Clinical manifestations

Presentation and course

Visual-induced seizures are the commonest type of reflex seizures. They are triggered by the physical characteristics of certain visual stimuli and not by their cognitive effects. Photosensitivity and pattern sensitivity are the 2 main categories (with frequent overlap) of simple reflex seizures with short (typically within seconds) time between stimulus and response.

According to the predominant visual stimuli, patients with visual-sensitive epileptic seizures can be grossly divided as:

Photosensitive epilepsy. Photosensitive epilepsy is a broad term comprising numerous heterogeneous situations in which seizures are triggered by light (73; 190; 162; 121; 68; 124; 109). It is not an epilepsy syndrome. Photosensitivity is documented with EEG photoparoxysmal responses induced by intermittent photic stimulation (IPS).

Photoparoxysmal responses may be asymptomatic throughout life or manifest with clinical epileptic seizures. It is well documented that photoparoxysmal responses with or without clinical photosensitivity are associated with greatly variable types of epilepsy and mainly juvenile myoclonic epilepsy (30%), Dravet syndrome (70%), Unverricht-Lundberg disease (90%), and a number of autosomal recessive progressive myoclonic epilepsies (188; 147; 73; 07; 179; 67; 155; 101).

Of patients with epileptic seizures and photoparoxysmal responses, 42% have only photically induced reflex seizures without spontaneous seizures (pure photosensitive epilepsy); 40% have spontaneous and reflex photosensitive seizure, and the remaining 18% have spontaneous seizures only (73).

Precipitants of seizures and environmental stimuli.

Flickering lights. Flickering lights is the main triggering stimulus in photosensitive patients, and many artificial or natural light sources can provoke epileptic attacks. Video games, computer screens, television, discotheques, and natural flickering light (across the trees or reflecting from the sea waves) are in that order common precipitants of seizures. Television viewing was by far the commonest of all precipitating factors (television epilepsy) until the introduction of video games, which are now the more frequent precipitants of reflex seizures.

Individuals attending electronic dance music festivals are also at risk of having an epileptic seizure, particularly in festivals occurring in darkness and in those individuals who are sleep deprived and/or are under the influence of ecstasy or similar stimulant drugs (150).

The first reports of photosensitive epilepsy were concerned with patients who had seizures when in bright sunlight (63; 100) or suddenly exposed to bright light (134). Gastaut argued that fluctuation of light stimulus (intermittent light stimulation) was the responsible factor to induce epileptic seizures and that any claims of observing seizure induction by continuous light should probably be discounted as due to interruptions of the light by flutter of the eyelids (58). However, some photosensitive patients are sensitive to eye closure alone in the presence of uninterrupted light (see the article on eyelid myoclonia).

Facilitating factors. Though flickering lights is the principle stimulus to trigger seizures, patients are more vulnerable in the presence of facilitating factors such as deprivation, excessive alcohol consumption, and fatigue (emotional and physical).

Photically induced seizures. Generalized seizures (absences, myoclonic jerks, and GTCS) are more frequent than focal photically induced seizures.

Generalized seizures. Myoclonic jerks, absences, and GTCS, in this order by prevalence, can occur in photosensitive patients. Some patients may have only 1 seizure type, but most have any combination, particularly myoclonic jerks and GTCS. That myoclonic jerks are by far the most common may appear contradictory to the usually stated view that GTCS are more common (55% to 84%) than absences (6% to 20%), focal seizures (2.5%), and myoclonic jerks (2% to 8%) (73). This prevalence is based on clinical historical evidence, which is likely to over-exaggerate GTCS in relation to minor seizure events even though they predominate. Over half of known photosensitive epilepsy patients questioned immediately after stimulation denied having brief but clear-cut seizures induced by intermittent photic stimulation and documented by video-EEG monitoring (86). In my personal experience with video-EEG, photoparoxysmal responses are commonly associated with eyelid manifestations (a blink, fluttering, flickering, myoclonia) and less often with jerks of the head, eyes, body, or limbs.

Absences followed in prevalence, and only 1 patient had an accidental GTCS. Patients are often unaware of the minor seizures although some of these are marked.

Focal seizures. Photically induced occipital seizures are much more frequent than originally appreciated after the use of intermittent photic stimulation in EEG testing.

These may occur alone or progress to symptoms from other brain locations and GTCS. See clinical article on photosensitive occipital lobe epilepsy.

Extraoccipital focal seizures from the onset are exceptional (11).

Subjective symptoms during photoparoxysmal responses are of doubtful significance; some are ictal phenomena, but most are not and many patients cannot tolerate the light at all (181).

Pure photosensitive epilepsy. Pure photosensitive epilepsy is a term used only for patients whose seizures are always photically induced without spontaneous, unprovoked seizures (73). Pure photosensitive epilepsy has a prevalence of 42% amongst the photosensitive epilepsies. The resting EEG is normal (half of the patients) or shows generalized discharges on eye closure (20%) with photoparoxysmal responses occurring in all untreated patients. The seizures are usually infrequent, and the prognosis is often excellent. Avoidance of precipitating factors may be the only treatment.

Pure photosensitive occipital lobe epilepsy is considered in the article on photosensitive occipital lobe epilepsy, in a separate review by Parmeggiani and Guerrini (130), and in a report by Politi-Elishkevich and colleagues (137). Occipital photosensitivity has been documented in adults (124; 96).

In a report on adult-onset photosensitivity, Koutroumanidis and colleagues studied the clinical and EEG data of patients who were referred due to a possible first seizure and who had a photoparoxysmal response on their EEG (96). Patients with clinical evidence of photosensitivity before the age of 20 were excluded. Out of 30 patients, 4 had acute symptomatic seizures, 2 had vasovagal syncope, and 24 were diagnosed with epilepsy. Nine of the 24 patients had idiopathic (genetic) generalized epilepsies and predominantly generalized photoparoxysmal response, but they also had rare photically-induced seizures, whereas 15 had exclusively, or almost exclusively, reflex photically-induced occipital seizures with frequent secondary generalization and posterior photoparoxysmal response. Other important differences included a significantly older age at seizure onset and paucity of spontaneous interictal epileptic discharges in patients with photically-induced occipital seizures; only a quarter of these had occasional occipital spikes, in contrast to the idiopathic (genetic) generalized epilepsy patients with typically generalized epileptic discharges. On the other hand, both groups shared a positive family history of epilepsy, common seizure threshold modulators (such as tiredness and sleep deprivation), normal neurologic examination and MRI, a generally benign course, and good response to valproic acid. Thus, this study demonstrated that photosensitivity can first occur in adult life and manifest either as idiopathic (possibly genetic) photosensitive occipital epilepsy with secondary generalization or as an EEG, and less often, it can occur as a clinical/EEG feature of idiopathic (genetic) generalized epilepsies. Identification of idiopathic photosensitive occipital epilepsy fills a diagnostic gap in adult first-seizure epileptology and is clinically important because of its good response to antiepileptic drug treatment and fair prognosis (96).

Pattern-sensitive epilepsy. Pattern-sensitive epilepsy refers to epileptic seizures induced by patterns, typically stripes (73; 181; 18; 142; 143; 22). It is not a particular epileptic syndrome, and it is closely related to photosensitivity. Almost all patients with clinical pattern sensitivity epilepsy show photoparoxysmal responses.

Conversely, 30% of clinically photosensitive patients are also sensitive to stationary patterns of stripes, and 70% to appropriately vibrating patterns. Patterns enhance the effect of photic stimulation, whether under test conditions or in real life (82). Pattern sensitivity without photosensitivity, sensitivity to nongeometric patterns, and self-induced pattern-sensitive epilepsy are rare.

Demographic data. Pure pattern-sensitive epilepsy with clinical attacks induced only by patterns is rare:

This is despite the relatively high incidence of pattern-induced EEG paroxysmal activity in photosensitive patients.

Clinical manifestations. The majority of patients exhibit absence, myoclonic seizures, or GTCS alone or in various combinations (73; 142; 143; 22). In 1 study, focal seizures with complex visual hallucinations followed by unresponsiveness and automatisms occurred in around 11% of patients (143). However, in another study though the most common type of spontaneous seizures were generalized (60%), reflex seizures were more frequently focal (74%) with elementary or complex visual hallucinations (22).

Self-induced pattern-sensitive epilepsy has been reported (113; 102).

Environmental stimuli. Environmental stimuli that induce seizures in pattern-sensitive patients are those that best match the properties of the provocative patterns used in relevant EEG testing and best suit and create the conditions of their spatial and directional presentation to the eyes. These are striped clothes (eg, shirts, jackets, or ties), escalators, wallpaper, furnishings, venetian blinds, air-conditioning grills, and radiators. Any activity visually involved with these patterns, such as ironing, is likely to induce seizures. Less direct, but often very significant, is the role of patterns in more complex stimuli, such as TV viewing and video games (73; 181; 18).

Pattern is recognized as a seizure precipitant less often by patients, caregivers, and physicians than environmental flicker and specific agents, such as the TV, discotheque lighting, or video games. Direct questioning implicates pattern as a seizure trigger in 6% to 30% of photosensitive individuals (186; 83).

Pathophysiology. Elaborate and intelligent methodological studies, mainly in patients with photically induced seizures, revealed many aspects of pattern seizure susceptibility and its pathophysiology (181; 18). See pathophysiology of photosensitive epilepsy.

Diagnostic procedures. The EEG with appropriate pattern presentations is the key test.

As in photosensitive patients, binocular is much more potent than monocular stimulation, and the patient should fixate on the presenting patterns. See diagnostic procedures of visual-induced seizures.

Video-induced seizures. The term “video-induced seizures” is preferable to “video game epilepsy” because this is not a syndrome. There is clearly heterogeneity in seizure types, seizure syndromes, precipitating and facilitating factors, and underlying mechanisms.

Video games comprise a multibillion-dollar industry with the demographics of players extending in nearly all ages, including significant numbers of women.

Video-induced seizures refers to epileptic seizures precipitated by playing interactive computerized “video games,” a term used to include not only those games using an interlaced video monitor but also small, hand-held, liquid crystal displays, arcade games, smart phones, and other new electronic devices that use noninterlaced displays and higher refresh rates than television (16; 47; 48; 65; 90). With video games pervading life today, the number of patients continues to rise. The risk has been highlighted by media reports. Manufacturers now warn of the danger and have sponsored research into the subject. The results of these studies prior to 1999, mainly sponsored by Nintendo Corporation under the auspices of the Japanese Epilepsy Association, were published in Epilepsia (53; 87; 104; 133; 145).

Clinical aspects. There are many mechanisms by which video games may induce seizures. These are photosensitivity, pattern sensitivity, emotional and cognitive excitation, and proprioceptive stimulation (movement). All these factors, alone or in combination, may induce a seizure. Fatigue, sleep deprivation, and hunger are significant contributing or facilitating factors (16; 47; 48; 65).

Video-induced seizures are typically absences, jerks, and GTCS. Occipital lobe seizures with or without (mainly without) photosensitivity is the second most frequent type (29%); this is more often the case in handheld LCD displays and arcade games.

Photosensitivity, which often combines with pattern sensitivity, is considered the main provocative factor in video-induced seizures (140). However, only 70% of patients with well-documented video-induced seizures are photosensitive on intermittent photic stimulation. One third of patients are not photosensitive in the sense that appropriate intermittent photic stimulation does not evoke photoparoxysmal responses. Four particular groups of non-photosensitive patients have been identified (47; 48):

Idiopathic generalized epilepsy without photosensitivity. This group consists of patients with various forms of idiopathic generalized epilepsy who also have seizures while playing video games though they are not photosensitive on intermittent photic stimulation in the EEG laboratory. These probably account for one third of the patients. Cognitive and emotional factors seem crucial provocative agents.

Typical in this sense is 1 of my patients, an intelligent 45-year-old man with juvenile absence epilepsy, which is well controlled with appropriate medication. On 5 occasions he had a GTCS at the same part of a car racing video game; the GTCS occurred at the moment the car was about to crash.

Idiopathic occipital epilepsy. This group consists of patients with focal seizures originating in the occipital lobes that manifest with visual hallucinations and visual deficits. These patients are generally not photosensitive in the laboratory but have spontaneous occipital spikes in their resting EEG. They constitute a third of all patients with video game seizures. The seizure-provoking mechanism is unknown but might relate to an association with playing games in arcades.

Low-threshold to seizures. Patients in this group are young adults with 1 to 3 video-induced seizures occurring only when a number of seizure precipitating factors cluster together in the time of the incident (47; 48). These are sleep deprivation, thirst, hunger, mental and emotional heightening, and prolonged playing of exciting and provocative games. In such patients, it is unlikely that 1 of these factors on its own is sufficient to provoke seizures but that a number of them together is.

A typical example is a 23-year-old man who has had 2 GTCS, one preceded by an absence seizure, while playing different video games. Three EEG examinations have all been normal, without photosensitivity. Fatigue, prolonged playing, hunger, and thirst were identified as possible factors.

Pure pattern epilepsy without photosensitivity. This may be extremely rare (74).

Practical implications. The practical implication of these observations is that not all patients who have seizures while playing video games will be helped by the advice recommended for photosensitive patients. A thorough clinical and EEG evaluation is needed to identify likely precipitants and enable individual guidelines to be offered (48).

Risk of video-induced seizures in non-photosensitive patients with epilepsy. Millet and colleagues examined systematically whether exposure to video game material is a risk factor for seizures in patients with chronic epilepsy without visual sensitivity (104). Of 212 chronic epilepsy patients without EEG photosensitivity, 13 had seizures during periods of video game play and 12 during alternative leisure. The authors concluded that seizures during video game play in more than 95% of the epilepsy population without visual sensitivity are most likely to represent a chance occurrence, although, as always, each individual should be carefully assessed.

Television-induced seizures. TV-induced seizures were the most common form of photosensitive epilepsy prior to the introduction of video games. TV-induced seizures mainly affect children aged 10 to 12 years. There is a two-fold preponderance of girls. Seizures are more likely to occur when the patient is watching a faulty (ie, flickering) TV set or is sitting very near to the TV screen (less than 1 meter distance).

Screen content may be more important than the characteristics of the screen itself (23). Programs with flickering lights are particularly dangerous, and occasionally their effect on eliciting seizures may reach epidemic proportions. TV-induced seizures came to global public attention in December 1997 when approximately 700 children and adolescents in Japan were admitted to the hospital because of seizures provoked by a Pokémon (Pocket Monster) television cartoon (78; 161). The offending sequence was caused by a character’s actions in a rocket launch scene with a flashing red then blue screen, changing 12.5 times per second for 4 seconds.

Clinical manifestations. Generalized seizures (myoclonic jerks, GTCS, or absences--in that order) are more frequent than focal visual seizures.

Myoclonic jerks often precede the GTCS, and the patient history often reveals that these may have occurred in the past without GTCS:

A substantial number of these patients also have spontaneous attacks. In pure TV epilepsy, 1 or a few overt TV-induced seizures occur without evidence of any other type of spontaneous seizure or seizures induced by other means.

Ten percent of patients report being “drawn like a magnet,” and when they reached a certain nearness to the screen, the GTCS occurs. This is called “compulsive attraction.”

There are a few reports of TV-related self-induced seizures (28).

Technical aspects and explanations of television-induced seizures. A television picture consists of 2 frames of horizontal lines that intercalate and alternate at half the frequency of the mains AC supply. They are produced by changes in the intensity of the spot of light that draws the lines across the television, and these lines are drawn alternately (ie, the first, third, fifth, etc., then the second, fourth, sixth, etc.). Where the mains AC frequency is 50 Hz (Europe and Japan), each alternate set of lines takes 0.02 seconds to complete, and the 2 lines produce an alternating flicker at 25 Hz. Where the mains AC frequency is 60 Hz (United States), this flicker has a frequency of 30 Hz. This difference in the frequency of the alternating current in the mains electricity supply explains why proportionally more people have television-induced seizures in Europe and Japan than in the United States (73). Special 100 Hz television screens, marketed in Europe, reduce the risk of television-induced seizures (146).

Television-induced seizures, however, are not only related to alternating current frequency flicker. Wilkins and colleagues studied patients who were not sensitive to the alternating current frequency flicker but who responded to the vibrating pattern of interleaved lines at half the alternating current frequency (25 Hz in Europe and 30 Hz in North America) to which about 75% of photosensitive subjects are sensitive and which can be discerned only in close proximity to the screen (187).

Color is important even without luminance changes; photoparoxysmal EEG responses can be elicited in sensitive subjects by noncolor-opponent stimuli even if they are isoluminant (72). Sensitivity is greater with red stimulation at wavelengths greater than 700 nm, and red stimulation was important in the Japanese cartoon incident. Red-cyan flicker, even when isoluminant, is reportedly even more provocative of epileptic discharge (153). Using a special color stimulator, Parra and colleagues suggested that color sensitivity follows 2 different mechanisms activated by different frequencies of stimulation: less than or greater than 30 per second (132). They suggested a theoretical basis for using colored spectacles to treat photosensitive patients.

Television-induced seizures are more likely to occur when the set is being watched from a close distance, less than 1 meter away from the screen. The main reasons for this are that at this distance (186; 187):

Management. See management section of the article discussing visual-induced seizures.

Self-induced seizures. Patients with all types of visually induced seizures may willingly induce attacks to themselves.

Self-induction is a mode of seizure precipitation employed by entirely normal or mentally impaired patients to produce seizures for themselves. Maneuvers for self-induction aim to provoke a seizure by producing optimal conditions of stimulation by flickering light (self-induced photosensitive epilepsy), patterns (self-induced pattern-sensitive epilepsy), proprioceptive stimuli, or higher brain functions (self-induced noogenic epilepsy). Intensely pleasurable sensations have been reported with these types of seizure, and some patients induce seizures to relieve stress or to gain attention (164).

Demographics. The exact prevalence of self-induced seizures is difficult to determine and may have been overestimated. In many cases, “self-induced” behaviors do not appear to be willful or consciously generated, and eyelid blinking or forced eyelid deviation towards the light has been unquestionably taken as evidence of self-induction. Of 442 patients with onset of nonfebrile seizures from birth to 15 years of age, 5 (1.1%) had self-induced seizures (119).

Age of onset varies from infancy to mainly early childhood, and females (70% to 80%) predominate.

Etiology. Etiology is unknown. It happens in both structural and idiopathic cases of photosensitive epilepsy.

Clinical manifestations. Absences and myoclonic jerks are the most common seizures in self-induction. GTCS, when they occur, are usually accidental events that were not desired. GTCS usually follow deliberately self-induced absences or jerks.

The objective of self-induced seizures is relief of tension and anxiety or escape from a disturbing situation.

One maneuver for self-induction in photosensitive epilepsy is looking at a bright light source, usually the sun, and voluntarily waving the abducted fingers in front of the eyes (sunflower syndrome) in order to produce optimal intermittent photic stimulation. Other techniques are: repetitive opening and closing of the eyes or lateral or vertical rhythmic movements of the head in front of a bright light source; making the television picture roll; quickly changing television channels while watching from a close distance; and playing video games.

A controversial aspect of self-induction is whether slow and forcible eye-closure is another maneuver used by photosensitive patients (34; 15) or part of the seizure (127). (See Clinical vignette 1.)

Differential diagnosis. Self-induced maneuvers should be differentiated from tics as well as genuine ictal phenomena, such as eyelid myoclonia or eye closures of occipital seizures (127).

Early forced eyelid blinking and flutter, eyelid jerks, and oculoclonic activity may be ictal manifestations of the occipital lobes that may not show in surface EEG but can be documented with deep stereo-EEG recordings.

Eyelid blinking and gaze fixation to light may be a normal “attraction movement” when light is presented and other manifestations of the optic fixation reflexes when volitional movements of the eyes are unattainable or weak.

Blinking functions may be a complex indicator of phasic responses to stress, such as that produced by listening to emotionally laden words.

Management. See management section of article on visual-induced seizures.

Fixation-off sensitivity. Fixation-off sensitivity (FOS) is a term coined by Panayiotopoulos to denote the form(s) of epilepsy or EEG abnormalities that are elicited by elimination of central vision and fixation (115; 116; 121). “Elimination of central vision and fixation”’ is a specific precipitating stimulus, which, even in the presence of light, induces high-amplitude occipital or generalized paroxysmal discharges. (See clinical vignette 2.)

Fixation-off sensitivity is suggested in the routine EEG by abnormalities, which consistently occur as long as the eyes are closed but not when the eyes are opened.

Clinical and EEG correlations in patients with fixation-off sensitivity. Fixation-off sensitivity is a reflex EEG activation with some preference for certain epileptic conditions. From clinical and video-EEG documentation, there are 3 types of patients with seizures and EEG abnormalities of fixation-off sensitivity (116):

(1) Patients with occipital paroxysms, such as those seen in EEG of some patients with Panayiotopoulos syndrome and more frequently in patients with idiopathic childhood occipital epilepsy of Gastaut, which are the model examples of fixation-off sensitivity (124; 80). It was in these cases that fixation-off sensitivity was first documented as a new type of activating stimulus in reflex epilepsies (114). Fixation-off sensitivity abnormalities are mainly localized in the occipital regions and are not associated with overt ictal clinical manifestations.

(2) The second type constitutes a rare but “pure” and distinct clinical form of fixation-off sensitivity, cryptogenic generalized epilepsy. Patients are women of borderline normal intelligence with frequent eyelid myoclonia (with or without atypical absences), absence status epilepticus, and GTCS. The eyelid myoclonia manifests with fast, small amplitude, clonic movements of the eyelids associated with tonic spasm of the eyelids and eyes that occasionally spread to the neck muscles (115). Absence status epilepticus is preferentially catamenial (02; 105). The EEG–FOS abnormalities consist mainly of diffuse alpha-like rhythms at 7 Hz, mixed with bisynchronous sharp and spike/polyspike components. These are often associated with clinical ictal manifestations.

Patients are not photosensitive and differ markedly from those with Jeavons syndrome (eyelid myoclonia with absences).

(3) The third type concerns patients with idiopathic generalized epilepsy and photosensitivity without fixation-off sensitivity-related overt clinical ictal manifestations (03; 121).

In the second and third types, the typical abnormalities related to fixation-off sensitivity are mainly diffuse or generalized, with “dropout” in sleep stages simultaneous with the alpha rhythm.

However, fixation-off sensitivity has also been described in patients with structural epilepsy (99; 98; 46; 75; 148; 20; 45) and more recently in a child with CHD2 mutation and mild developmental impairment (26). Furthermore, fixation-off sensitivity may occur in individuals without seizures (95; 97). In such an asymptomatic adult with fixation-off sensitivity, continuous bilateral occipital paroxysms during elimination of central vision were associated with transitory cognitive impairment, demonstrated by neuropsychological testing (97).

In a report, Koutroumanidis and colleagues aimed to define the spectrum of the epileptic syndromes and epilepsies (other than the idiopathic epilepsies of childhood with occipital paroxysms) that can be associated with fixation-off sensitivity, delineate the electrographic types of fixation-off sensitivity abnormalities, identify the patterns that can be associated with clinical seizures, and examine whether there may be a pure form of fixation-off sensitive epilepsy (95). Nineteen of about 8,500 patients had 1 or more video-EEG with fixation-off sensitivity, yielding an approximate incidence of 0.2%. From the 14 patients with full clinical and EEG data available, 12 had various epilepsies that included idiopathic generalized epilepsy phenotypes (7), symptomatic or probably symptomatic focal (3), cryptogenic generalized (1), and adult-onset idiopathic photosensitive occipital (1); 2 patients had no seizures. Seven patients (50%) were photosensitive. Fixation-off sensitivity-related EEG abnormalities were occipital in 6 patients, generalized in 8, and generalized with posterior emphasis in 2. These were associated with habitual seizures in 7 patients, but actual fixation-off sensitivity-induced seizures (absences) were documented with video-EEG in only 1 patient; 3 others had some historical evidence suggesting that, under some circumstances, their fixation-off sensitivity-EEG abnormalities might generate clinical seizures (95).

In another study, Fattouch and associates retrospectively evaluated the clinical data, EEG, and MRI findings of patients with epilepsy and FOS persisting in adult life to better define the spectrum of syndromes (45). They selected 15 consecutive patients (12 female/3 male; age range 19 to 59 years). They found a female prevalence (F/M = 12/3). Eight patients presented both simple and complex focal seizures, whereas 7 had only complex focal seizures. Focal seizures evolved into generalized seizures/hemiconvulsions in 9 cases. The FOS pattern consisted of spike-and-wave and slow-wave abnormalities with posterior localization (bilateral in 8 and monolateral in 7). Seizures were recorded in 10 out of 15 patients. All showed a posterior onset (bilateral in 2/left in 2/right in 6). FOS was prevalent in symptomatic epilepsy (cortical malformations in 7; celiac disease in 3; calcified vascular malformation in 1). One patient presented cryptogenic epilepsy, and only 3 presented with idiopathic epilepsy (Gastaut syndrome). The authors concluded that FOS can be observed in adult life in idiopathic epilepsy, representing the "prolongation" of the same phenomenon developed during childhood. Nevertheless, it often represents the EEG expression of symptomatic epilepsies (cortical malformations/celiac disease) (45).

Wang and colleagues described the electroclinical features of idiopathic generalized epilepsy patients presenting with fixation-off sensitivity (180). In a 4-year period they found only 8 patients with fixation-off sensitivity and idiopathic generalized epilepsy; 4 with eyelid myoclonia/Jeavons syndrome, 2 with juvenile myoclonic epilepsy, 1 with photosensitivity epilepsy, and 1 with epilepsy with generalized tonic-clonic seizures only. Fixation-off sensitivity coexisted with photosensitivity in 6 patients as independent EEG features. Neuropsychological testing revealed transitory cognitive impairments associated with fixation-off sensitivity.

Pathophysiology. The underlying mechanisms of fixation-off sensitivity are not known, but they may be related to an abnormality of the alpha-rhythm generators (116). Fixation-off sensitivity has the opposite characteristics of photosensitive epilepsies (Table 1), but conversion from 1 to the other may rarely occur (112).

Table 1. Fixation-Off Sensitivity (FOS) Versus Photosensitivity Epilepsy

(124)

Fixation-off sensitivity paroxysms studied with functional MRI were correlated with activation of parieto-occipital and frontal brain areas (97) and a significant increase of blood oxygen level-dependent signal in the extrastriate cortex (77). Magnetoencephalography of visual-evoked fields in fixation-off sensitivity revealed abnormal activation of the visual corticocortical pathway via the insular cortex (98). Strigaro and colleagues studied cortical excitability changes associated with fixation-off sensitivity in a woman with generalized fixation-off sensitivity (158). They measured by transcranial magnetic stimulation (TMS) the excitability level of the primary motor area and explored her visual system by pattern-reversal and flash visual-evoked potentials. They found that both outside and within fixation-off sensitivity the cortical silent period was dramatically short, indicating defective GABA-B inhibition as a persistent background factor. The same was true for the short-interval intracortical inhibition, a TMS marker of cortical GABA-A inhibition. The fixation-off sensitivity state corresponded then to a pathologic enhancement of intracortical facilitation, a TMS marker of Glu/Asp transmission. During fixation-off sensitivity, the flash visual-evoked potentials exhibited a hugely enhanced afterdischarge, expressing a pathologic overactivity of secondary visual areas. The authors concluded that these findings support a grossly imbalanced cortical excitability, in both the frontal and posterior areas, as an important correlate of fixation-off sensitivity.

Scotosensitive epilepsy. Scotosensitivity (skotos in Greek means darkness) denotes forms of epilepsy, seizures, or EEG abnormalities that are elicited by the complete elimination of retinal stimulation by light. Pure scotosensitive patients are rare (09). Most patients described as scotosensitive probably have fixation-off sensitivity (116).

Techniques for documenting fixation-off sensitivity (116; 124; 93; 94). First, it is essential to confirm that the EEG abnormalities observed in routine EEG recording are related to the eyes-closed state.

The patient is asked to open and close his or her eyes every 5 seconds, 6 times consecutively. Instructing the patient to look at a fixed point, such as the tip of a pencil, ensures fixation in the eyes-opened state (116).

Then, fixation-off sensitivity is evaluated by instructing the patient to perform the same sequence of eyes-opened and eyes-closed states in conditions that eliminate central vision and fixation. There are many practical ways to achieve this, such as underwater goggles covered with opaque tape (this achieves complete darkness) or semitransparent tape (which allows light in, but obscures any other visual input).

Warning. Complete darkness can be difficult to achieve in routine EEG departments. Even a small spot of red light on which the eyes may fixate can totally inhibit EEG abnormalities induced by complete darkness. Switching off the lights in the EEG recording room is not adequate and may explain conflicting results in the literature. Complete darkness can be produced with underwater goggles covered completely with opaque tape (116).

Prognosis and complications

Prognosis of patients with visual-induced seizures varies considerably depending on the underlying conditions. It is excellent in some and severe in others.

The overall view is that pure photosensitive epilepsy as a whole has a good prognosis for seizure control with or without antiepileptic drug treatment, but photoparoxysmal responses persist in adult life (71; 173). In 1 of the largest (100 photosensitive patients) and longest follow-up studies (14 years average duration of follow-up and mean age of 27 years), the following results were obtained (71):

However, different prognoses are discussed in the clinical summaries for eyelid myoclonia, idiopathic photosensitive occipital lobe epilepsy, juvenile myoclonic epilepsy, Dravet syndrome, Unverricht-Lundborg disease.

Prognosis of pattern-sensitive epilepsy is also uncertain though a good prognosis is usually reported (143; 22). Brinciotti and associates studied 35 patients with follow-up of more than 5 years. The epilepsy was generalized in 18 cases (17 idiopathic, 1 structural) and focal in 17 (13 idiopathic, 4 structural). Five patients (14%) had only reflex seizures. Patients with only reflex seizures were instructed to avoid precipitating stimuli and were not treated with antiepileptic drugs. Treatment was gradually withdrawn in 10 of 30 treated patients, with relapse in only 2 cases. At the end of follow-up, 28 patients (80%) were seizure-free (22). In the study of Radhakrishnan and colleagues, during a median follow-up period of 15.7 years, 25 (45.5%) of 55 patients who were followed up for 5 or more years achieved complete seizure remission. The median age at remission was 24.4 years. The absence of progressive neurologic disease was correlated significantly with remission; a family history of seizures showed a trend in favor of remission. More than two thirds of the patients did not consider the seizures an impediment to their family life or to educational and occupational achievements (143).

Complications of convulsive seizures induced by visual stimulation are the same as those of spontaneous convulsive seizures and may exceptionally be severe or lethal.

Clinical vignette

Vignette 1. At the age of 5 years, this now 39-year-old woman experienced onset of frequent seizures manifested with brief but marked eyelid myoclonia and absences with mild or moderate impairment of consciousness. Absences improved, but eyelid myoclonia continued daily with ethosuximide treatment. At the age of 26 years, an attempt by a neurologist to substitute ethosuximide with carbamazepine resulted in nonconvulsive status epilepticus. With continuous eyelid myoclonia and absences, she became confused: “My eyes were continuously jerking. I was unable to look after myself and was drowsy and off work for a few days.”

She suffered a total of 6 GTCS throughout her life, starting at the age of 13 years. Two GTCS were induced by lights, and the others occurred after sleep deprivation, alcohol indulgence, or inappropriate change of medication. Eyelid myoclonia and GTCS occurred mainly in the morning after awakening. Seizures improved dramatically when valproate was added to ethosuximide at the age of 31 years, but she continued to have brief seizures of eyelid myoclonia without absences or GTCS. Ethosuximide was later replaced by clonazepam 0.5 mg nightly, which controlled the eyelid myoclonia. At 39 years of age, she has been seizure-free for 4 years on clonazepam 0.5 mg and valproate 2000 mg daily.

Over the years, a number of experts suggested that she was self-inducing her seizures. I have frequently questioned her regarding self-induction, which she thoroughly denies. “Why?” she said “It gives me no pleasure and it is socially embarrassing.” Similar clinical and EEG seizures also occurred while the eyes were closed as well as that the onset of the generalized discharges preceded the eyelid myoclonic jerks. Furthermore, it is unlikely that self-induction would be attempted in social situations, such as at her wedding.

Vignette 2. This normal boy with childhood occipital epilepsy of Gastaut had, from the age of 10 years, experienced:

(1) Ictus amauroticus--brief, infrequent episodes of complete blindness without warning or impairment of consciousness.

(2) Visual seizures--frequent (1 to 3 per week), transient visual disturbances lasting for 10 to 30 seconds.

It looked like a rectangle filled with colored small circles. This time I saw the colors. They were blue, green, red and yellow. While I was reading, I started seeing the words all stuck together. I blink a lot to see more clearly. It is a familiar vision, sometimes bright or dark. It replaces, obscures the real images. They are large objects, probably people, which I cannot identify. They are always in my right eye and draws my right eye and my head to the right.

(3) Syncope-like epileptic seizures--four brief (1 to 2 minutes) episodes of loss of consciousness without convulsions while sitting or standing. He falls down and becomes unresponsive, but there are no postictal symptoms. One episode witnessed by a physician who described it as, “clumsiness, vacant, unresponsive for a minute or so. No convulsions.” Another seizure was witnessed by his parents: “he was next to us in a shop. We heard a bang and saw him on the ground. Color white not blue. He was out for a few seconds.”

No further episodes occurred after initiation of treatment with carbamazepine, which was stopped 3 years later. At last follow-up at 18 years of age, he was well and a good student; he was taking no treatment and had normal EEG and brain MRI.

Biological basis

Etiology and pathogenesis

Significant insights on the etiology and genetics of visual-induced seizures and particularly photosensitivity have been gained through extensive animal and human studies (109). However, there are still important matters that are not yet fully understood.

Genetics. Photosensitivity appears to be genetically determined. In particular, the genetic basis for photoparoxysmal responses is well documented. Photosensitivity occurring in patients with identifiable epileptic syndromes is inherited separately from the other epileptic disorders.

A single gene for photosensitivity has not yet been identified, and the genetics of human photosensitivity probably involve several genes in different chromosomes. However, autosomal dominant inheritance with reduced penetrance was proven in several families with photosensitivity (79).

A family with a novel autosomal dominant familial epilepsy syndrome "myoclonic occipital photosensitive epilepsy with dystonia" has been described (149). This involves a spectrum of phenotypes from juvenile myoclonic epilepsy, sometimes overlapping with idiopathic photosensitive occipital epilepsy to progressive myoclonus epilepsy with paroxysmal dystonia.

Photoparoxysmal responses. It is likely that photoparoxysmal responses follow an autosomal dominant transmission with age-dependent penetrance.

In a large-scale study, 32 clinically photosensitive mothers had 67 children during the follow-up period; 13 children (20%) had photoparoxysmal responses, and 4 also had clinical photosensitive seizures (71).

In another important study in families with a single photosensitive parent, photoparoxysmal responses were significantly more common in 5- to 10-year-old siblings of proband offspring from a parent with a photoparoxysmal responses (50%) than in siblings of photoparoxysmal response-positive children from parents without photoparoxysmal responses (14%) (179).

The 2.5-fold higher prevalence of photoparoxysmal responses in girls than boys (73) indicates that hormonal or specific sex-chromosomal properties may be important (155; 167; 171). The incidence of epileptic seizures is elevated in female relatives of photoparoxysmal response carriers (when siblings of the mother and father are compared) (41).

That external factors may play a role in the clinical expression of photosensitive epilepsy has been exemplified by a pair of monozygous twin brothers. Though both brothers had photoparoxysmal responses, only 1 of them developed clinical photosensitivity after a period of weekly exposures to high-intensity light flashes (37).

Molecular genetic studies on EEG photoparoxysmal response identified putative loci on chromosomes 6, 7, 13, and 16 that seem to correlate with peculiar seizure phenotype (79).

Genetics of photoparoxysmal responses in epileptic syndromes. It is well documented that photoparoxysmal responses (PPR) with or without clinical photosensitivity are associated with greatly variable types of epilepsy and mainly with idiopathic generalized epilepsy (eg, juvenile myoclonic epilepsy), genetic epileptic syndromes (eg, Dravet syndrome), and a number of autosomal-recessive progressive myoclonic epilepsies (155). Therefore, the molecular genetics of photoparoxysmal responses may also vary.

Pinto and colleagues collected 16 photoparoxysmal response-multiplex families of mainly idiopathic generalized epilepsy with myoclonic jerks and conducted a genome-wide linkage scan using a broad model (all PPR types) and a narrow model (exclusion of PPR types I-II and the occipital epilepsy cases) of photoparoxysmal responses (136). They found empirical genome-wide significance for linkage for 2 chromosomal regions, 7q32 at D7S1804 and 16p13 at D16S3395, respectively. These genomic regions contain genes that could be important for the neuromodulation of cortical dynamics in humans, such as the genes encoding the metabotropic glutamate receptor 8 (GRM8) and the cholinergic-muscarinic type 2 acetylcholine receptor M2 (CHRM2) (131).

Tauer and colleagues, in an effort to explore genetic relations between photoparoxysmal responses and idiopathic generalized epilepsy, studied 60 families with at least 2 siblings displaying photoparoxysmal responses; 19 families with predominantly pure photoparoxysmal responses and photosensitive seizures (PPR-families) and 25 families in which photoparoxysmal response was strongly associated with idiopathic generalized epilepsy (PPR/IGE-families) (165). They found 2 PPR-related susceptibility loci, depending on the familial background of idiopathic generalized epilepsy. The locus on 6p21.2 seemed to predispose to PPR itself, whereas the locus on 13q31.3 also confers susceptibility to idiopathic generalized epilepsy. In particular, MOD score analyses provided significant evidence for linkage to the region 6p21.2 in the PPR families (empirical p = 0.00004) and suggestive evidence for linkage to the region 13q31.3 in the PPR/IGE families (p = 0.00015), both with a best-fitting recessive mode of inheritance. In the PPR/IGE families, linkage evidence was even stronger (p = 0.00003) when the trait definition was broadened by idiopathic generalized epilepsy traits.

De Haan and colleagues suggested oligogenic inheritance in a well-studied family with photosensitivity and juvenile myoclonic epilepsy (37). According to the authors, the clinical phenotype in the offspring could be explained by a combination of photosensitivity and epilepsy traits that segregated independently of each other.

De Kovel and associates performed a whole-genome linkage scan for epilepsy-related photosensitivity (38). They combined 2 previously published genome-wide linkage studies (which have loci for photoparoxysmal responses at 6p21, 7q32, 13q13, 13q31, and 16p13) augmented with additional families in a mega-analysis of 100 families. Nonparametric linkage analysis identified 3 suggestive peaks for photosensitivity, 2 of which were novel (5q35.3 and 8q21.13), and 1 has been found before (16p13.3). No evidence for linkage was detected at 6p21, 7q32, 13q13, and 13q31. The authors concluded that different family data sets are not linked to a shared locus. Detailed analysis showed that the peak at 16p13 was mainly supported by a single subset of families whereas the peaks at 5q35 and 8q21 had weak support from multiple subsets.

Dibbens and associates screened NEDD4-2 (Neuronally Expressed Developmentally Downregulated 4) for mutations in a cohort of 253 families with idiopathic generalized epilepsy (39). They identified three NEDD4-2 missense changes in highly conserved residues, S233L, E271A, and H515P in families with photosensitive generalized epilepsy, and concluded that photosensitive epilepsy may arise from defective interaction of NEDD4-2 with as yet unidentified accessory or target proteins.

A trend toward association of transient receptor potential cation 4 (TRPC4) variants and photoparoxysmal responses of idiopathic generalized epilepsy has been reported (174).

Taylor and colleagues used family studies to investigate the clinical genetics of photosensitivity to understand the interrelationship of different photosensitive epilepsy syndromes (166). Twenty-nine families were recruited in which at least 2 members had idiopathic epilepsy and either clinical or electrical photosensitivity on EEG studies. The authors performed electroclinical analysis of these individuals and all other affected family members and analyzed the phenotypic patterns in families. They found an earlier age at seizure onset in photosensitive patients compared with nonphotosensitive individuals. A significant female bias for photosensitivity was confirmed. All subjects with visual seizures were photosensitive. Subjects could be classified into 3 main photosensitive phenotypes: genetic (idiopathic) generalized epilepsies (GGE), idiopathic photosensitive occipital epilepsy (IPOE), and mixed GGE/IPOE. Within each category, subjects with purely photosensitive seizures were observed. The authors also reported a distinctive syndrome of early-onset photosensitive absence epilepsy, with onset beginning by 4 years of age, which was more refractory than childhood absence epilepsy. It was concluded that the clinical genetics of the idiopathic photosensitive epilepsies show a phenotypic spectrum from the GGEs to IPOE with overlap between the focal features of IPOE and all the GGE syndromes. Shared genetic determinants are likely to contribute to the complex inheritance pattern of photosensitivity, IPOE, and the GGEs (166).

CHD2 mutation is the first identified cause of eyelid myoclonia with absences (54). Furthermore, unique CHD2 variants are also associated with photosensitivity in common epilepsies (54).

Localization and pathophysiology of photosensitivity. Photosensitivity results from functional abnormalities in the cortical mechanisms that control the response to strong visual stimulation (185). It is now well documented that the visual cortex is the primary site of epileptogenesis in occipital photosensitivity and pattern seizure sensitivity. This may also be true for the onset of photically-induced generalized seizures in syndromes of idiopathic generalized epilepsy.

The pathophysiology of human photosensitivity has gone through several stages. Initially, the predominant view was that it is “centrencephalic” (generalized epilepsy) with the nonspecific thalamocortical reticular system activated from the lateral geniculate body (14; 56). This view, which dominated the literature at that time, was mainly based on the findings that photoparoxysmal responses are usually synchronous and generalized. The first clear evidence of the occipital origin of photoparoxysmal responses came from our studies of photically-induced occipital spikes often preceding generalized photoparoxysmal responses (128; 110; 110).

Evidence that occipital spikes are preferentially elicited in photosensitive patients documents the primary or exclusive role of the occipital cortex in the initiation of photically induced seizures. This may be the only cortical region involved in occipital photosensitive seizures or the initial trigger zone of generalized photosensitive epilepsy. The visual stimuli eliciting occipital spikes depend on the number of flashes of light per second (or the number of pattern image changes per second), spatial frequency, orientation, contrast, and the line width ratio (73). All these factors indicate that the visual cortex is involved, and this is the earliest site at which integration occurs.

The primary role of the visual cortex in photosensitive epilepsy has been confirmed with elaborate documentation of pattern sensitivity, mainly in patients with photically induced seizures (182; 184; 181; 18; 143). This has revealed many aspects of pattern seizure susceptibility and its pathophysiology:

Generalized seizures can occur if normal excitation of visual cortex involves a "critical mass" of cortical area with synchronization and subsequent spreading of excitation from the occipital lobe trigger.

Present knowledge on pathophysiology of epileptic photosensitivity points to 2 types of mechanisms--mediated by the magnocellular and parvocellular systems--that contribute either synergistically or independently to elicit a photoparoxysmal response (85). Generalized photoparoxysmal responses and intermittent photic stimulation-induced occipital spikes appear to be generated independently by the parvocellular and magnocellular visual systems, respectively (72).

Color sensitivity depends on 2 mechanisms: 1 related to color modulation, intervening at low frequencies and the other dependent on single-color light intensity modulation and related to white light sensitivity that is activated at higher frequencies (132).

In Jeavons syndrome (eyelid myoclonia with absences), as opposed to other photosensitive epilepsies, eye closure is more potent than photic stimulation as a triggering factor (62; Panayiotopoulos et al 1996.)

However, eye closure requires the presence of light, and it is entirely ineffective in darkness. Another intriguing feature is that some patients may manifest with both features of photosensitivity and fixation-off sensitivity, which have opposing characteristics. It is possible that in patients with Jeavons syndrome, the alpha-rhythm generators malfunction and both the magnocellular and parvocellular systems are functionally disturbed (125). See the article for eyelid myoclonia for further discussion.

According to 1 study, the coexistence of paroxysmal eyelid movements, photoparoxysmal EEG responses, increased blinking, and epileptic eyelid myoclonia in patients with visual-sensitive reflex seizures suggests an underlying dysfunction involving cortical-subcortical neural networks of a system epilepsy (21).

Failure of inhibitory or excitatory mechanisms or both? The intermittent photic stimulation flash time-locked occipital spikes appear on the descending arm of the P100 component of the visual-evoked response (VER) (128; 110; 110).

This, we postulated, was a failure of postsynaptic inhibition. However, the findings that “the P100 VER component is enhanced in photosensitive patients and that valproate slightly reduces its amplitude while occipital spikes are unaffected” were interpreted as evidence of “at least normal, if not supranormal, post-inhibitory potentials,” which suggests that the occipital spikes represent an excitatory phenomenon rather than a failure of inhibition (73), a view that was later challenged (06).

In 1 report, the effect of IPS at a common activating frequency (ie, 20 Hz) on motor cortex excitability was assessed by means of transcranial magnetic stimulation in 15 photosensitive patients with idiopathic generalized epilepsy (157). Nineteen normal subjects of similar age and sex acted as controls. After the resting motor threshold was measured, the corticomotor excitability was studied in 2 conditions randomly delivered, during IPS (5 s) at 20 Hz and without IPS. Motor evoked potentials were recorded from the right first dorsal interosseous muscle. The following parameters were determined: the cortical silent period, the short-latency intracortical inhibition at the interstimulus interval of 3 and 4 ms, and the intracortical facilitation at interstimulus interval of 12 and 14 ms. It was found that IPS at 20 Hz is significantly shortening the cortical silent period in normal subjects, whereas no significant changes were detected in patients. The resting motor threshold was significantly higher in patients than controls, as expected by the concurrent antiepileptic treatment. Other corticomotor excitability measures were unaffected. Thus, it was confirmed that IPS has a weak influence on the motor cortical output in patients. The authors concluded that the loss of the normal shortening of the cortical silent period, otherwise present in healthy subjects in response to IPS, may have a possible protective nature (157).

Another report found altered recovery from inhibitory repetitive transcranial magnetic stimulation in subjects with photosensitive epilepsy. Visual responses recovered more quickly in the stimulated hemisphere, and disinhibition persisted in the contralateral side of photosensitive subjects (19).

Perry and colleagues have found evidence for increased visual gamma responses in photosensitive epilepsy (135). A sustained gamma (30 to 70 Hz) oscillation induced in the occipital cortex by high-contrast visual stimulation has been well characterized in animal local field potential recordings and in healthy human participants using magnetoencephalography. The spatial frequency of a static grating stimulus that gives maximal gamma is also that most likely to provoke seizures in photosensitive epilepsy. The authors used magnetoencephalography to study visual responses induced by grating stimuli of varying contrast and size in 12 patients with photosensitive epilepsy and 2 matched control groups--one with epilepsy but no photosensitivity, the other healthy controls. They used a beamformer approach to localize cortical responses and to characterize the time-frequency dynamics of evoked and induced oscillatory responses. A greater number of patients with photosensitivity had particularly amplitude gamma responses compared to controls. Formal statistical testing failed to find a group difference. One photosensitive patient, tested before and after sodium valproate, had a peak gamma amplitude when drug naive over 4 times larger than the group mean for controls; this high amplitude was substantially decreased after treatment with sodium valproate. There was no difference in the frequency of the sustained gamma response between the 3 groups. It was concluded that altered power, but not frequency, in induced cortical responses to a static grating stimulus may be a characteristic of photosensitive epilepsy. The failure to find a group difference on statistical testing may have been due to a wide intersubject variability and heterogeneity of the photosensitive group. A high amplitude response would be in keeping with previous evidence of altered contrast gain and increased spatial recruitment in photosensitive epilepsy (135).

Multi-modal imaging documented a functional link between the circuits that trigger the posterior alpha rhythm and visual sensitivity (70; 170), a link that was also suggested for fixation-off sensitivity (125). Vaudano and colleagues investigated (A) the hemodynamic correlates of the spontaneous alpha rhythm, which is considered the hallmark of the brain resting state, in photosensitive patients and in people without photosensitivity and (B) the whole-brain functional connectivity of the visual thalamic nuclei in the various populations of subjects under investigation (170). Forty-four patients with epilepsy and 16 healthy control subjects underwent an electroencephalography-correlated functional magnetic resonance imaging study, during an eyes-closed condition. The following patient groups were included: (1) genetic generalized epilepsy with photosensitivity, 16 subjects; (2) genetic generalized epilepsy without photosensitivity, 13 patients; and (3) focal epilepsy, 15 patients. For each subject, the posterior alpha power variations were convolved with the standard hemodynamic response function and used as a regressor. Within- and between-groups second level analyses were performed. Whole brain functional connectivity was evaluated for 2 thalamic regions of interest, based on the hemodynamic findings, which included the posterior thalamus (pulvinar) and the medio-dorsal thalamic nuclei. Genetic generalized epilepsy with photosensitivity demonstrated significantly greater mean alpha-power with respect to controls and other epilepsy groups. In photosensitive epilepsy, alpha-related blood oxygen level-dependent signal changes demonstrated lower decreases relative to all other groups in the occipital, sensory-motor, anterior cingulate, and supplementary motor cortices. Coherently, the same brain regions demonstrated abnormal connectivity with the visual thalamus only in epilepsy patients with photosensitivity. These findings indicate that the cortical-subcortical network generating the alpha oscillation at rest is different in people with epilepsy and visual sensitivity. This difference consists of a decreased alpha-related inhibition of the visual cortex and sensory-motor networks at rest. These findings represent the substrate of the clinical manifestations (ie, myoclonus) of the photoparoxysmal response (170).

Neurotransmitters. A selective dopaminergic mechanism in human epileptic photosensitivity has been postulated (139; 138). Apomorphine, a dopamine receptor agonist, blocked photoparoxysmal responses in patients with idiopathic generalized epilepsy, and this effect was not modified by naloxone, a specific opiate antagonist, thus, suggesting that apomorphine acts on cerebral dopaminergic receptors. Conversely, apomorphine did not block spontaneous generalized spike-wave discharges in patients with non-photosensitive idiopathic generalized epilepsy.

Impaired white-matter integrity. In 1 study, diffusion tensor-imaging data from MRI brain scans were collected from 8 photosensitive patients and 16 gender- and age-matched non-epileptic controls using a SIEMENS Trio 3.0-Tesla scanner (43). Compared with the control subjects, the corpus callosum of patients had significantly lower fractional anisotropy values, indicating abnormal white-matter in the corpus callosum of patients.

Pathophysiology of photosensitive epilepsy in animals. Photosensitive papio-papio baboons have long been used as an animal model of photosensitive epilepsy (103; 107). They suffer from intermittent photic stimulation-induced myoclonus and have a natural predisposition to “nonepileptic myoclonus,” which is not accompanied by electrical discharges or seizures. Clinical manifestations do not show any signs of localized origin of the epileptogenic processes, and the photoparoxysmal responses are also bilateral and synchronous. However, experimental data document that the origin of photoparoxysmal responses and seizures is in the motor cortex. Photosensitive baboons also show a different pattern of activation and inhibition in blood-flow PET studies compared to normal controls, suggesting involvement of specific cortical-subcortical networks in photosensitivity (160; 159). However, studies in ferret visual cortex support a cortical etiology of pattern sensitivity (152).

In their study, Koepp and colleagues provide more details regarding ictogenic mechanisms and networks involved in photic-induced seizures (92).

Epidemiology

The prevalence and incidence of photosensitivity have been roughly estimated for those with (1) photoparoxysmal responses and clinical epileptic seizures (clinical photosensitivity) and (2) photoparoxysmal responses without firm evidence of clinical epileptic seizures (EEG photosensitivity).

The reported figures vary significantly according to reviews of existing evidence (73; 36; 124; 171), and they are highly dependent on the age and sex of the population studied, criteria for normality, and definition of EEG abnormal responses (191). The numbers may be underestimated for those with minor and infrequent seizures who do not have GTCS.

Clinical photosensitivity. Clinical photosensitivity probably affects 1 in 4000 of the population (5% of patients with epileptic seizures); two thirds are women (video game induced seizures occur more often in men); and the onset has a peak age at 12 to 13 years (73).

In a well performed demographic study in Great Britain, the annual incidence of cases of epilepsy with generalized photoparoxysmal responses on their first EEG was conservatively estimated to be 1.1 per 100,000 (representing approximately 2% of all new cases of epilepsy). When restricted to the age range 7 to 19 years, the annual incidence rose to 5.7 per 100,000 (approximately 10% of all new cases of epilepsy presenting in this age range) (141). This means that photosensitivity is found in 2% of patients of all ages presenting with seizures and 10% of patients presenting with seizures in the age range 7 to 19 years (141). Another demographic study of the same group of authors was on seizures induced by electronic screen games (140). Of 118 patients who had a first seizure while playing an electronic screen game during two 3-month periods, 3 groups were identified: (Group A) 46 patients for whom there was thought to be a definite causal relationship (type 4 photoparoxysmal response); (Group B) 25 patients for whom there was a probable causal relationship (types 1 to 3 photoparoxysmal response, clinical evidence of photosensitivity, subsequent recurrent seizures on repeat exposure to electronic screen games, or occipital spikes in the resting electroencephalogram; and (Group C) 47 patients for whom there was no apparent causal relationship. The number of patients in Group C did not exceed that expected by the chance occurrence of 2 common events (playing electronic screen games and incidence of epilepsy). Most (103/118) of the patients were in the age range of 7 to 19 years. Within this age group, the annual incidence of first seizures triggered by playing electronic screen games (Groups A and B combined) was estimated to be 1.5/100,000.

In Japan, approximately 1% of people under the age of 18 years had seizures after watching a Pocket Monster cartoon in 1997 (78; 52). However, some had migraines, visual distortions, nausea and motion sickness, or other nonseizure symptoms, and more than half the children who experienced a previous convulsion had a history of a seizure induced by television (78).

In our studies, (1) clinical photosensitive epilepsy was found in 15 (3.4%) of 442 patients with onset of 1 or more afebrile seizures between birth and 15 years of age (119); (2) idiopathic photosensitive occipital lobe epilepsy was found in 11 (0.7%) of 1,550 adult and child patients with epilepsy (118); and (3) eyelid myoclonia with absences happened in about 3% of adults with epilepsy and 13% of those with idiopathic generalized epilepsy with absences (62).

EEG photosensitivity in patients with epileptic seizures. The prevalence of EEG photosensitivity is probably 5% amongst patients with clinically evident epileptic seizures (73).

EEG photosensitivity is significantly higher in (a) pediatric than adult patients with epilepsy, (b) females than males, and (c) those with generalized rather than focal epilepsies. It is well documented that photoparoxysmal responses with or without clinical photosensitivity are associated with greatly variable types of epilepsy and mainly juvenile myoclonic epilepsy (30%), Dravet syndrome (70%), Unverricht-Lundberg disease (90%), and a number of autosomal recessive progressive myoclonic epilepsies (188; 147; 07; 73; 179; 67; 155; 101; 171).

EEG photosensitivity in normal subjects. Photoparoxysmal responses were found in 1.3% to 7.6% of normal school-aged children and 0.35% to 2.4% of normal Air Force candidates.

For such normal people with EEG photoparoxysmal responses, the likelihood to develop epilepsy is very low. In 1 report, the average prevalence of EEG photoparoxysmal responses (of any type, posterior or generalized) was 7.6% in healthy children aged 1 to 16 years, but only 3% of them develop seizures to age 20 years (42). In 2 other smaller scale studies, none of the normal subjects with photoparoxysmal responses developed epileptic seizures over a follow-up period of 6 to 12 years (154; 171). However, it may be possible that “asymptomatic photosensitive subjects” have unnoticed minor reflex seizures triggered by stimuli encountered in daily life.

Differential diagnosis

The differential diagnosis first involves the distinction of genuine epileptic seizures from nonepileptic paroxysmal attacks (124) and then the documentation of precipitating factors.

Eyelid myoclonia induced by eye closure is often misdiagnosed as tics or attempts for self-induction of seizures. (See clinical vignette 1.) Nonepileptic paroxysmal eyelid movements may occur in children and adults with generalized photosensitive epilepsy and may be mistaken for absence seizures. There is often a family history of eyelid movements, and EEG monitoring readily distinguishes the 2 (24).

Reflex and spontaneous visual seizures are commonly misdiagnosed as migraine with visual aura.

Pure photosensitive epilepsies cannot be diagnosed on the basis of an epileptiform response to flicker alone. This EEG finding occurs in asymptomatic subjects (especially children) in several forms of epilepsy and with different seizure types, which are usually easily distinguished from pure photosensitive epilepsies on clinical and EEG grounds. Moreover, some patients with pattern sensitive epilepsy may not be sensitive to flash intermittent photic stimulation.

Photosensitivity with generalized seizures may accompany idiopathic generalized epilepsies with spontaneous seizures, especially juvenile myoclonic epilepsy, and is typical in eyelid myoclonia with absences. It also may occur with symptomatic generalized epilepsies, such as severe myoclonic epilepsy of infancy (Dravet syndrome), or with degenerative gray matter encephalopathies, such as Lafora disease, Unverricht-Lundborg disease, Kufs disease, the neuronal ceroid lipofuscinoses, and others collectively known as the progressive myoclonus epilepsies in which photosensitivity at low flash frequencies is typical. These syndromes are associated with photic cortical reflex myoclonus, and the patients also have clear-cut action myoclonus.

Generalized seizures and EEG abnormalities induced by visual stimulation are conventionally differentiated from idiopathic photosensitive focal seizures with typical secondary generalization, but detailed clinical and EEG studies may be needed to make this distinction, and it should be recalled that the initial stimulus activates occipital lobe structures in both types.

Diagnostic workup

The majority of patients with visual-induced seizures have no evident lesions with brain imaging, which is rarely needed. Properly applied intermittent photic stimulation or other stimuli during EEG is the most important test.

The practical objective of this is to determine the following:

EEG intermittent photic stimulation procedures. Intermittent photic stimulation (IPS) techniques and stroboscopes has differed significantly in various publications and departments. An attempt to standardize the procedure has been published (85).

Intensity, frequency, and duration of the IPS are the most significant determinants of the response. An abnormal photoparoxysmal response is more likely to occur with light of high intensity, frequency of mainly 12 to 20 Hz, and longer trains of IPS (with reasonable limitations of no more than 5 seconds). Combining light IPS with appropriate geometric patterns makes it much more potent than diffused or white light. Distance and ambient light mainly influence the intensity of the photic stimuli.

Binocular is far more effective than monocular stimulation. This is why it is recommended that photosensitive patients cover 1 eye when in epileptogenic environmental photic conditions such as discotheques.

Stimulating central vision (fixating on the light source) is much more potent than stimulating peripheral vision. This is because photosensitivity is mainly mediated through central vision and fixation. This is also the reason for the recommendation that, in IPS testing, “a central marker on the diffuser to aid fixation is specified, as photosensitivity is unlikely to be demonstrable in the eyes-open state unless the stimulator is in central vision.”

The state of the eyes during IPS is probably the most significant internal factor that modifies the response to IPS (111; 120); eye closure (closing of the eyes while IPS continues) is by far the most potent. Of the other 2 states, eyes open is more susceptible than eyes closed to patterned flickering lights. Conversely, eyes closed appeared to be more susceptible than eyes open to direct unpatterned light, probably because of a diffusion effect of light by the eyelids. When light with a diffuser is applied, eyes open is again more effective than eyes closed because of an intensity loss by the closed eyelids.

Failure to achieve maximal provocative IPS may produce false negative results in the testing of patients for photosensitivity. However, prolonged photic stimulation that may expose the patient to a major seizure should be totally discouraged. To continue a train of photic stimulation after the appearance of EEG ictal discharges or ictal clinical manifestations is unacceptable. There is nothing to learn or benefit from this practice. There are plenty of examples of individuals having an IPS-induced seizure during an EEG that was performed for reasons other than epilepsy.

The most comprehensible procedure for photic stimulation is that of Jeavons (81):

EEG abnormalities in photosensitive patients. The resting EEG of patients with idiopathic photic reflex seizures is usually normal or frequently (20% to 30%) shows eye closure-related paroxysms occurring within 1 to 3 seconds after closing the eyes. These are usually brief EEG paroxysms lasting for 1 to 4 seconds and having similar features to those elicited by intermittent photic stimulation for each individual patient. They disappear if eye closure occurs in total darkness.

Photoparoxysmal responses are broadly categorized as:

Generalized spike or polyspike waves. They are of higher amplitude in the anterior regions, but onset, particularly if patterned intermittent photic stimulation is employed, is often with occipital spikes. They are highly associated (90%) with clinical photosensitivity, particularly if they outlast the stimulus train.

Generalized photoparoxysmal responses often (60%) associate with clinical events such as jerks, impairment of cognition, or subjective sensations, but their detection may require video-EEG.

Posterior (temporoparieto-occipital with occipital emphasis) spike or polyspike-waves. This is the mildest form of photoparoxysmal response and does not spread to the anterior regions. It consists of occipital spikes, polyspikes, or slow waves mixed with small, larval spikes.

Occipital spikes are often time-locked to the flash with a latency of approximately 100 ms, coinciding with the positive P100 of the visual-evoked response (110).

Half the patients with posterior photoparoxysmal responses have epileptic seizures (spontaneous or photically elicited or both).

Some authors follow the subclassification of photoparoxysmal responses into 4 types as proposed by Doose (41; 178):

Photomyoclonic responses are not cerebral responses (60). They are spike or polyspike-like muscle activity that appears in the frontal-central EEG electrodes. They occur only when the eyes are closed and are inhibited by opening of the eyes. Provocation of photomyoclonic responses requires very high intensity intermittent photic stimulation with the stroboscope positioned very close to the eyes. They are unlikely to occur when intermittent photic stimulation is applied at recommended levels. Clinically, they manifest with predominant jerking of the facial muscles, especially around the eyes (eyelid fluttering), a phenomenon called photomyoclonus. They may end in generalized convulsions if the stimulation is continued and the eyes of the subject are kept closed. Photomyoclonic responses are a nonspecific finding reported in normal people (0.3%), psychiatric patients (17%), epileptic patients (3%), and patients with brainstem lesions (60).

Risk of seizures associated with photoparoxysmal responses. Ictal clinical manifestations during photoparoxysmal responses (PPR) may be 1 of the most important factors with regard to risk of seizures, but this has not been studied and emphasized in expert consensus.

Occipital spikes and other posterior abnormalities induced by intermittent photic stimulation are considered of much lower epileptogenic capacity than generalized PPR (see above). The wider the PPR range, the more the patient is at risk of seizures in daily life (83). Emphasis is often given to whether PPR outlast the stimulus train or whether they are self-limited, ie, they stop before or with the end of the intermittent photic stimulation. The rationale is that PPR that outlast the stimulus train may strengthen their association with epilepsy. This may be artificial because the duration of the discharge as a rule depends on the duration and strength of the intermittent photic stimulation and the time that this is stopped after the onset of PPR.

Warning. Responses can be greatly attenuated or abolished by some antiepileptic drugs, especially valproate, and this must be considered when interpreting the EEG for follow-up reasons and for decisions on antiepileptic drug withdrawal.

Testing of pattern-sensitive patients. A significant number of patients with pattern sensitivity also show PPR in routine EEG with intermittent photic stimulation.

However, the documentation of pattern sensitivity requires specific testing with appropriate pattern presentations. Pattern sensitivity depends on the spatial frequency, orientation, brightness, contrast, and size of the pattern. An optimally epileptogenic pattern consists of black-and-white stripes of equal width and spacing.

The most epileptogenic patterns and their spatial/directional relations to the eyes have been defined by Wilkins and Binnie as follows (181; 18):

Syndromes and diseases in which the seizure type occurs

Visual-induced seizures occur in patients of all ages of normal or abnormal neurocognitive state and/or with normal brain imaging. They may be a one-off event or life-long, focal or generalized, responsible or resistant to antiepileptic drug treatment. They frequently occur together with spontaneous seizures.

Photoparoxysmal responses with or without clinical photosensitivity are associated with greatly variable types of epilepsy, mainly juvenile myoclonic epilepsy (30%), Dravet syndrome (70%), Unverricht-Lundberg disease (90%), and a number of autosomal recessive progressive myoclonic epilepsies (188; 147; 07; 73; 179; 67; 155; 101; 124; 93).

Pure reflex syndromes with visual-induced seizures are:

Management

Avoidance, prevention, or modification of the provocative stimulus. Avoidance, prevention, or modification of the provocative stimulus may be sufficient for many patients with visual-induced seizures (73; 33; 124; 171). Usual advice about avoiding sleep deprivation, excessive alcohol use, and recreational drugs should be given.

Patients with television-induced seizures should be advised to watch television in a well-lit room; maintain a maximum comfortable viewing distance (typically, more than 2.5 m from a 19-inch screen); use the remote control and, if necessary, approach the screen by covering 1 eye with their palm; and avoid prolonged watching, particularly if sleep-deprived and tired. Occlusion of 1 eye is also advised when photosensitive subjects are suddenly exposed to flickering lights, as in discotheques, for example.

Patients with video game-induced seizures can often do without video games or significantly restrict the time spent playing. They should not play when sleep-deprived or tired.

Conditioning treatment or wearing appropriate tinted glasses has been recommended (183; 91; 25; 171). A commercially available blue lens, named Z1, was found to be highly effective in controlling photoparoxysmal discharges (25; 171). An optometric technique, colorimetry, enables the perceptual effects of ophthalmic tints to be evaluated subjectively, optimized, and then prescribed in tinted spectacles (181; Wilkins et al. 1999).

Management of visual-sensitive epilepsy also necessitates the development and implementation of guidelines to minimize exposure of susceptible populations to provocative stimuli (85). In fact, the Pokémon incident in Japan stimulated a debate on the need for regulations and protective measures for video material, particularly for television programs, to prevent seizure precipitation. Since then, official guidelines exist in some countries, notably in Japan and Europe, and have been shown to protect photosensitive subjects from the broadcasting of risky program content (163). Updated guidelines and recommendations should consider the role of parameters such as modulation depth and stimulus wavelength at provocative frequencies and the increasing availability of modern audiovisual technology that employs large screen without flicker effects but with significant changes of other variables (for instance, luminance of the screen and the separate stimulation of the 2 eyes). To pursue this goal, sensitization and cooperation from the industry are necessary as is the involvement of broadcasters and producers (85).

One publication presents a spatiotemporal pattern detection algorithm that can detect hazardous content in streaming video in real time (05). A tool is developed for producing test videos with hazardous content, and then those test videos are used to evaluate the proposed algorithm, as well as an existing post-processing tool that is currently being used for detecting such patterns. To perform the detection in real time, the proposed algorithm was implemented on a dual core processor, using a pipeline/parallel software architecture. Results indicate that the proposed method provides better detection performance, allowing for the masking of seizure inducing patterns in real time (05).

The risk of seizures evoked by 3D video displays or virtual reality headsets in children appears to be very low, even for those with EEG photosensitivity (169).

Prophylactic antiepileptic medication. Prophylactic antiepileptic drugs are needed when triggering stimuli are not preventable and when seizures are frequent or also occur spontaneously. The choice of an antiepileptic drugs depends on their specific efficacy on the type or types of reflex and spontaneous seizures and the particular epileptic syndrome (124). This is no different from treatment of spontaneous seizures of the same type.

In photosensitive and pattern-sensitive patients and in those with syndromes of generalized epilepsies, valproate, levetiracetam, and lamotrigine are the main antiepileptic drugs to recommend in that order of efficacy. Valproate controls all types of seizure in more than 80% of patients (124; 172). Levetiracetam controls all types of seizure, with well-established efficacy in EEG and clinical photosensitivity (89; 33); it is more efficacious in myoclonic and GTCS than absence seizures. Lamotrigine also appears to be effective (61), but it may exaggerate jerks. In an old report, suppression of photoparoxysmal responses with lamotrigine has been seen in 5 patients, of whom 4 were also taking valproate (17) with which there is a beneficial pharmacodynamic interaction (126). Ethosuximide and lamotrigine may be particularly useful in absence seizures when valproate is ineffective or undesirable. Clonazepam is effective in photically or spontaneously induced myoclonic seizures (but is ineffective against and may exaggerate GTCS). Most of the other antiepileptic drugs (ie, carbamazepine, gabapentin, oxcarbazepine, phenytoin, pregabalin, tiagabine, and vigabatrin) are contraindicated because they are ineffective (ie, they induce side effects without providing any therapeutic benefit in addition to depriving patients of appropriate treatment).

One study examined the effects of valproate monotherapy on seizure response/control and photosensitivity in 55 adolescents with newly diagnosed epilepsy with generalized tonic clonic seizures only and photosensitivity (172). Two phases of the study were defined and analyzed separately. In the phase I, the electroclinical data of patients were compared over 3 time points: T1 (at 6 months of treatment); T2 (at 12 months of treatment); and T3 (at 36 months of treatment). In the phase II, only patients who stopped valproate were evaluated over a period of 12 months. At both T2 and T3 there was a significantly great percentage of seizure-free patients compared with that at T1 (78.2% vs. 69.1%, p < 0.01; and 85.5% vs. 69.1%, p < 0.001), and a similar trend was also noted according to photosensitive-free patients (70.9% vs. 52.7%, p < 0.01; 80.0% vs. 52.7% p < 0.001). At the end of the phase II, 46.5% and 32.6% out of 43 patients who stopped valproate had seizure relapses and reappearance of photosensitivity, respectively. In particular, 78.6% of the 14 patients with photosensitivity reappearance presented the same type of EEG response showed at study entry. The authors concluded that valproate monotherapy is effective for both seizure outcome control and photosensitivity reduction in adolescents with epilepsy with generalized tonic clonic seizures only. Treatment discontinuation induces relapse of seizures and photosensitivity in a certain number of patients (172). However, valproate is unwanted in women of childbearing age.

Human photosensitive epilepsy models have been used as proof of principle drug trials for epilepsy and can be useful as early and informative indicators in anti-epileptic drug discovery and development (89; 88; 189; 69). Photosensitive patients are exposed to intermittent photic stimulation, and the reduction in sensitivity to the number of standard visual stimulation frequencies is used as an endpoint.

Patients with self-induced visual-induced seizures are difficult to treat and may need psychiatric or behavioral intervention (108). The attacks may be pleasurable, and there is clear secondary gain for some patients (Binnie 1988; 164). On anecdotal evidence, fenfluramine (a serotonin-releasing drug) has been recommended for the treatment of self-induced epilepsy in combination with valproate (27).