arousal often observed in these seizures depend on the ad hoc recruitment of a critical mass of epileptogenic tissue to produce a clinical seizure or paroxysmal EEG activity in response to the different characteristics of an effective triggering stimulus. In seizures induced by reading, thinking, IPS, and striped patterns, the relatively localized trigger induces generalized or bilateral EEG abnormalities and seizures. The recruitment that produces these seizures, however, need not be confined to physically contiguous brain tissue or fixed neuronal links. Instead, it may depend on activity of a function-related network of both established and plastic links between brain regions, modified by the effects of factors such as arousal. These two approaches share much common ground.
Disorders of cortical development based on genetic predisposition may be present in some patients with reflex seizures: musicogenic reflex epilepsy was shown in a patient with Dravet syndrome (70).
Reportedly normal imaging may be misleading; subtle changes or dysplastic lesions may be missed without special magnetic resonance imaging (MRI) techniques or may only be found in a surgical pathology specimen (71–75).
Of special interest is also the case history of a 19-year-old male with thinking-induced seizures that occurred exclusively during acute bacterial meningitis (76). SPECT and MRI studies in patients with eating epilepsy have shown lesions near the perisylvian region (77). Future fMRI and PET studies should elucidate the preferred networks of the reflex stimulus, although many individual (genetic) differences may exist.
SEIZURES INDUCED BY VISUAL TRIGGERS
Epilepsy with reflex seizures evoked by a variety of visual stimuli in daily life (78,79) is clearly the most common reflex epilepsy. Even epidemics have occurred: after a specific Pokemon cartoon TV transmission in Japan, studies have shown that about 1% of the viewers, mainly children, had provoked seizures (in about 75% it was the first seizure) and 10% had symptoms of headache, nausea, blurred vision, vertigo, etc. (80,81). In about 40% of these cases, photoparoxysmal EEG responses or photoparoxysmal responses (PPRs) were found.
Complaints of headache are found significantly more often in photosensitive patients (twice as often) and it is also a regular finding during a PPR (78,82). Headache can also be the sole manifestation of an epileptic event and remaining complaint after AED treatment (83), especially in families with both migraine and epilepsy (84). Conversely, if children and adolescents complain about headache and are diagnosed as having migraine with aura, a PPR is found in about 30% (85).
Following the Pokemon incident, subjects known to have visually induced seizures were examined whether color modulation could be an independent factor in human epileptic photosensitivity. Among photosensitive epilepsy patients sensitive to flash and pattern stimulation, 25 of 43 were sensitive to color stimulation, particularly at frequencies below 30 per second. Red was the most effective color, and red-blue was the most provocative alternating stimulus. They concluded that “color sensitivity follows two different mechanisms: one, dependent on color modulation, plays a role at lower frequencies (5 to 30 Hz). Another, dependent on single-color light- intensity modulation, correlates with white light sensitivity and is activated at higher frequencies.” This also suggests a mechanism to explain observations that colored spectacles adapted to the patient’s color sensitivity may be useful for treatment (86).
Of the several abnormal EEG responses described in the laboratory, IPS and especially evoked generalized paroxysmal epileptiform discharges (e.g., spikes, polyspikes, spike-and-wave complexes) are clearly linked to epilepsy in humans. About 5% of all patients with epilepsy show this response to IPS (78,87) and 1.4% of healthy schoolchildren. No follow-up studies have been performed in these children in order to know if clinical seizures did occur later in life.
Family studies indicate an autosomal dominant mode of inheritance of PPRs with age- and sex-dependent penetrance (88,89). Several molecular genetic studies in generalized epilepsy patients with a PPR have shown evidence for linkage to chromosomes 7q32; 16p13 (90) and to 6p21.2; 13q31.3 (91). Combination of both studies and additional samples from other European countries identified 16p13.3, 5q35.3, and 8q21.13, with replication of only 16p13.3 (92). Continued research is necessary taking EEG patterns and clinical information into account.
Photosensitivity occurs most frequently in adolescents and women, but studies of the epileptic response to IPS are complicated by differences in how IPS is performed. An expert panel has recommended a protocol for performing IPS and guidelines for interpreting the EEG responses (93,94). This is also shown in a review with video (55).
Sensitivity to IPS is often divided into three groups: patients with light-induced seizures only, patients with photosensitivity and other seizure types, and asymptomatic individuals with isolated PPRs (nonepilepsy patients or family members of epilepsy patients). However, half of known photosensitive patients questioned immediately after stimulation, denied having had brief but clear-cut seizures induced by IPS and documented by video–EEG monitoring (95). In combination with the observation that patients can start having overt (light- induced) seizures several years after discovery of a PPR as seen in family studies, any subdivision should be taken as flexible as possible. Until now, no controlled studies have been performed on evolution of the various physiologic phenomena in combination with clinical data. A flexible system using all available information can be used as a temporary informative classification. The following working categories can be discerned (82):
1. Normal individuals (no history of epileptic seizures) with a PPR in the EEG
2. A first visually induced seizure in special circumstances (TV, stroboscopic lights, video games, etc.), with or without a PPR
3. Recurrent visually induced seizures, with or without a PPR
4. Spontaneous seizures and a PPR
5. Visually induced and spontaneous seizures, without a PPR
6. Visually induced and spontaneous seizures, with a PPR
Decision about treatment should, however, always be made on an individual level based on prevention of seizures as primary goal, rather than classification in a subgroup. Criteria like extent of sensitivity to visual stimuli (range of flash frequencies, size of patterns, reaction to TV/computer screens) and severity of seizures or complaints in combination with age and lifestyle are important.
For educational purposes, we discuss two commonly encountered discriminative clinical manifestations in more detail: (i) recurrent visually induced seizures and (ii) spontaneous seizures and a PPR. In reality and dependent on available EEG, clinical, and follow-up information data, there is an overlap. Although most patients are sensitive to various visual triggers (96), patients can similarly be roughly divided in those primarily sensitive to flicker (the majority) and those sensitive to pattern. A specific category of patients are the self-inducers; they use their sensitivity to visual stimuli to evoke PPRs. Self-induction is therefore described in a subsection.
Patients with Recurrent Visually Induced Seizures
According to Jeavons (97), 40% of photosensitive patients have generalized seizures provoked exclusively by a flickering light source, and television is the most common precipitating factor. Video games may also trigger these generalized seizures (96,98). Other typical environmental stimuli include discothèque (stroboscopic) lights and sunlight reflected from snow or the sea or interrupted by roadside structures or trees. Rotating helicopter rotors and tower-mounted wind turbines, which can reflect or break up light into flicker, also present risk (99,100).
In this category, female adolescents are typically overrepresented. Reviews of the topic have been provided by several authors (78,87,95,101). The seizures are generalized tonic–clonic in 84% of patients (102), absences in 6%, partial motor seizures (possibly asymmetric myoclonus in some cases) in 2.5%, and myoclonic seizures in 1.5% of patients. Subtle myoclonic seizures may go unnoticed until an obvious seizure occurs. The developmental and neurologic examinations are normal. The resting EEG may be normal in about one-half of patients, but spike-and-wave complexes may be seen with eye closure. IPS evokes a PPR in virtually all patients. Depending on the photic stimulus and on the patient’s degree of photosensitivity, the clinical response ranges from subtle eyelid myoclonus to a generalized tonic–clonic convulsion.
These photosensitive patients are typically conceptualized as having a variety of genetic generalized epilepsy, but cases occur in which EEG and clinical evidence favors the occipital lobe origin, as predicted by theoretical and animal models (103,104). IPS can also induce clear-cut partial seizures originating in the occipital lobe (105,106). As in more typical photosensitive subjects, environmental triggers include television and video games. Many of these patients have IPOE, a relatively benign, age-related syndrome without spontaneous seizures, although cases with occipital lesions have been reported, including patients with celiac disease. The clinical seizure pattern depends on the pattern of spread. The visual stimulus triggers initial visual symptoms that may be followed by versive movements and motor seizures; however, migraine-like symptoms of throbbing headache, nausea, and sometimes vomiting are common and can lead to delayed or incorrect diagnosis.
Since the seizures are provoked by visual stimuli, the best prevention is avoidance, or modification of the environmental light stimuli, such as increasing the distance from the television set, watching a small screen in a well-lighted room, using a remote control so that the set need not be approached, and monocular viewing or the use of polarized spectacles to block one eye, should provide protection (97,107). Colored spectacles may be useful in selected cases (108,109).
Drug treatment is needed if these measures are impractical or unsuccessful or if photosensitivity is severe. The drug of choice, in particular in male patients, is valproate. Valproate abolishes PPRs in 54% of patients and markedly reduces it in additional 24% (110). In females, preference goes to levetiracetam (95) and lamotrigine. Benzodiazepines such as clobazam (111) may be also useful. About one-fourth of patients lose their photosensitivity by 25 years (112). Because this resolution usually occurs only in the third decade, too early withdrawal of treatment may facilitate seizure recurrence; serial EEG recording to determine the photosensitivity range may be helpful in assessment and during follow-up (94,95).
Patients with Spontaneous Seizures and PPR
Jeavons and Harding (102) found that about one-third of their photosensitive patients with environmentally precipitated attacks also had spontaneous seizures. Spike-and-wave activity was common in the resting EEG patterns of patients with spontaneous seizures, and only 39% of patients had normal resting EEGs. PPRs may accompany idiopathic generalized epilepsies, especially JME, and is associated with onset in childhood and adolescence, normal intellectual development and neurologic examination, normal EEG background rhythm, and generally good response to treatment. Photosensitive benign myoclonic epilepsy may also begin in infancy, with a generally good prognosis though the events may be overlooked by the parents for some time before diagnosis (113). PPRs also may occur with severe myoclonic epilepsy of infancy or Dravet syndrome (114) or with diseases associated with progressive myoclonic epilepsy syndromes like Lafora body disease, Unverricht–Lundborg disease, and the neuronal ceroid lipofuscinoses (114). Photosensitivity is generally seen in eyelid myoclonia with absences (EMA) but not in benign occipital epilepsies of childhood of the Gastaut or Panayiotopoulos types despite the florid EEG abnormalities (114).
Seizures Induced by Patterns
Absences, myoclonus, or more rarely, tonic–clonic seizures may occur in response to epileptogenic patterns. These are striped and include common objects such as the television screen at short distances, video games, curtains or wallpaper, escalator steps, and striped clothing. Pattern sensitivity is seen in about 70% of photosensitive patients tested with patterned IPS in the EEG laboratory, but sensitivity to stationary striped patterns affects only about 30% (56). However, clinical pattern sensitivity is not easily recognized: patients often may not make the association, the family may be unaware of it, and physicians may not inquire about it.
Wilkins et al. (55,57,115–117) studied the properties of epileptogenic patterns, isolating visual arc size, brightness, contrast, orientation, duty cycle, and sensitivity to movement and binocularity. They concluded that the seizures involve excitation and synchronization of a sufficiently large number of cells in the primary visual cortex with subsequent generalization. We can compare this with the previously described animal experiments and Wieser’s theory. Pattern sensitivity optimally requires binocular viewing, and treatment may be aided by avoidance of environmental stimuli (admittedly often impractical) as well as by alternating occlusion of one eye with polarizing spectacles and increased distance from the television set. A high degree of pattern sensitivity requires antiepileptic drug treatment, as described for the predominantly flicker-sensitive patients.
Seizures Induced by Television and Other Electronic Screens
Television likely remains the most common environmental trigger of photosensitive seizures. A television screen produces flicker at the main frequencies, generating IPS at 60 Hz in North America and 50 Hz in Europe. Jeavons and Harding (102) found that photosensitivity was more common at the lower frequency, which partly explained the higher incidence of television-induced seizures in Europe as compared to North America. Television-induced seizures, however, are not related to alternating current (AC) frequency flicker alone. Wilkins et al. (116,117) described two types of television-sensitive patients: those sensitive to IPS at 50 Hz, who apparently were sensitive to whole-screen flicker even at distances >1 m from the screen, and patients not sensitive to the mains frequency flicker but who responded to the vibrating pattern of interleaved lines at half the AC frequency, which can be discerned only near the screen. They emphasized that increased distance from the screen decreased the ability to resolve the line pattern and that a small screen evoked less epileptiform activity than a large one. Binocular viewing was also needed to trigger attacks.
Not surprisingly, domestic video games using the home television screen viewed at close distances for long periods and sometimes under conditions of sleep deprivation and possible alcohol or nonmedical drug use can trigger seizures in predisposed individuals. Some individuals are not photosensitive and may have seizures by chance or induced by thinking or other factors. These events, however, have caused many patients with epilepsy to believe erroneously that they are at risk of seizures evoked by video games, and they need accurate information about their personal risk (118).
Not all seizures triggered by television and similar screens fit this pattern. Seizures can be triggered even at greater distances and by noninterlaced screens without flicker. Flashing or patterned screen content is implicated in such episodes including that from video games. Nevertheless, the 50/25-Hz frequency appears to be a powerful determinant of screen sensitivity, and in countries with 50-Hz AC, special 100-Hz television sets have been shown to reduce the risk of attacks (119). Other preventive measures include watching a small screen from afar in a well-lighted room, using a remote control to avoid approaching the set, and covering one eye and looking away if the picture flickers or if myoclonia occurs (120).
Broadcasting of certain forms of flashing or patterned screen content has been responsible for outbreaks of photosensitive seizures, most notably in Japan where 685 people, mostly children and young adults with no history of epilepsy, were hospitalized after viewing a cartoon (121). The details of triggering factors in screen images have been summarized (122) and were used to develop broadcast standards in the United Kingdom and Japan, which now reduce this risk. Electronic filters have also been proposed (123). Modern displays, like liquid crystal, organic and polymer light-emitting diodes, and plasma-driven panels, all generate their light output in a different way, and it will probably depend on the size of the screen, the contrast, and the luminance as to whether these provoke seizures in photosensitive patients.
Further outbreaks are to be expected if viewers, especially mass audiences of adolescents, are exposed to such screen content when guidelines do not exist or are violated (124). The diversity in screens and programs combined with at random exposure to potentially epileptogenic triggers might make it much more difficult to recognize this form of photosensitivity.
Self-Induced Seizures
Reports of self-induced epileptic attacks using visual sensitivity antedated the discovery of the photoconvulsive EEG response (125). Regarded as rare, self-induction was reported particularly in mentally retarded children and adolescents, with a female preponderance (78,87,126,127). More recent information, however, shows that although some affected patients are mentally challenged, most patients have relatively normal or only mildly delayed development (128–130). When carefully sought, the syndrome is not rare; it was found in about 40% of photosensitive patients studied by Kasteleijn-Nolst Trenité et al. (128). The EEG usually shows spontaneous generalized spikes or spike-and-wave complexes, and about 75% of patients are sensitive to IPS. The self-induced seizures are usually myoclonic, especially with palpebral myoclonus, or absences, and some patients have EMA. Patients induce seizures with manoeuvers that cause flicker, such as waving a hand with fingers spread apart in front of their eyes or gazing at a vertically rolling television image. Monitoring (130,131) shows that these behaviors, once thought to be part of the seizure, precede the attacks and are responsible for inducing them. The compulsive nature of this behavior has been observed often and has been likened to self-stimulation (132) in experimental animals. Patients have reported intensely pleasant sensations and relief of stress with self-induced photosensitive absence seizures (129,130). Frank sexual arousal has been described (133,134). Patients are often unwilling to give up their seizures, and noncompliance with standard, well-tolerated antiepileptic drugs is common (128,129). Treatment is difficult, however, even in compliant patients (127). Drugs that suppress self-stimulation in animals, such as chlorpromazine and pimozide, may block the pleasurable response without affecting the response to IPS and have partially reduced or completely terminated self-induction (127,135). The effectiveness of valproate in reducing or abolishing photosensitivity has resulted in virtual disappearance of this form of self-induction, which is now encountered in patients for whom the drug has not been prescribed and in those with inadequate drug levels for any reason.
Quesney et al. (136) proposed a dopaminergic mechanism in human epileptic photosensitivity based on the transient abolition of photosensitivity with apomorphine, and bromocriptine and parenteral L-dopa have been reported to alleviate photosensitivity (137,138). Based on these findings, a candidate linkage study on the five dopamine receptor gene regions of DRD1 to DRD5 was performed in families with PPR, but none of these loci were found to be linked to the PPR (139).
In Dravet patients who self-induce, treatment with the serotonergic drug fenfluramine and the voltage-gated calcium channel blocker verapamil has been effective (140,141). However, many patients appear not to want treatment for their self-induced attacks. Photosensitive patients may induce seizures with maneuvers that do not produce flicker. These attacks are similar to flicker-induced seizures, but the inducing behaviors are not. Pattern-sensitive patients may be irresistibly drawn to television screens, which they must approach closely to resolve the epileptogenic pattern of vibrating lines, or they may spend hours gazing through venetian blinds or at other sources of pattern stimulation. Those sensitive to eye closure have been observed to use forceful slow upward gaze with eyelid flutter (142,143) to induce paroxysmal EEG discharge and, at times, frank seizures. These patients are often children, who describe the responses as pleasant: “as nice as being hugged, but not as nice as eating pudding” (Kasteleijn, personal observation). We have noticed that these tonic eyeball movements are always associated with spike-and-wave activity in young individuals. As they mature, their movements may persist but no longer elicit epileptiform activity and can be likened to a tic learned in response to positive reinforcement. These observations and the compulsive seizure-inducing behavior of many such patients suggest that, as in flicker- induced seizures, the self-induced attacks provide pleasure or relieve stress. Experience suggests that treatment is similarly difficult (128). They must be distinguished from rare seizures occurring with eyes closed or with loss of central fixation. Panayiotopoulos et al. (144,145) studied these extensively and described the syndromes in which they occur.
SEIZURES INDUCED BY COMPLEX NONVISUAL ACTIVITY
Reflex epilepsy with nonvisual stimuli is rare though reflex seizures with JME are more common than previously thought (146,147). In a Brazilian study, the following triggers were found to be provocative in adult JME patients based on questionnaires: thinking and concentration (23%), praxis (20%), visual stimuli (15%), speaking in public (11%), reading (7%), calculating and writing (5%), music-related activities (5%), and drawing (3%).
Seizures may be classified as those with relatively simple somatosensory triggers (bathing, rubbing, etc.) and those triggered by complex activity, such as thinking, eating, listening to music, or gaming [for reviews, see Gastaut and Tassinari (142) and Striano et al. (148)]. There are clear differences between populations regarding prevalence of specific reflex stimuli: in India, eating and hot water bathing (pouring hot water on the head) are the most prevalent triggering factors (149), while in Caucasians, language-related stimuli like reading and writing are more common.
Seizures Induced by Thinking and Gaming
Wilkins et al. (58) introduced the term seizures induced by thinking to describe a patient who reported seizures induced by mental arithmetic but who proved also to be sensitive to tasks involving manipulation of spatial information with or without any motor activity. Other complex mental activities have been reported to trigger seizures, such as card games and board games, such as checkers (British, draughts), mahjong, go and Baduk games, or making complex decisions. A rather consistent electroclinical syndrome emerges, most succinctly called seizures induced by thinking, reviewed in Andermann et al. (150). EEG monitoring during detailed neuropsychological testing is not always performed, and this peculiar type of reflex epilepsy might thus be more common than originally estimated. Reading is not usually an effective trigger, and unlike reading epilepsy, most patients also have apparently spontaneous attacks. The seizures are typically generalized myoclonus, absences, or tonic–clonic attacks, and the induced EEG abnormalities are almost always generalized spike-and-wave or polyspike-and-wave activity. Focal spiking is found in only about 10% of patients, and photosensitivity is seen in about 25%. Although numbers are small, most subjects are men. The mean age of onset is 15 years. Family histories of epilepsy are neither typical nor helpful in the diagnosis. Avoidance of triggering stimuli is practical only when activation is related to cards or other games, but drugs effective in idiopathic generalized epilepsies have been most useful. Epileptogenic tasks in these patients involve the processing of spatial information and possibly sequential decisions. The generalized seizures and EEG discharges may depend on initial involvement of parietal or, possibly, frontal cortex and subsequent generalization, much as pattern-sensitive seizures depend on initial activation of primary visual cortex. Recent studies provide more detail on the cerebral representation of calculation and spatial thought and document a bilateral functional network activated by such tasks (151). Similar results were found in a Korean study in 11 patients with seizures exclusively being experienced while playing Go–stop or Baduk games with one interesting difference: the mean age at onset was 53.7 years, and they had played the games without any problem for many years since adolescence (152).
Praxis-Induced Seizures
Japanese investigators (153) have described praxis-induced seizures as myoclonic seizures, absences, or generalized convulsions triggered by activities as in seizures induced by thinking but with the difference that precipitation depends on using a part of the body to perform the task (e.g., typing). Hand or finger movements without “action-programming activity” (defined as “higher mental activity requiring hand movement” and apparently synonymous with praxis) are not effective triggers (154). EEG responses consist of bisynchronous spike or polyspike-and-wave bursts at times predominant over centroparietal regions. Most subjects have JME; some had another idiopathic generalized epilepsy syndrome. None had clear-cut localization-related epilepsy. In its milder forms, such as the morning myoclonic jerk of the arm manipulating a utensil (M. Seino, personal communication, 1999), this phenomenon resembles cortical reflex myoclonus as part of a “continuum of epileptic activity centered on the sensorimotor cortex” (155). It also appears to be another manifestation of triggering of a generalized or bilateral epileptiform response by a local or functional trigger (156), in this case requiring participation of the rolandic region of one or both hemispheres, which may be regionally hyperexcitable in JME (157). The seizures of idiopathic generalized epilepsy may involve only selected thalamocortical networks (158), and this seems especially so in JME (159).
Seizures Induced by Reading
Bickford et al. (160), in 1956, first identified primary and secondary forms of reading epilepsy. The primary form consists of attacks triggered exclusively by reading, without spontaneous seizures. Age at onset is typically between 12 and 25 years. Patients report characteristic jaw jerks or clicks. If reading continues, a generalized convulsion may occur. Prolonged reading-induced partial seizures with ictal dyslexia, bilateral myoclonic seizures, and absences have been reported. The resting EEG pattern is normal, but during reading, abnormal paroxysmal activity is recorded, often consisting of sharp theta activity that may be generalized (142,160–163) or localized to either temporoparietal region, especially on the dominant side (164,165). These abnormalities frequently are correlated with the jaw jerks, and monitoring also shows perioral reflex myoclonus similar to that seen in JME. Bilateral or asymmetric myoclonic attacks or jerks of the arms and head also similar to those of JME may also occur, with bilaterally synchronous spike-and-wave activity.
Patients with primary reading epilepsy are typically developmentally normal, with normal neurologic examinations. No structural lesions have been demonstrated. A family history of epilepsy is common, and familial reading epilepsy has been reported (165,166). Patients with secondary reading epilepsy also have spontaneous seizures without jaw jerking and often have an abnormal baseline EEG. Primary reading epilepsy was classified as an idiopathic, age-related, localization-related epilepsy, but its focal nature has recently been questioned (16,167). Attacks are induced by reading and may be reproduced easily in sensitive subjects. Functional MRI has shown (168) activations in most subjects in areas overlapping or adjacent to those physiologically activated during language and facial motor tasks, including subcortical structures as also noted by Archer et al. (169). Reading epilepsy seems to be an example of activation of a hyperexcitable network, which can produce seizures when sufficient critical mass is incorporated by adequate stimuli to produce a seizure, at times a seizure of apparently generalized epilepsy. We have noted that it may rely on both existing and reorganized functional links between brain regions and need not be confined to physically contiguous brain sites or established neuronal links.
The triggering stimulus in reading epilepsy is unknown, but several authors have speculated about the origin. Bickford et al. (160) proposed that normal sensory stimuli influenced some hyperexcitable cortical focus. Critchley et al. (164) emphasized several factors: the visual pattern of printed words, attention, proprioceptive input from jaw and extraocular muscles, and conditioning. Forster (64) theorized that the seizures were evoked by higher cognitive functions; however, patients with primary reading epilepsy are not photosensitive, deny other precipitating cognitive stimuli, and do not appear to have thinking-induced seizures. Patients with the latter almost always deny activation by reading. A recent detailed study in one patient showed that the alphabetical nature of written stimuli triggered his seizures (170). Another single patient with otherwise clear-cut primary reading epilepsy reported induction by card playing while drinking beer (171). Comprehension of the material being read is essential in some cases and irrelevant in others, suggesting that attention is not sufficient to precipitate seizures. Studies suggest that increased difficulty, complexity, or duration of a task increases the chance of EEG or clinical activation (59,60).
Functional imaging has shown that these seizures result from activation of parts of a speech and language network in both hemispheres (172), confirming that the hyperexcitable neuronal tissue forming the critical mass is not necessarily contiguous but is functionally linked, as discussed by Salek-Haddadi et al. (168), by Rémillard et al. (173), and by Safi et al. (170). A mechanism similar to that in pattern-sensitive epilepsy, in which generalized activity is activated by the occipital cortical stimuli, may operate in some cases of primary reading epilepsy in which bilateral myoclonic attacks or bilaterally synchronous epileptiform activity is triggered.
Primary reading epilepsy generally responds well to valproate, and benzodiazepines or lamotrigine is expected to be useful as well, but patients often decline treatment especially if they have only jaw jerks. Exquisite treatment response to levetiracetam was reported in two patients, one with primary and one with secondary reading epilepsy (174).
Language-Induced Epilepsy
Geschwind and Sherwin (175) described a patient whose seizures were induced by three components of language: speaking, reading, and writing. Some other cases have been reported since. Similar to those in primary reading epilepsy, the seizures consist of jaw jerks, with focal (161,176–178) or generalized (175) abnormal paroxysmal EEG activity during language tasks. In some patients, isolated components of language were the only effective seizure triggers. Writing (179,180), typing (132), listening to spoken language (181), and singing or recitation (182) have been reported as isolated triggers. Writing or speaking may activate patients with reading epilepsy (178,183), and exceptionally, reading epilepsy occurred in a patient who was also activated by card games (172). We consider activation by drawing (184) to be part of seizures induced by thinking, and other patients believed to have language-induced epilepsy may have thinking-induced seizures. This heterogeneity suggests that the definition of a language-induced epilepsy is not clear-cut. Cases may form part of relatively more stereotyped syndromes of reading epilepsy, whose definition should be broadened. Alternatively, Koutroumanidis et al. (172) suggested that primary reading epilepsy might be classified as a variant of a more broadly defined language-induced epilepsy. The association of reflex language-induced epilepsy and idiopathic generalized epilepsy was explored by Valenti et al. (184) and is of interest since some patients with reading epilepsy also seem to have an underlying generalized epilepsy.
Seizures Induced by Music
The rare musicogenic epilepsy consists of seizures provoked by hearing music. The music that triggers seizures is often remarkably specific in any one patient, and no consistent epileptogenic features of musical sound can be identified. A startle effect is not required. Many patients have spontaneous attacks as well. Some attacks can be provoked by music and by nonmusical sounds such as ringing or whirring noises from telephones or vacuum cleaners (6,7). In some patients, an effective musical stimulus often induces emotional and autonomic manifestations before the clinical seizure begins. Patients may report triggers with personal emotional significance. However, in some patients, the triggers have no particular connotations (185), while in others, they may (186). Triggers without particular emotional significance can induce the typical autonomic features before the clinical attack (187,188). Establishment of the seizure as a conditioned response has also been suggested (142,185,187,189), but this view is not generally accepted (190). A case with self-induction possibly motivated by emotional factors has been described (191). Musicogenic attacks may appear only in adulthood, often in the context of a preexisting symptomatic localization-related epilepsy. Many case reports antedate intensive monitoring and modern imaging, but the seizures appear to be simple or complex partial, and epileptiform EEG abnormalities are recorded focally from either temporal lobe but more frequently over the right side. Mesial temporal and lateral temporal seizure onsets have been documented (192).
The pathophysiology of musicogenic epilepsy is obscure. Studies in epileptic subjects not sensitive to music show that musical stimuli may have widespread effects on neuronal activity in human temporal lobes, extending well beyond the rather restricted primary auditory area (193); that different components of music have different effects, possibly with specialized lateralization and localization; and that the effects of music differ from those of speech (194,195). Components of musical stimuli such as melodic contour and perception of unfamiliar pitch patterns are processed by cortical subsystems rather than by a nonspecific music area of the brain (196–198). Functional imaging of musical perception has been reviewed (199). Wieser et al. (200) suggested a right temporal predominance for musicogenic seizures. Right anterior and mesial hyperperfusion during ictal single- photon emission computed tomography has been documented (199,201), and later, detailed coregistration functional imaging supported a privileged role for right temporolimbic activation (202). Zifkin and Zatorre (203) note that more complex musical processing tasks activate more cortical and subcortical territory bilaterally, although with right hemisphere predominance. Hyperexcitable cortical areas could be stimulated to different degrees and extents by different musical stimuli in patients sensitive to these triggers. Gloor (204) suggested that responses to limbic stimulation in epileptic subjects depend on widespread neuronal matrices linked through connections that have become strengthened through repeated use of interest in considering the delay from seizure onset to the development of sensitivity to music and the extent of the networks involved in musical perception.
The extreme specificity of the stimulus in some patients and the delay from stimulus to seizure onset can be useful in preventing attacks, but these seizures usually occur in patients with partial seizures, and appropriate antiepileptic drugs are generally required. Intractable seizures should prompt evaluation for surgical treatment. Furthermore, in musicians, right temporal lobectomy can cause loss of musicality (205).
Seizures Induced by Eating
Boudouresques and Gastaut (206) first described eating epilepsy in four patients who experienced seizures after a heavy meal. Gastric distention may have been at least partly responsible for the attacks (207), but many such seizures occurred early in the meal and were unrelated to gastric distention (208,209). The clinical characteristics are usually stereotyped in individual patients, but there are few common features among patients. Some patients have seizures at the very sight or smell of food, whereas others have them only in the middle of a meal or shortly afterward. In some patients, the seizures may be associated with the emotional or autonomic components of eating; in others, they are associated with sensory afferents from the tongue or pharynx. These seizures have also been documented in young children, in whom they can be mistaken for gastroesophageal reflux (210).
Seizures with eating are almost always related to a symptomatic partial epilepsy. Cases in which the seizures were generalized from onset are exceptional (211). Rémillard et al. (173) suggested that seizures in patients with temporolimbic epilepsy are activated by eating from the beginning of their seizure disorder and continue to have most seizures with meals. In contrast, patients with localized extralimbic, usually postcentral, seizure onset develop reflex activation of seizures later in their course, with less constant triggering by eating and more prominent spontaneous seizures. These patients typically have more obvious lesions and findings on neurologic examination.
The mechanism of eating epilepsy is unclear. Several investigators suggest that interaction of limbic and extralimbic cortices (212) and contributions from subcortical structures, such as from the hypothalamus (67,206,213), are particularly important. Other proposed triggering mechanisms include a conditioned response, mastication (213), stimulation of the esophagus (214), and satisfaction of a basic drive (210). Rémillard et al. (173) suggested that seizures with extralimbic, suprasylvian onset, often involving obvious structural lesions, may be activated by specific thalamocortical afferents. Detailed studies of three male and three female patients with eating epilepsy in India showed ictal rhythmic slowing/fast activity in parietotemporal (n = 2) or frontotemporal (n = 4) regions with subsequent secondary generalization in three. Ictal and interictal SPECT imaging showed changes in frontal lobe (n = 1), anterior temporal lobe (n = 1), and parietoinsular region (n = 1), suggesting that these areas are part of the ictal onset zone (77).
That obvious combinations of several stimuli are required in some cases (215,216
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
