Search

Photosensitivity is usually tested by intermittent photic stimulation. Photoparoxysmal responses in the EEG are enhanced when the eyes remain closed during stimulation. We tested the hypothesis that this is due to diffusion of light by the eyelids. In 25 photosensitive patients, conditions 'eye closure', 'eyes closed', 'eyes open' and 'eyes open with diffuser' were tested for frequencies of 2-60 Hz. Additional influences of a red filter and fixation were also examined. The photosensitivity range was maximal in the condition 'eyes open with diffuser', due to an increase of the upper limit to a median 60 Hz (range 25-60), from 35 (15-50) Hz with eyes open, 37.5 (25-60) Hz with eyes closed and 40 (23-60) Hz with eye closure (P = 0.0002). This effect was attenuated in patients on valproic acid and with the use of a red or white filter. Testing with a diffuser was more sensitive than without, except for one patient who was photosensitive only in the eye closure condition. The influence of the eyelids on photosensitivity can be explained by diffusion of light, attentuated by an intensity loss. Use of a diffuser may simplify testing for photosensitivity in the EEG laboratory. The diffusion effect may explain seizure susceptibility in front of 50 and 60 Hz television screens.

We undertook a prospective nationwide study to determine the incidence of photosensitive epilepsy (PE). Virtually all EEG departments in Great Britain (providing approximately 90% coverage of all EEGs performed on people with newly diagnosed seizures) screened cases referred to them over a 3 month period and identified all new cases of epilepsy (defined as one or more recognised seizures) whose first EEG showed a photoparoxysmal response (PPR) on intermittent photic stimulation (IPS). 191 cases were identified, 143 of whom had type 4 PPRs (generalised spike and wave on IPS) on their first EEG. The annual incidence of cases of epilepsy with type 4 PPRs 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 7-19 years, the annual incidence rose to 5.7 per 100,000-approximately 10% of all new cases of epilepsy presenting in this age range. To ascertain if there was a significant seasonal variation in PE, 5 EEG departments (which together contributed 15% of cases in the first study period) were visited during a second 3 month study period to identify all new cases of epilepsy with type 4 PPRs on their first EEG. No significant seasonal variation in incidence between summer and winter was found.

In order to investigate the significance of different patterns of the photoparoxysmal response (PPR), an electroencephalographic family study on 135 probands and 371 relatives was carried out. The PPR was subclassified in 4 phenotypically different types. Considering all types of PPR, the incidence was equal in siblings of probands with epilepsy and in siblings of probands without epilepsy. Type 4 (generalized spikes and waves) occurred more often both in probands with epilepsy and their siblings than the respective controls. Thus, the coincidence of photosensitivity will appear as higher if only a PPR with generalized spikes and waves is considered to be indicative of photosensitivity. A striking age dependency of the phenotypical expression of the PPR was found. These findings suggest that the phenotypical expression of the PPR is age-related and modified by other factors predisposing to generalized epilepsy, the varied patterns of the PPR only representing different levels of expression of the same genetically determined trait. The importance of an adequate technique of intermittent light stimulation is discussed.

A group of 15 volunteers with both photosensitive epilepsy and pattern sensitivity were shown patterns of stripes in the left and right visual hemifields. In the majority of patients the resulting paroxysmal epileptiform EEG activity was maximal in the posterior temporal and occipital regions contralateral to the stimulus. Patterns in the upper and lower hemifields were also presented and in two patients there was a difference in the vertical distribution of the response such as to suggest that in these patients the discharges were largely confined to the striate and prestriate cortex. The probability with which paroxysmal activity was evoked was usually different for the two lateral hemifields. This difference between the hemifields (which was dramatic in some patients) was mirrored in a lateralisation of the responses to diffuse intermittent photic stimulation. The association between the asymmetries in pattern sensitivity and in the responses to photic stimulation suggests that the two cerebral hemispheres often differed in hyperexcitability. It is therefore interesting that the putative differences in hyperexcitability were observed in patients with primary generalised epilepsy as well as those with other types.

The triggering of epileptiform EEG discharges by pattern is thought to depend on the intensity of excitation within the visual cortex. The present study investigates the role of the synchronisation of neuronal activity by the stimulus. In 10 pattern-sensitive subjects the effects of the following patterns have been compared: (1) static gratings, (2) gratings oscillating in a direction orthogonal to the lines (which should synchronise activity in direction-sensitive cortical units), and (3) gratings drifting at the same angular velocity (which should produce little or no synchronisation, because the contours enter and leave the overlapping receptive fields of individual neurones asynchronously). The oscillating gratings were most, and the drifting least, epileptogenic. In 2 further subjects oscillating and phase-reversing patterns were more epileptogenic than drifting gratings. Although open to alternative explanations, the findings conform with predictions from the hypothesis that synchronisation of individual cortical neurones by the stimulus contributes to epileptogenesis in photosensitive subjects.

Of a consecutive series of 138 photosensitive epileptic patients 57 showed a unilateral preponderance of asymmetrical discharges on IPS, significant in each individual at the 5% level. There was no association between asymmetry of IPS-induced discharges and clinical evidence of cerebral pathology (partial or secondary generalized epilepsy, neurological deficit, left handedness, etc.). It is concluded that these results, together with our previous findings of asymmetrical pattern sensitivity suggest that the cortical hyperexcitability postulated in primary cortico-reticular epilepsy is not uniformly distributed.

Seven patients with pattern-sensitive epilepsy were examined repeatedly over a period of 4–12 months during which the dose of sodium valproate was adjusted. Patterns were presented in series in which the size of successive patterns was progressively increased until paroxysmal activity appeared, or until the largest pattern (radius 24° visual angle) had been presented. As valproate dose increased, paroxysmal activity was less likely to occur. When it did occur it was less likely to include a spike; it had a lower voltage, involved fewer electrodes, and lasted a shorter time. The size of the pattern just sufficient to induce paroxysmal activity showed relatively little change with valproate dose. The degree to which the various electrodes were involved in the discharges showed considerable stability. The paroxysmal response to patterns presented in one visual half-field was almost invariably maximal over the contralateral posterior quadrant, usually at the posterior temporal electrode.

The acute pharmacological effects of melatonin (5-methoxy-N-acetyl-tryptamine) were evaluated on behaviour and electrographic signs in the photosensitive baboon (Papio papio), including visually evoked potentials (VEPs). Doses of 5 and 10 mg/kg i.v. had little effect, contrasting with the enhacement of spontenous paroxysmal activity and latency of clinical signs elicited by intermittent ligh stimulation (ILS) produced by the 20 and 50 mg/kg doses. Moreover, a modification of the electrographic pattern of discharge during ILS-induced myoclonic responses was observed. These observations indicate a slight decrease of light sensitivity after melatonin. Only frontal VEPs were modified, with increase in amplitude of their late components for al doses tested. These effects lasted for 2–4 h. The participation of serotoninergic mechanisms involved in melatonin effect and in this model of experimental epilepsy are discussed.

During routine EEG investigation of patients with photosensitive epilepsy and of patients with EEG abnormalities induced by intermittent photic stimulation (IPS) it was noted that two different photostimulators had different efficiencies in provoking occipital spikes and photoconvulsive responses. The difference appeared to be related to the presence of a metal grid. Comparison was made of the intensity of the photic stimulus of each lamp and the method is described. Ten patients were studied in detail in order to establish which patterns, when combined with IPS, were the most effective in inducing a photoconvulsive response. It was found that most patterns enhanced the effect of IPS, but the grid and “small squares” pattern were the most potent. It is therefore suggested that a square pattern in which the squares subtend approximately 20′ should be fitted to photostimulators for routine use.

In fifteen patients with photosensitive epilepsy the characteristics of the EEG epileptiform discharges induced by intermittent photic stimulation (IPS) were investigated. In all the patients generalized EEG epileptiform discharges were induced by IPS. In eight of them (1st group) the discharges appeared first in the occipital area, whereas in the remaining seven patients (2nd group) they occured simultaneously over all head regions or appeared earlier over the anterior areas of both sides. Averaged visual evoked responses (VERs) to single flashes of light in the patients were similar morphologically to those of normal subjects. In these, the VER recorded from the mid-occipital area consisted of 4–6 rhythmic wavelets occurring within 100 msec from the flash and of larger deflexions following such wavelets. The initial component of the response is positive and has a peak latency of 20–27 msec. In six patients of the 1st group certain components of the VER were extraordinarily augmented. The earliest spikes of the EEG epileptiform discharges appearing first over the occipital regions were interpreted as unusually augmented VERs. The photosensitivity of the patients decreased remarkably during drowsiness. No epileptiform discharges were induced during slow wave sleep. In the REM stage of sleep the discharges were on the contrary as promptly induced as in the waking state. This phenomenon did not seem to be due to pupillary miosis. Monocular stimulation markedly reduced the epileptogenic effect of IPS as compared to binocular stimulation. Background illumination appeared to modify the epileptogenic effects of IPS. Ophthalmological examination revealed incomplete red-green color blindness in one patient and quadrantic hemianopsia in another. In the remaining subjects no pathological conditions of the retina were found. Some considerations on the neural mechanisms involved in epileptic photosensitivity are presented.

Red-coloured flicker is claimed to be more epileptogenic than white or that of other colours matched for subjective intensity. A feature of the colour opponent system is that the response of luminosity-sensitive cortical units to stimulation of ganglion cells of a particular spectral sensitivity is reduced when cells of other sensitivities are simultaneously stimulated. We hypothesized that the apparent effect of colour on photosensitivity was not a property of red light per se but arose simply from the fact that, with commercially available filters a light can be provided to stimulate only red sensitive cones, but owing to the overlap of the absorption spectra of the visual pigments it is difficult to stimulate only green or blue sensitive cones. Such stimulation of a single cone population can be achieved by the 'silent substitution method' which has been used for evoked response studies. In 12 photosensitive epileptic patients we find that, using stimulus intensities (less than 20 nits) at which white flicker is without effect, stimulation of either red or green cones by the silent substitution method may produce epileptiform discharges, there being a slight (and not significant) excess of patients showing a greater sensitivity for green than for red cone stimulation. The findings are considered to support the hypothesis set out above.

The effects of the intravenous or intracerebroventricular injection of the stereoisomers, and the racemic mixture, of allyglycine (2-amino-pent-4-enoic acid) have been studied in baboons, Papio papio , with photosensitive epilepsy. Enhancement of the natural syndrome of photosensitive epilepsy is seen 1–12 h (maximally at 3–8 h) after l -allylglycine, 100 mg/kg, intravenously, or d,l -allyglycine, 200 mg/kg intravenously. Such enhancement is seen with a slower onset, and to a lesser, and more variable, extent after d -allylglycine, 500–750 mg/kg, intravenously. Brief focal or generalised seizures occured (in the absence of intermittent photic stimulation) after l -allylglycine, 150–200 mg/kg, intravenously. This effect is similar to that previously observed after d,l -allylglycine, 300–400 mg/kg. d -Allylglycine, 780 mg/kg, intravenously produced episodes of vertical nystagmus with increased extensor motor tone, but no ‘spontaneous’ seizures. Intracerebroventricuar injection of l -allylglycine, d -allylglycine or d,l -allylglycine, 100 mg in 1 ml saline, did not modify the natural syndrome of photosensitive epilepsy. d -allylglycine, or d,l -allylglycine, 100 mg intracerebroventricularly, after 1–2 h gave rise to a syndrome with vomiting, sustained vertical nystagmus, and intermittent extensor spasms. The results are interpreted in terms of regional differences in the metabolism of the two isomers to active compounds that can inhibit glutamic acid decarboxylase. d -Allylglycine is active only at the brain stem and cerebellum because d -amino acid oxidase is largely confined to these brain areas.

Patients with photosensitive epilepsy were asked to view normally functioning 625-line televisions while the EEG was monitored. In the first of two studies paroxysmal EEG activity was reliably induced by television at a viewing distance related to a patient's sensitivity to intermittent photic stimulation (IPS); patients who were sensitive to diffuse IPS at 50 Hz were sensitive to the television at greater viewing distances than those who were not. No such relationship was obtained with patterned IPS. On the other hand, patterned IPS was generally more epileptogenic than diffuse IPS with the same luminance. In the second study, where the angular subtense of the television screen and the subtense of its lines were manipulated independently, the convulsive response was found to be a function of both factors, the relative contribution of each depending on the viewing distance at which the patient was sensitive. For patients sensitive at normal viewing distances, where 50 Hz diffuse flicker appeared to be responsible for the induction of paroxysmal activity, the probability with which paroxysmal activity was induced was closely related to the subtense of the screen. For patients sensitive only at closer viewing distances the probability was influenced not by the subtense of the screen but by the subtense of its lines, suggesting that the paroxysmal activity was induced by the 25 Hz pattern alternation produced by the scan. A television with a small screen was considerably less epileptogenic than one with a large screen for all patients, presumably due to the reduced contribution of both diffuse flicker and pattern alternation.

The responses to intermittent photic stimulation (IPS) at rates of 1–10 fl/sec were studied in seventy-two subjects, eighteen being normal controls, thirty-eight having photosensitive epilepsy and sixteen having epilepsy which was not clinically photosensitive. All patients showed abnormal responses to IPS. In most of the patients photic stimulation evoked negative occipital spikes with or without photoconvulsive responses. Comparison was made between the occipital spikes and the VERs in the same patient. This showed that there was no simple time relation between occipital spikes and components of the VERs. The VERs of the normal subjects showed no component with a latency corresponding to that of the negative occipital spikes seen in the patients. It is suggested that the non-specific thalamo-cortical system may be responsible for the genesis of the “epileptogenic” occipital spikes and that increased sensitivity of the occipital cortex is present in most photosensitive patients.

1.1. Drugs modifying γ-aminobutyric acid (GABA) metabolism have been administered intravenously to baboons (Papio papio) manifesting a syndrome of photosensitive epilepsy. Myoclonic and seizure responses to intermittent light stimulation (ILS) have been observed before and after drug administration and compared with epileptic phenomena directly induced by the drugs. 2.2. Isoniazid (INH), 65–80 mg/kg, increased the tendency to respond to ILS with generalized myoclonus or tonic-clonic seizures. ILS given 30–60 min after isoniazid 100 mg/kg induced seizures which commonly originated asymmetrically in the occipital cortex whereas in animals without drug treatment seizures usually originated in the fronto-rolandic cortex. Seizures occurred spontaneously (i.e., in the absence of ILS) 7–70 min after isoniazid 130–150 mg/kg. Apart from changes associated with the seizures, no acute effects on behaviour were observed and there were no modifications in the background EEG rhythms. 3.3. Enhanced responsiveness to ILS was observed 1–6 h after thiosemicarbazide (TSC), 4–5 mg/kg. Spontaneous seizures with very variable cortical points of origin were observed 1–4 h after thiosemicarbazide 7.5–10 mg/kg. After either TSC or INH it was possible to induce seizures with ILS repeatedly at intervals of a few minutes. 4.4. Pyridoxine administered acutely in doses of 50–250 mg/kg did not diminish the tendency of photosensitive baboons to respond to ILS with generalized myoclonus. Thus the natural syndrome is not due to a dietary deficiency of pyridoxine or to a metabolic disorder leading to an abnormal requirement for pyridoxine. Pyridoxine, 150–300 mg/kg, given before isoniazid, did not modify its seizure-promoting effect. Pyridoxine, 200–250 mg/kg, completely blocked both the spontaneous and the ILS-induced seizures normally produced by thiosemicarbazide, 10 mg/kg. 5.5. Amino-oxyacetic acid (AOAA) (5–15 mg/kg) had a marked protective effect against myoclonus induced by ILS. This effect began 30–40 min after AOAA administration and lasted up to 24 h. Spontaneous or ILS-induced cortical spikes or spikes and waves were not diminished by AOAA. A larger dose of AOAA (20 mg/kg) had a toxic effect, producing convulsions after 15–30 min. 6.6. Drugs lowering brain GABA turnover (INH and TSC) enhance myoclonic responses to ILS and one which raises the brain GABA content (AOAA) blocks myoclonic responses. However, the occurrence of multiple seizures and the shift to an occipital point of origin for the seizures after INH and TSC suggest that it is unlikely that the same biochemical derangement is responsible for the photically-induced and drug-induced seizures. The persistence of spontaneous spikes and waves after AOAA suggests that this drug does not correct the primary abnormality but modifies its clinical expression.

A comparison of the group mean visual evoked potential (VEP), somato-sensory evoked potential (SEP) and electroretinographic (ERG) response was made between photosensitive epileptic subjects and age and sex-matched normal controls. The VEP of epileptic subjects was similar in wave form but differed significantly in a number of other features from that of normals. In the occipital region, components II, III, VI and VII were of larger amplitude, component VI showed a longer latency; and there was a tendency for the response to spread anteriorly to the central electrode. Midline central (vertex) responses were generally dissimilar in wave form from occipital responses and showed greater differences than did occipital potentials between the two groups, all components being of larger amplitude in photosensitive epileptic subjects. There was evidence also for alteration of the ERG in the form of an earlier a wave. The changes indicate participation in photosensitivity of: the retina; visual pathways probably of primary, secondary and associative type; unspecific diffuse projection systems projecting maximally to the vertex; or perhaps local changes of the central cortex. The possibility of contamination of cerebral responses by ERG, eye and muscle potentials was assessed by various recording techniques and, in two cases, by comparing cortex and scalp responses during diagnostic cerebral biopsy. Differential stages of sleep and states of arousal produced quite stereotyped changes of the VEP. The SEP was altered even more than the VEP. Individual components showed marked changes in amplitude, latency and distribution, some having mean amplitudes of up to 8 times normal. It is suggested that “photosensitivity”, at least in epilepsy, represents a diffuse multi-modal alteration in cerebral excitability affecting various levels of different sensory systems.

In eleven patients with photosensitive epilepsy, photosensitivity during various phases of sleep and wakefulness was investigated. During the waking state, epileptiform discharges were promptly induced by intermittent photic stimulation (IPS) in all of the patients. During non-REM sleep, five patients showed no evidence of photosensitivity. The remaining six patients showed some EEG evidence of sensitivity to IPS, but this was reduced remarkably as compared with that in the waking state in five of them. The induced discharges during non-REM sleep were usually most prominent on the anterior regions and were found only in the first cycle of nocturnal sleep. In one patient, prominent epileptiform discharges were induced during non-REM sleep both in the first and last cycles, but IPS failed to induced paroxysmal discharges during non-REM sleep in the other cycles of nocturnal sleep. During REM sleep, all patients except one who was treated with anticonvulsants showed prominent photosensitivity, although this was in many instances relatively reduced as compared with that in the waking state. In some cases the induced discharges were of larger amplitude in the posterior than in the anterior region. Possible explanations for previous conflicting reports on this subject are discussed.