The Benign Occipital Epilepsies of Childhood: An Overview of the Idiopathic Syndromes and of the Relationship to Migraine

Transcription

1 Epilepsia, 39(SuppI. 4):S%S23, 1998 Lippincott-Raven Publishers, Philadelphia 8 International League Against Epilepsy The Benign Occipital Epilepsies of Childhood: An Overview of the Idiopathic Syndromes and of the Relationship to Migraine Frederick Andermann and Benjamin Zifkin Departments of Neurology and Pediatrics, McGill University, Montreal Neurological Hospital and Institute, Montreal, Quebec, Canada Summary: Benign occipital epilepsy of childhood is an idiopathic partial epilepsy syndrome with elementary visual symptomatology, frequently associated with other ictal phenomena. Seizures are usually followed by postictal headache and are often associated with interictal occipital rhythmic paroxysmal EEG activity that appears only after eye closure. In some children the ictal visual symptoms or the interictal EEG abnormalities may not be demonstrated. The clinical and/or EEG manifestations of other forms of idiopathic partial or generalized epilepsy may be found in association. Occipital spikes in nonepileptic children with defective vision, occipital slow spikeand-wave found in some patients with the Lennox-Gastaut syndrome, focal epilepsy due to occipital lesions, seizures originating in the temporal lobe secondary to an occipital abnormality, and complicated or basilar migraine must be considered in the differential diagnosis. Early-onset benign occipital epilepsy or seizure susceptibility syndrome deserves to be considered separately. It has been defined by Panayiotopoulos as consisting of brief, infrequent attacks or prolonged status epilepticus and characterized by ictal deviation of the eyes and/or head and vomiting, occurring in children usually between the ages of 3 and 7 years. Advances in molecular genetics will help decide whether these two disorders are indeed distinct. Benign occipital and benign rolandic epilepsy are commonly associated with migraine. The selective involvement of the occipital lobe in migraine has not been fully explained. The association between benign occipital epilepsy and migraine is likely related to this underlying mechanism as well. The fixation off phenomenon or blocking of occipital epileptic discharges by eye opening is not specific to benign occipital epilepsy of childhood and may be found in symptomatic epilepsies as well. Migraine and epilepsy are distinct disorders, both as far as their pathophysiologic mechanisms and clinical symptomatology are concerned. There is however an overlap in some patients and a causal relationship may exist in some, leading to clinically distinct migraine epilepsy syndromes. Here too, clarification of the molecular basis of migraine and of epilepsy will throw light on the nature of the relationship between the two conditions. Key Words: Benign occipital epilepsy- Children-Migraine-Molecular genetics. An increasing body of literature describes the occipital epilepsies in childhood (1-15). Attempts have been made to delineate separate syndromes on the basis of the clinical manifestations, the electrographic changes, the natural history, or the relationship with migraine. This has led to some degree of uncertainty in the diagnosis, management, and prognosis of these children and young people. This article attempts to clarify these issues and to provide a bridge to the not so benign forms of occipital epilepsy and the lesional epilepsies involving the occipital lobe. Gastaut and Zifkin (16) defined benign epilepsy of childhood with occipital spike-and-wave complexes as partial seizures with elementary visual symptomatology, frequently associated with other ictal phenomena. They Address correspondence to Dr. F. Andermann at Montreal Neurological Hospital and Institute, 3801 University Street, Montreal, Quebec, Canada H3A 2B4. are at times followed by postictal headache and are often associated with interictal occipital rhythmic paroxysmal EEG activity that appears only after eye closure. The existence of this benign epilepsy of childhood with occipital paroxysms (BEOP) had been foreseen years ago by many investigators who nevertheless did not recognize it as a distinct clinical and electrographic syndrome. In 1950, Gastaut (17) reported several cases of benign childhood epilepsy with visual seizures and occipital spike-and-wave complexes. Emphasis was placed on the EEG pattern of high-amplitude occipital spike-and-wave complexes that repeated rhythmically at 2-3/s and appeared with eye closure or intermittent photic stimulation. This pattern was found in the absence of any demonstrable cortical lesion. Gibbs and Gibbs (1 8) noted that Seizure foci in one or both occipital lobes are most commonly found in young children... Occipital foci tend to disappear in adult life, and the subsidence of the electroencephalographic abnormality is usually accom- s9

2 SIO F. ANDERMANN AND B. ZIFKIN panied by a cessation of seizures. Sore1 and Rucquoy- Ponsar (19) described an occipital epilepsy, along with rolandic epilepsy, as a part of their age-dependent idiopathic epilepsy of childhood. Rodin (20) suggested two types of focal epilepsy, one based on a focal structural lesion and the other possibly genetic. Ludwig and Ajmone-Marsan in 1975 (21) suggested the existence of a benign occipital epilepsy in infants and children. Aicardi and Chevrie (22) distinguished a form of benign partial epilepsy in which bilaterally synchronous, symmetric, patietooccipital spike-and-wave complexes, at times very abundant, were seen in place of the usual rolandic spikes. Beaumanoir et al. (23) emphasized the possible relationship between a functional occipital focus and a benign parieto-occipital epilepsy with the same characteristics as benign epilepsy with rolandic spikes. Delwaide et al. (24) described a benign occipital epilepsy of childhood in which the ictal visual phenomena were at times followed by headaches and even vomiting. That so distinct a form of epilepsy had not previously been described is no doubt explained by the fact that its two fundamental characteristics are not always obvious and must be carefully sought. The typical visual symptoms may be forgotten or poorly described by children, especially younger ones, and they may not be elicited when the history is obtained. Moreover, the attenuation of the occipital spikes with opening of the eyes may not be obvious in a routine EEG recording, during which the eyes may be opened only once or twice for a few seconds at a time. CLINICAL OBSERVATIONS In 1987, Gastaut and Zifkin reviewed 63 clinical and EEG observations and divided the cases into three types: a complete syndrome, including seizures with ictal visual symptoms and interictal occipital paroxysmal activity, and two incomplete syndromes, in which either the ictal visual symptoms or the interictal EEG abnormalities were absent (16). About half of the patients had the complete syndrome, 17 had incomplete forms, and in 13 the clinical and EEG features were associated with other forms of epilepsy. Illustrative case histories are drawn from Henry Gastaut s and Ben Zifkin s detailed descriptions. case 1 The complete syndrome was found in this case. The patient was a 23-year-old woman. Seizures began at age 9, with formed visual hallucinations that occurred when she closed her eyes at night. These hallucinations caused her to hide under her bedclothes; they were sometimes followed by a hemiclonic seizure of either side of the body. These seizures occurred before her menses for 9 years, although she was treated with phenobarbital and carbamazepine. Addition of clobazam resulted in com- plete control of the seizures. Antiepileptic drugs (AEDs) were stopped at age 19 and there have been no seizures in the past 4 years. Over a 9-year period, 20 EEGs were performed, all showing typical bilaterally synchronous and symmetrical occipital 3 s spike-and-wave complexes (Fig. 1). With the disappearance of her seizures at age 19, the EEG also returned to normal. Case 2 This patient, a 14-year-old boy, had two brothers with epilepsy and one sister with migraine. Seizures began at age 6, with four attacks of complex visual hallucinations over 2 days. During these attacks the patient attempted to play with toys that he imagined were on the kitchen table. These hallucinations were followed by loss of vision lasting for 4-5 min and headaches with nausea and vomiting. At age 8 he noted sparkling flashes throughout his visual field for several seconds. At age 10, while.he was on an ocean cruise, the flashes were followed by hemiclonic seizures that continued for 2 h until stopped by i.v. diazepam. Over a 7-year period, seven EEGs showed typical occipital spike-and-wave complexes, either bilaterally synchronous, asynchronous, or with shifting lateralization. Several seizures with elementary visual symptoms were recorded. The patient has been free of seizures and epileptiform activity during the past 4 years since the time that phenobarbital was prescribed. Benign epilepsy of childhood with occipital paroxysms but without visual symptoms is illustrated by the following two cases. case 3 This patient was a 9.5-year-old girl. Seizures began at age 7, with 20- to 30-s psychomotor attacks characterized by gestural and verbal automatisms with no reported visual symptoms. Several EEGs showed typical occipital sharp waves over both hemispheres with clear rightsided predominance, which repeated at about Us, as well as independent right-sided rolandic spikes and occasional generalized 3/s spike-and-wave complexes. On one occasion the EEG recorded a psychomotor attack that began during sleep with right occipital discharges and then rapidly generalized (Fig. 2). Seizures persisted for 2 years despite carbamazepine 400 mg/day. The patient has been free of seizures and interictal EEG abnormalities for 1 year since clobazam 20 mg/day was added. case 4 A 14-year-old girl has had seven generalized seizures that occurred during early sleep over a period of 7 years, each followed by severe headache. The two EEGs ob tained have shown typical bilateral occipital spike-andwave activity. An eighth, most recent seizure, which occurred when the patient went to bed before the light was turned out, was immediately preceded by a 1 -min loss of vision (Fig. 3A,B). Epilepsia, VoI. 39, SuppI

3 OVERVIEW OF IDIOPATHIC SYNDROMES SI I P P FIG. 1. Case 1, age 14.5 years. The typical interictal activity of BEOP: bilateral, synchronous, symmetric, surface-negative high-amplitude spike-and-wave complexes repeating at Ws, localized to the occipital and posterior temporal regions. They appear within only 2 s of eye closure. From ref. 16. common average r+ferrnce derivations C 4 4 * c 3 f 4 4 r 3 r z c In these two cases, the absence of visual symptoms at the beginning of the attacks was due to the fact that the first child was amnesic for the attacks and in the second child the seizures began during sleep. Benign epilepsy of childhood with occipital spikeand-wave complexes without occipital paroxysmal activity on eye closure are illustrated next. The seizure pattern in these children was typical but the EEG was normal. They also had normal neurologic examination. This group of patients is illustrated by the following two cases. Case 5 This patient was a 10-year-old boy. A single seizure occurred at 4 A.M. The child awakened with nausea and vomiting and then fell asleep next to his mother, who was awakened an. hour later by her son's generalized seizure. Several minutes after it ended, nausea and vomiting recurred and were followed by a loss of vision lasting for several minutes. An EEG performed the next day and another 1 month later were both normal. Case 6 This patient was a 21-year-old woman whose seizures began at age 8. She had progressive loss of vision over a 10-min period, followed by a lengthy right-sided hemiclonic seizure and a severe headache that lasted for several hours. Phenobarbital and phenytoin were given over 13 years; each attempt to taper the medication was followed by a seizure (Fig. 4A,B). The two EEGs recorded the day after a right hemiclonic seizure showed the expected left posterior slow activity, but seven further EEGs have all been normal. Gastaut concluded that even in the most typical and complete forms of benign occipital epilepsy of childhood the occipital epileptiform activity may not be seen in some recordings. He suggested that such cases are examples of the otherwise typical syndrome in which the EEG manifestations have not yet been recorded. Patients with benign epilepsy of childhood with occipital spike-and-wave complexes associated with clinical and or EEG manifestations of idiopathic generalized epilepsy are illustrated by the following two patients. Case 7 The patient was a 13-year-old girl who had experienced typical absence attacks since age 6. While the patient was changing television channels, three other attacks occurred, which were heralded by phosphenes in

4 s12 F. ANDERMANN AND B. ZlFKlN I J5OY 1 see FIG. 2. Case 3, age 13. Repetitive rhythmic sharp waves over the right posterior quadrant, beginning 6 s after eye closure. Note the spread to homologous areas of the left side after the sudden increase in amplitude of the right-sided discharge. a-activity is also reduced on the right side. From ref. 16. all visual fields and were followed by right-sided hemiclonic seizures with postictal headaches and vomiting that lasted several hours. In seven EEGs, bursts of generalized bilaterally synchronous spike-and-wave activity were seen only during hyperventilation accompanied by clinical absence. The amplitude of visual and somatosensory evoked responses was significantly higher over the left hemisphere. The patient had been initially treated with low doses of barbiturates and, after the substitution of valproate 900 mg/day, no seizures of any kind occurred during the following year and no abnormal EEG activity was seen. Case 8 This patient, an 11-year-old girl, has one paternal uncle with seizures. After age 7, four identical attacks occurred, each in the evening. Sudden brief losses of vision, which the patient herself described, were followed by loss of consciousness and a tonic seizure lasting about 20 s with incontinence of urine, followed by a slow recovery of consciousness with confusion, headache, and vomiting lasting 1-2 h. Of her five EEGs, two were normal and three showed generalized spike-andwave and polyspike-and-wave activity occurring both spontaneously and also with intermittent photic stimulation. No AEDs have ever been prescribed. The association between benign occipital epilepsy and benign rolandic epilepsy is illustrated by the following two cases. Case 9 The patient was a 20-year-old man whose attacks began at age 13. The seizures were heralded by 3-60 s of sparkling flashes in all visual fields, followed by adversion to the left and at times by complete rotation of the body and a subsequent generalized seizure. All nine EEGs obtained have shown generalized, bilaterally synchronous sharp waves with predominance contralateral to the adversion, as seen characteristically in primary generalized versive seizures of adolescents (25). Valproate and phenobarbital were prescribed 6 years ago, and no seizures have since occurred. Case 10 This patient, a 7-year-old boy, had a paternal uncle with primary generalized epilepsy. The patient s seizures first occurred at age 4. He awakened around midnight and had an adversive seizure to the right, followed by

5 OVERVIEW OF IDIOPATHIC SYNDROMES S13 FIG. 3. A: Case 4, age 13. Occipital and posterior temporal ictal discharge with right sided predominance. At A, the child said, Here it comes! I see yellow! At a1 and a2 he was asked what he saw and described lights of a car at night. B: at B, he returned his head and eyes to the left and did not respond. At C, he suddenly asked, Where am I? Treated with phenobarbital 150 rng/day and phenytoin 250 mg/day, he had no further seizures or EEG abnormalities during 2 years of follow-up. From ref. 16. Epilepsia, Vol. 39, Suppl. 4, 1998

6 S14 F. ANDERMANN AND B. ZIFKIN -. the figure 57, which gradually faded away and was followed by loss of vision lasting 20 s, after which she did not respond to questions. During the final 40 s she had oral and gestural automatisms. From ref. 16. Epilepsia. Vol. 39, Suppl. 4, 1998

7 OVERVIEW OF IDIOPATHIC SYNDROMES S15 nausea, vomiting, and sudden copious diarrhea, before losing consciousness for over an hour. Since then, many other attacks of right hemifacial twitching, salivation, and anarthria have occurred. Two EEGs showed typical left occipital spike-and-wave complexes attenuated by eye opening, and left-sided rolandic spikes. The patient was initially treated with low doses of barbiturates. The addition of clobazam 15 mg/day 2 years ago has resulted in seizure control and the disappearance of the epileptiform EEG abnormalities. Gastaut recognized early the association of benign epilepsy of childhood with the clinical and/or EEG evidence of a brain lesion. This is illustrated by the following case histories. Case 11 The patient was a 12-year-old girl with mild developmental delay. The first seizure occurred at 5.5 years, beginning with a loss of vision while the patient was playing in the sea. An EEG performed immediately thereafter showed spike-and-wave activity only at the occipital poles with left-sided predominance, attenuated by eye opening. At age 8, frequent diurnal atonic seizures and nocturnal tonic seizures appeared. The EEG then showed diffuse spike-and-wave complexes with posterior predominance, unaffected by opening of the eyes. With AED treatment the seizures became less frequent and eventually ceased. Four EEGs recorded at ages 8.5, 9, 10, and 10.5 have shown the gradual disappearance of the diffuse spike-and-wave activity and the appearance of typical left occipital spike-and-wave complexes with eye closure. A CT scan revealed mild diffuse cerebral atrophy. Case 12 This patient, a 19-year-old man, had a twin brother and a father with primary generalized epilepsy. During an attack of scarlet fever at age 2, an episode of left-sided status epilepticus occurred, followed by a residual left hemiparesis. The patient was left-handed and had an IQ of 77. Further seizures were noted at age 5, consisting of the hallucination of a luminous ball, followed by deviation of the head and eyes to the left and dysphasia. Since age 15, psychomotor seizures followed by dysphasia without initial visual phenomena have occurred. The EEG showed sharp and slow-wave complexes over both occipital regions, with right-sided predominance seen with eye closure and repeating at Us. A CT scan revealed right hemisphere atrophy similar to that seen with the hemiconvulsion hemiplegia epilepsy (HHE) syndrome. INTERICTAL SYMPTOMS OF BEOP There were 33 males (52%) and 30 females (48%). The age at onset was 15 months to 17 years (mean age 7 years 5 months). Investigation showed that 23 patients had a family history of epilepsy, including febrile convulsions, and 10 cases had a history of migraine. There were seven cases (1 1 %) of prematurity and/or mild perinatal distress, nine cases (14%) of febrile convulsions, three cases (5%) of breath-holding spells, one case of posttraumatic subdural hematoma, one case of skull fracture, and one case of HHE syndrome. Neurologic examination Neurologic examination was normal in 57 patients (90%). Of the six who had abnormal examinations, all were developmentally delayed, five of them only slightly so; two patients had a hemiparesis and one a hemiplegia. Of the patients with a normal examination, 11 (17%) appeared to be of above normal intelligence, and nine had mild behavioral disturbances. Ophthalmologic examination Visual acuity was normal in all patients. The visual fields were normal in the 33 in whom they were examined interictally. Neuroradiologic examination CT scans of 39 patients showed 35 cases to be normal. Angiography, pneumoencephalography, and isotope brain scans were normal in all patients who underwent these procedures (1 1 angiograms, four pneumoencephalograms, and seven isotope brain scans). Visual evoked potentials Flash evoked responses were obtained in 11 patients whose ictal symptoms and/or ictal EEGs suggested a unilateral origin for their seizures. In nine patients these responses were asymmetric, with higher-amplitude late responses seen over the involved hemisphere. Somatosensory evoked responses were obtained in three patients with hemiclonic attacks and these were similarly of higher amplitude over the involved hemisphere. Electroencephalography Each patient underwent several EEGs separated by intervals ranging from several days to 12 years. Background activity was unremarkable for each patient s age, but abundant and unusual interictal epileptiform activity was recorded. There were spike-and-wave complexes in 48 of 63 cases (76%), sharp waves in 19%, and rhythmic posterior slow waves in 5%. Amplitude sometimes reached 300 mv. Distribution was over the occipital and posterior temporal regions (standard electrode positions 01 and/or 02 as well as T5 and/or T6 of one hemisphere, most often the left, or over both hemispheres simultaneously or independently). It was usually seen as rhythmically repeated paroxysms from 1 to 3/s, occurring in bursts or in trains that might continue for virtually the entire recording with eyes closed. On rare occasions, isolated paroxysmal activity was seen at irregular intervals. Epilepsia, Vol. 39. Suppl. 4, 1998

8 S16 F. ANDERMANN AND B. ZIFKIN Prompt disappearance of the paroxysmal activity with eye opening was seen in 59 of the 63 patients (94%). The paroxysmal discharge was seen almost exclusively after eye closure, usually with a latency of 1-10 s. Very occasionally, latencies of up to 20 s were seen. Before BEOP had been fully described, Panayiotopou- 10s (26) showed that this interictal activity is not suppressed with the eyes open in darkness but that subsequent fixation on a very small light source abolishes it. He concluded that the loss of macular vision with eye closure was responsible for the appearance of the epileptiform activity. He introduced the term fmation-off sensitivity to describe this phenomenon, a term that is now widely accepted. The variability of the associated clinical and seizure manifestations was recently studied by Maher et al. (27). Intermittent photic stimulation Evoked bursts of occipital sharp waves unrelated to eye opening and closing occurred in seven patients who did not display such activity at rest. This effect of IPS was not related to the stimulus frequency and occurred several seconds after the beginning of stimulation, suggesting that the resulting bursts were related to an induced hyperexcitability of the occipital cortex rather than being a response locked to the photic stimuli. In seven other patients with typical occipital paroxysmal activity with eye closure, IPS evoked generalized bursts of spikeor polyspike-and-wave activity, at times with associated myoclonus. Hyperventilation had no significant effect on the occipital epileptiform activity. Sleep EEGs were obtained after partial sleep deprivation in 17 patients, and spontaneous sleep was recorded in five others. In 13 of these 22 patients, the occipital paroxysms disappeared with drowsiness and were not seen during the following slowwave sleep. Conversely, in five patients occipital paroxysmal activity, not seen in wakefulness, was noted. Five patients also had apparent electrographic seizures over occipital regions during slow-wave sleep, with no clinical accompaniment. In six patients, bursts of generalized bilaterally synchronous spike-and-wave complexes were seen. Generalized bilaterally synchronous spike-and-wave or polyspike-and-wave complexes, characteristic of primary generalized epilepsy and rolandic spikes, typical for benign partial epilepsy of childhood, were recorded in 19 of the 50 children (38%) who had no clinical evidence of either of these forms of epilepsy. The coexistence of rolandic spikes and typical occipital paroxysmal activity was seen principally in patients with hemiclonic seizures. It is noteworthy that a similar association was described by Beaussart (28) in children with benign partial epilepsy and nocturnal hemiclonic seizures. EEG ictal features Several examples are described by Gastaut and Zifkin (16). In Case 13, a seizure lasting 80 s began with phosphenes in all visual fields, followed by turning of the head and eyes to the left, and was accompanied by right posterior temporal and occipital ictal activity that then spread throughout the right hemisphere. In Case 14, a 3-min attack began with visual hallucinations, followed by mild confusion associated with left occipital ictal activity. In Case 15, an attack lasting 30 s occurred in stage 3 of sleep and consisted of motor and verbal automatisms. It was associated with ictal activity that began in the right occipital region and rapidly spread over both hemispheres. Pmtictal symptoms The postictal symptoms of BEOP were quite specific and were only rarely encountered in any other type of epilepsy. Diffuse headache, only rarely hemicranial, was found in 33% of the cases, whereas nausea and vomiting, at times associated with headache, were seen in 17%. Postictal diarrhea was observed in 3%. These postictal phenomena became less severe and disappeared over several hours without any disturbance in the level of consciousness. Their postictal quality was further emphasized by the absence of any ictal discharge in the 13 cases in which the EEG was recorded during the postictal headache. This association of headache, nausea, and vomiting lasting several hours, so characteristic of migraine, is further discussed below. Frequency and precipitating factors The frequency of BEOP attacks is extremely variable, ranging from a few attacks daily over several months to occasional seizures followed by seizure-free intervals of several years. Because recognition of this syndrome depends on a characteristic visual aura, it may not be recognized when attacks occur during sleep. However, typical occipital electmgraphic seizures have been recorded during sleep without any clinical accompaniment. It is therefore possible that some of the seizures of BEOP may occur during sleep but go unreported as such because the typical history of visual disturbance cannot be obtained. Although there are no clear precipitating factors, ambient light and the menstrual cycle contribute to the occurrence of BEOP attacks. Seizures that were apparently precipitated by light were reported by 16 patients. Some occurred ongoing from a dark area into a brighter one, or, conversely, from a well-lit area into a darkened one, as described by Pazzaglia et al. (29). Three occurred while on a sunlit beach or at sea, or while watching television. In five of 12 pubescent girls with BEOP, seizures began with puberty and continued to occur with menstruation. Epilepsia. Vol. 39, Suppl

9 0 VERVIE W OF IDIOPATHIC SYNDROMES S17 Prognosis of BEOP BEOP generally carries a good prognosis, although it is not as benign as childhood epilepsy with rolandic spikes. In none of these cases have the seizures persisted into middle adulthood, and other types of recurring seizures in adulthood were seen in only three patients. The prognosis, as expected, is poorer in patients with other evidence of cerebral disturbance, such as mental retardation or abnormal imaging, or in those whose EEGs show an additional secondary epileptiform disturbance. Complete seizure control with AEDs was achieved in 38 patients (60%), and seizures persisted in three patients followed beyond age 19. Seizure control can be achieved, although the epileptiform EEG may persist for several months or years after the seizures have ceased. Almost all available AEDs have been tried, but none is the obvious drug of choice. The rapid response to clobazam described by Gastaut and Low (30) is noteworthy. In seven of nine patients in whom it was used, both seizures and interictal spikes ceased after only a few days of treatment. Whether the last-generation AEDs offer any advantages because of lower effective doses or fewer side effects is as yet unclear. Diagnosis of BEOP Diagnosis is not difficult in patients with the complete syndrome. However, because the presence of occipital interictal discharges increases suspicion and directs the physician in seeking a history of visual disturbance, diagnosis is more difficult in cases in which the syndrome is incomplete, particularly when the characteristic EEG pattern is absent. Therefore, many cases of BEOP may be missed if a history of visual phenomena is not carefully sought. Differential diagnosis The diagnosis of BEOP should not be made when seizures first occur in adult life. Although several authors (31-34) have described occipital seizures evoked by intermittent photic stimulation in adults without evidence of brain lesions, these cases are rare and may not constitute a separate type of epilepsy. Five conditions must be differentiated from BEOP. (a) Occipital needle-like spikes (18,23,35) may be seen with eye closure in nonepileptic children with defective vision. This unusual pattern is not associated with seizures. (b) Very rarely, the EEGs of patients with secondarily generalized epilepsy suggesting the Lennox- Gastaut syndrome show occipital slow spike-and-wave complexes with eyes open as well as closed, in place of the more usual generalized epileptiform activity. (c) Focal epilepsy caused by occipital lesions is much rarer in childhood than BEOP. It is characterized by a less complex ictal pattern, usually marked by visual seizures followed by tonic head and eye deviation. In such patients, the neurologic and neuro-ophthalmologic examinations, as well as the brain imaging, may be abnormal. The interictal EEG shows intermittent spikes, which do not react to eye opening, over a single occipital lobe where continuous polymorphous 8 activity, or attenuation of background activity, also suggests a focal lesion. (d) Complex partial seizures arising in the temporal lobe are common in children. The two may be confused in those rare instances in which the psychomotor attacks are preceded by visual illusions or hallucinations. The two can be differentiated clinically at times, but the EEG is especially useful in showing the typical anterior sylvian abnormalities associated with epilepsy arising in the temporal lobe. There also is, however, preferential electrographic spread to the temporal lobe from infracalcarine regions, often leading to electroclinical manifestations suggestive of temporal origin. Independent contralateral temporal interictal and ictal discharges may also be found (36). These abnormalities, evidence of secondary epileptogenesis, were often a source of misdiagnosis until the significance of posterior temporal or occipital lesions was recognized (36,37). (e) Basilar migraine (38) is usually seen in adolescent girls who subsequently develop common migraine. It is characterized by a prodrome, typically lasting up to 30 min, of visual loss or flashing lights, together with signs of pontine and midbrain dysfunction such as vertigo, dysarthria, bilateral paraesthesia, or transient loss of consciousness, followed by headache lasting several hours. No paroxysmal EEG activity is often seen during or between these attacks, but bilateral posterior slow waves may be seen for hours or days afterwards (39). The absence of interictal epileptiform occipital activity attenuated by eye opening in many patients with basilar migraine (38,40-44) is of major importance in differentiating basilar migraine from BEOP. In some patients with a convincing history of basilar migraine, seizures may occur in the course of an attack, representing one of the migraine epilepsy syndromes discussed below (Fig. 5) (45). HOW MANY SYNDROMES OF BENIGN OCCIPITAL ONSET ARE THERE? Determined attempts were made in recent years to subdivide the syndrome of benign epilepsy of childhood with occipital spike-and-wave complexes and these have originated mainly from Panayiotopoulos and his colleagues. In a recent article, Ferrie et al. (14) attempted to delineate a group of children with early onset benign occipital seizure susceptibility syndrome. They defined it as a syndrome of brief or prolonged (status), infrequent partial seizures characterized by ictal deviation of the eyes andor head and vomiting. Seizures are usually nocturnal and frequently evolve to hemiseizures or generalized tonic-clonic attacks. Onset is usually between 3

10 A G.P. Aga 11 YC.K F3C3 14 I loop4 a P3-01 F d C A k B G.P. Aga 10 M.C.H P E/O Drowsy C G.C. Age 11% MCR F3C3 1.6 I1oopv C 3 - P 3 W W P F&C4 4 C4-P4& P FIG. 5. A-C: A 10-year-old girl developed attacks at age 9. In these she saw circular lights of many colors over the visual field, although occasionally only to the left. They resembled the afterimage that results from pressing or rubbing the eyes. After about 10 min everything would suddenly go black. The attack might stop at this point, with vomiting and gradual clearing of vision over a period of 1-15 min. On rare occasions the attack progressed to a gradual fading of consciousness. She slowly sat or lay down and became flaccid for as long as 40 min, with saliva running from her mouth. On some Occasions such episodes were associated with a headache that lasted 1-2 h. This headache was usually bitemporal, although occasionally it consisted of a unilateral pressure sensation. Her father had common migraine. Several times a year he was sick for a day at a time with pounding headache and nausea. The patient was a pleasant, intelligent adolescent with a normal neurologic examination. CT scan suggested mild cerebellar atrophy, inferred from an increase in the size of the fourth ventricle. She was treated with phenytoin and phenobarbital. EEGs at ages 10 and 11 demonstrated striking and very active bilaterally synchronous, shape and slow-wave and spike-and-wave epileptogenic discharges at 1.5 to 2 Hz over posterior head regions (A,@. These were enhanced by hyperventilation and during drowsiness were replaced by repetitive and at times rhythmic spikes blocked by arousal or eye opening. Whenever epileptogenic activity was not present, the background activity was normal. An EEG taken at age 11 was completely normal (C).

11 OVERVIEW OF IDIOPATHIC SYNDROMES S19 and 7 years of age, with remission after 1-2 years. All patients are seizure-free by 12 years. The sexes are equally affected. EEG shows occipital paroxysms, which consist of high-amplitude runs of repetitive occipital spike and sharp and slow-wave complexes, which are often bilateral but with a right-sided predominance. Occipital paroxysms attenuate on eye opening and fixationoff sensitivity is frequently demonstrated. On subsequent EEGs, over one-third of patients demonstrate centrotemporal and/or frontal and/or somatosensory evoked spikes. Thus, the ictal tonic deviation of the eyes and head, the vomiting, and the frequency of partial status epilepticus are characteristic. Ferrie et al. (14) stress the benign prognosis, with remission within 1-2 years and the frequency with which single seizures are the only manifestation. This obviously has considerable impact on the need to treat these children, which must be determined on an individual basis. The authors suggest early withdrawal of medication. Ferrie and Panayiotopoulos emphasize that headache is uncommon in this early form. This may relate to the younger age of the patients and may be due, just like the paucity of visual manifestations, to their limited ability to report symptoms. They also stres,s that seizures in the younger variant &e predominantly nocturnal, whereas they occur mainly in the daytime in children with benign epilepsy of childhood with occipital spike-and-wave. They analyzed the histories and findings in 139 patients and excluded six with symptomatic or cryptogenic origin and eight with ictal features classical of benign epilepsy of childhood with occipital spike and wave complexes. Of the remaining 125, 80 had tonic eye and head deviation and vomiting without visual symptoms, but their EEGs demonstrated occipital paroxysms, which attenuated on eye opening and fixation. They defined as having an incomplete syndrome 33 children in whom (a) either ictal vomiting or tonic eye head deviation was not described, (b) occipital symptoms were not demonstrated, or (c) EEG abnormalities (if demonstrated) failed to attenuate on eye opening or central fixation. They also distinguished a third group of 12 patients in whom ictal tonic eye and/or head deviation and vomiting were associated with visual symptoms. It is apparent from these descriptions that the younger the children are the less likely they are to describe visual symptoms and headache and the more likely they are to present with vomiting and with tonic eye and head deviation. Whether fhe tonic deviation is related to spread to frontal structures or whether it depends on the occipital epileptic discharge itself is not clear. As in other benign epilepsies of childhood, some patients are described both with the earlier and later variants to have normal EEGs. This is found mainly early in the course of the illness and is of course not terribly unusual in patients with benign rolandic epilepsy. Opinions are divided as to whether the syndrome of childhood epilepsy with occipital paroxysms should be split into an early and a late form. Many pediatric neurologists find merit in such further subdivision because there may be some indications for differences in prognosis and in possible approaches to treatment. However, the pathophysiology of the disorder appears to be quite similar in both groups of children, and perhaps further advances in molecular genetics will help to determine whether these are in fact separate entities. The relationship of benign occipital to benign rolandic and other benign partial epilepsies of childhood has recently been addressed by Panayiotopoulos (46) in an editorial, in which they were subsumed as benign childhood epilepsy susceptibility syndromes. MIGRAINE AND THE BENIGN IDIOPATHIC PARTIAL EPILEPSIES OF CHILDHOOD Bladin (47) has stressed the strong association that exists between classical and common migraine and benign rolandic epilepsy. This has also been our impression, that of Watters (personal communication), and of others. However, this opinion is not universally shared by pediatric neurologists. The reason for this difference of opinion is difficult to explain. It probably relates to the differences in what neurologists consider to represent migraine, despite the efforts of the International Headache Society to introduce a standardized classification. Further studies and, in particular, family investigations could certainly clarify this issue. It is, however, clear that most of the manifestations of classical migraine or migraine with aura involve the occipital lobe, although the specific reasons for this form of pathoclysis are not yet clear. Occipital epilepsy, whether idiopathic or symptomatic, is also closely associated with vascular headache. Vascular malformations in the posterior circulation have been shown to lead not only to migraine but to a migrainous aura with a classical march (see ref. 48 for review). Gastaut and Zifkin (16) discuss the etiology of benign epilepsy of childhood with occipital spike-and-wave complexes, and again we quote extensively from their 1987 report. In our patients, the majority of seizures began with brief unilateral teichopsia or hallucinations that were characteristically epileptic. This was confirmed in six patients who had seizures during EEG monitoring. Headache was seen in only one third of patients and, even in these, it only followed some, but not all of their attacks. Their headaches were not always typically migrainous. There has been debate for many years about the role of migraine in patients with benign occipital epilepsy. Gastaut modestly disclaimed expertise in migraine but did attempt to explain its occurrence in children with benign epilepsy and occipital paroxysms as being related to the

12 s20 F. ANDERMANN AND B. ZIFKIN epileptic discharge spreading into vasomotor centers in the brainstem. Such an explanation, however, would hardly explain the familial occurrence of migraine that is so often demonstrated. Another attempt at ignoring an association between migraine and occipital epilepsy is presented by Ferrie et al. (14) in their paper on early onset benign occipital seizure susceptibility syndrome. They recognized the occurrence of migraine in only 3% of families of children with occipital epilepsy, and this figure is considerably less than what is found in the population at large. They rightly stressed that this was not an epidemiologic study and that the relationship of migraine and epilepsy was not specifically addressed. A similar argument regarding the relation between benign rolandic epilepsy, and migraine was brought forward in a case study from the group of Giovanardi Rossi (49). Migraine and epilepsy are distinct conditions with specific and separate pathophysiologic mechanisms. However, they are associated in a number of patients and there may be a causal relationship in some. Clinical migraine epilepsy syndromes are listed below Epileptic seizures induced by a classical migraine aura. Epilepsy with continuing seizures no longer triggered by migrainous aura. Epilepsy due to gross cerebral lesions caused by migraine. Benign occipital epilepsy of childhood and the spectrum of the occipital epilepsies. Benign rolandic epilepsy. Malignant migraine, a manifestation of mitochondrial encephalomyopathy (MELAS syndrome). Migraine attacks following partial complex seizures. Alternating hemiplegia of childhood. Of particular interest are the epileptic seizures induced by a classical migraine aura that is almost always occipital (Fig. 6A-C). These have also been described as intercalated attacks (50). The concept that spreading depression is central to the mechanisms of the migraine aura has been proposed by Lauritzen (51). Spreading depression does not usually cross major sulci. The motor epileptic manifestations of intercalated attacks may be due to spreading depression extending anteriorly and exceptionally crossing the central sulcus to involve the precentral gyrus. This association illustrates the close relationship between occipital localization and epileptic seizures in patients with this migraine epilepsy syndrome. More recently, benign familial temporal lobe epilepsy has been identified (52). In many of these families, d6ja vu is overrepresented as an aura and many affected individuals may have only isolated d6ja vu. In these families, migraine is also overrepresented, and this has led in the past to the diagnostic impression that the d6ja vu may represent a manifestation of migraine rather than of epilepsy (53). Patients with not so benign occipital epilepsy The nonspecificity of occipital spike-and-wave complexes suppressed by eye opening has been addressed by Aicardi and Newton (54). They presented a series of patients who illustrate the association of this pattern with basilar migraine in some patients and with benign occipital epilepsy syndrome in others. They stressed the absence of visual phenomena in some of these individuals, perhaps because of age or lack of intellectual ability. They also stressed that the benign course previously emphasized by Gastaut (55) and by Beaumanoir and coworkers (5637) is by no means always found. Several of their patients had seizures that were difficult to control and an equal number had educational problems. They stress, in particular, that in some patients the prognosis was certainly not as good as in benign epilepsy of childhood with rolandic spikes. Aicardi and Newton (54) attribute these differences to the impact of their referral base, resulting in patients with more severe epilepsy being overrepresented in their series, which was collected at a center dedicated mainly to the care of children with severe neurologic disorders. They also stressed that the reliability of family histories of migraine in retrospective studies is notoriously limited. Aicardi and Newton conclude that, on the basis of their case material, the rhythmic occipital slow and sharp waves suppressed by eye opening are not representative of any specific clinical entity but rather were found in a wide spectrum of children with varied symptomatology, neurologic and intellectual deficits, and response to treatment. They also suggested that perceptual problems that might lead to learning difficulties should be suspected and tested for, even in children who show good response to treatment. Atypical evolution of benign partial epilepsy in children has also been discussed by Fejerman (58). CONCLUSIONS Benign occipital epilepsy of childhood is a wellestablished syndrome among the benign partial epilep sies. It shows a relationship to benign rolandic epilepsy and to idiopathic epilepsy with spike-wave discharges, but the fundamental basis of this relationship is imperfectly understood. Molecular studies will most likely clarify this issue. The relationship of benign occipital epilepsy of childhood to migraine is at present based on clinical observation (59,60). The selective involvement of the occipital lobe in migraine has not been fully explained. The association between benign occipital epilepsy and migraine is probably related to the predilection of migraine for occipital involvement as well. The fixation-off phenomenon or blocking of occipital epileptic discharges by eye Epilepsia. Vol. 39. Suppl. 4, 1998

13 FIG. 6. A-C: A 27-year-old man began to experience migraine at age 11. A dazzling white, round, multifaceted light began to appear in his left visual field and spread to the right, rendering him completely blind for 2-15 min although he remained fully conscious. Sometimes a tingling sensation developed on the left side of his body. On rare occasions, starting during his early attacks, he became confused but not unconscious and then developed tonic-clonic adversive seizures involving the left side of the face, with jerking of the head and eyes to the left for several minutes. Whether or not a seizure developed, these episodes terminated with a severe posterior pounding headache, sometimes with copious vomiting, followed by sleep for 1-2 h. His mother and father had classical migraine, with scintillating scotomata in one visual field, followed invariably by headache. One brother had also classical migraine, with concentric fortification spectra and retained central vision. Another brother had common migraine. Attacks gradually abated and AEDs were stopped at age 18. At age 27 he was a successful interior decorator. He had occasional phosphenes when he was tired and closed his eyes. His EEG at age 12 showed almost continuous spike, sharp-wave, and slow activity, the latter at Hz, from the right posterior temporal and parieto-occipital areas (A). At age 18 (B), continuous, sometimes rhythmic slow-sharp-wave activity, blocking with eye opening, was seen in the same distribution. Abnormalities were similar but of lower voltage at age 25 (C). B and C are continuous single-channel tracings. E/O, EO, eyes open; E/C, EC, eyes closed. tl 3 &

14 s22 F. ANDERMANN AND B. ZIFKIN opening is not specific to benign occipital epilepsy in childhood and may be found in some symptomatic epilepsies as well (Fig. 6A-C). Benign occipital epilepsy occurs in small children with a somewhat different clinical pattern from that in older children. This clinical difference justifies recognition of an early benign occipital seizure susceptibility syndrome or early benign occipital epilepsy of childhood, according to several pediatric epileptologists (N. Fejerman, J. Aicardi, personal communication). The distinction between benign occipital epilepsy and not so benign cryptogenic forms of partial epilepsy involving that area suggests a need for caution in expressing a prognosis. There clearly exists a range from the entirely benign to more severe forms of occipital epilepsy. The difference between the benign and more severe forms is more difficult to establish compared to what one finds in the rolandic epilepsies, perhaps because of the greater specificity of EEG discharges of BECRS compared to the EEG findings in the various forms of occipital epilepsy. Migraine and epilepsy are distinct conditions as far as their pathophysiologic mechanisms and clinical symptomatology are concerned. However, there is an overlap in some patients and a dausal relationship may exist in some. The distinction between a migrainous and an epileptic aura is based on clinical observation only, and at times distinction may be difficult because of the prolonged duration of some occipital epileptic auras (61). Clarification of the molecular basis of migraine and of epilepsy will most likely throw light on these issues and, indirectly, also on the mechanisms of the benign partial epilepsies of childhood. REFERENCES 1. Kuzniecky R, Gilliam F, Morawetz R, Faught E, Palmer C, Black L. Occipital lobe developmental malformations and epilepsy: clinical spectrum, treatment, and outcome. Epilepsiu 1997;38: Guemni R, Dravet C, Genton P, et al. Idiopathic photosensitive occipital lobe epilepsy. Epilepsiu 1995; Sveinbjornsdottir S, Duncan JS. Parietal and occipital lobe epilepsy: a review. Epilepsiu 1993;34: Holmes GL. Benign focal epilepsies of childhood. Epilepsiu 1993; 34(suppl 3):S Kivity S, Lerman P. Stormy onset with prolonged loss of consciousness in benign childhood epilepsy with occipital paroxysms. J Neurol Neurosurg Psychiatry 1992;55: Lerman P, Kivity S. The benign partial nonrolandic epilepsies. J Clin Neurophysiol 1991 ;8: Nagendran K, Prior PF, Rossiter MA. Benign occipital epilepsy of childhood: a family study. J R Soc Med 1990,93: Panayiotopoulos CP. Benign childhood epilepsy with occipital paroxysms: a 15 year prospective study. Ann Neurol 1989; Nalin A, Ruggerini C, Ferrari E, Galli V, Ferrari P, Finelli T. Clinical aspects, differential diagnosis and evolution of visual epileptic seizures in children. Neurophysiol Clin 1989; Panayiotopoulos CP. Benign nocturnal childhood occipital epilepsy: a new syndrome with nocturnal seizures, tonic deviation of the eyes, and vomiting. J Child Neurol 1989; Jaffe SJ, Roach ES. Transient cortical blindness with occipital lobe epilepsy. J Clin Neuro-ophthulmol 1988;8: Fois A, Malandrini F, Tomaccini D. Clinical findings in children with occipital paroxysmal discharges. Epilepsiu 1988; Kuzniecky R, Rosenblatt B. Benign occipital epilepsy: a family study. Epilepsiu I987;28: Fenie CD. Beaumanoir A, Guerrini R, et al. Early-onset benign occipital seizure susceptibility. Epilepsiu 1997;38: Takaishi Y, Hashimoto K, Enokido H. A study of idiopathic epilepsy of childhood with occipital electroencephalographic foci. J Nippon Med Sch 199 I ;58: Gastaut H, Zifkin B. Benign epilepsy of childhood with occipital spike and wave complexes. In: Andermann F, Lugaresi E, Eds. Migruine and epilepv. London: Butterworth Heinemann, I. 7. Gastaut H. Evidence tlectrographique d un mtcanisme souscortical dans certaines tpilepsies partielles-la signification clinique des secteurs artothalamiques. Rev Neurol 1950;83:3% Gibbs F, Gibbs E. Atlas of elecrroencephulogruphy. Vol. 2. Cambridge, MA: Addison-Wesley, Sore1 L, Rucquoy-Ponsar M. L Epilepsie fonctionelle de maturation. Rev Neurol 1969;121: Rodin E. Discussion. In: Alter M, Hauser WA, eds. The epidemiology of epilepsy. Bethesda: US Department of Health, Education, and Welfare Publication No. NIH , 1972:llO. 21. Ludwig B, Ajmone-Marsan C. Clinical ictal patterns with occipital EEG foci. Neurology 1975;25: Aicardi J, Chevrie J. Epilepsies partielles Wnignes de I enfant. In: Congrts de la societt de neurologie infantile. Marseille: Diffusion Gtntrale de Librairie, 1977: Beaumanoir A, Inderwildi B, Zaguri S. Paroxysmes EEG non tpileptiques. Med Hyg Geneve 1981;39: Delwaide PJ, Barragan M, Gastaut H. Remarques sur I epilepsie partielle occipitale. Acru Neurol Belg I97 1;71: Gastaut H, Lemolo F, Menendez-Gonzalez P. L Epilepsie versive a pointe-ondes generalistes de I adolescent. In: Epilepsies de 1 adolescent et adolescents epileptiques. Paris: Documentation Medicale Labaz, I98 1 : Panayiotopoulos CP. Inhibitory effect of central vision on occipital lobe seizures. Neurology 1981;3 I: Maher J, Ronen GM, Ogunyemi AO, Goulden KJ. Occipital paroxysmal discharges suppressed by eye opening: variability in clinical and seizure manifestations in childhood. Epilepsiu 1995; Beaussart M. Benign epilepsy of children with rolandic foci. Epilepsiu 1972; I3:795-8 I Pazzaglia P, Sabattini L, Lugaresi E. Crisi occipitali precipitate dal buio. Riv Neurol Psychiatry 1955;18: Gastaut H, Low M. Antiepileptic properties of clobazam. Epilepsiu 1979;20: I. Naquet R, Fergerstein L, Bert J. Seizure discharges localized to the posterior cerebral regions in man provided by intermittent photic stimulation. Electroencephulogr Clin Neurophysiol 1960,12: Gastaut H, Franck G, Krolikowska W, Naquet R. Regis H, Roger J. Etude des potentiels Cvoquts visuels chex les htmianopsiques prksentant des crises tpileptiques visuelles dans leur champ aveugle. Rev Neurol 1963;108: Fischer-Williams M. Bickford RG, Whisnant JP. Occipito-parietotemporal seizure discharge with visual hallucinations and aphasia, Epilepsiu 1964,5: Swanson JW, Vick NA. Basilar artery migraine. Twelve patients with an attack recorded electmncephalographically. Neurology 1978;28: Gibbs F, Gibbs E, Gibbs T. Relation between specific types of occipital dysrhythmia and visual defects. Johns Hopkins Med J 1968; 122: Palmini A, Andermann F, Dubeau F, et al. Occipitotemporal epilepsies: evaluation of selected patients requiring depth electrode studies and rationale for surgical approaches. Epilepsiu 1993;M Fish D, Andermann F, Olivier A. Complex partial seizures and Epilepsia, Vol. 39, Suppl