Multilobar polymicrogyria, intractable drop attack seizures, and sleep-related electrical status epilepticus

Epilepsy with continuous spike-and-wave activity during slow sleep (known as electrical status epilepticus during sleep; ESES)^1,2 is included among the epileptic syndromes with both focal and generalized electroclinical features. Children with this disorder have infrequent partial motor seizures during sleep and subsequently have frequent generalized atypical-atonic absence seizures. EEGs show both focal and generalized interictal discharges; sleep recordings show continuous, usually generalized spike-and-wave activity during slow-wave phases. Seizures and major EEG abnormalities appear between the ages of 2 to 10 years, with a peak at 4 to 5 years of age. The seizures can respond to treatment or be resistant, lasting months or years, and invariably remit by the age of 13 years.³

Neuropsychological impairment may emerge, especially with longer duration, and can persist indefinitely.⁴ Patients are almost equally divided between those with a cryptogenic and those with a symptomatic origin of ESES.^2,5 No specific factor has been suggested as more likely to cause ESES. Although epilepsy with ESES appears uncommon it might be underreported, because few centers perform routine sleep EEG recordings.

We describe nine patients with MRIs that demonstrated multilobar polymicrogyria who had ESES followed by stable seizure remission. This good outcome contrasts with that typically observed with polymicrogyria and other cortical malformations, which show a very low seizure remission trend.^6-9 Among our patients, electroclinical and imaging findings suggested age-related secondary bilateral synchrony (SBS), development of which may have been favored by the type, site, and extent of the structural lesion.

Methods and patients. Between 1988 and 1997, at four medical centers, we saw 11 patients with polymicrogyria and ESES. These patients represented 8% of the range of patients with malformations of cortical development and epilepsy^10,11 seen by the authors and 18% of those with polymicrogyria. We report on nine of the 11 patients whose follow-up periods (mean duration, 12 years; range, 4 to 19 years) extended beyond cessation of ESES. The other two patients, 4 and 6 years of age, had brief follow-up periods, and ESES was present at the time of this writing.

All patients underwent brain MRI with 0.5-tesla or 1.5-tesla apparatus. Noncontrast-enhanced spin echo, inversion recovery, and gradient echo sequences were performed in the axial, sagittal, and coronal planes. All patients were examined with standard 5-mm slice thickness. The diagnosis of epilepsy with ESES was made on the basis of both clinical and serial sleep EEG recordings. All patients underwent repeated video-EEG recording while awake and asleep. The 10-20 International Electrode Placement System was used. Eight patients had both all-night and afternoon recordings of at least one sleep cycle, and one had only afternoon recordings. For Patients 5 and 8, we studied the temporal and spatial characteristics of spike-and-wave activity. The EEG signal was filtered with a bandpass of 1 to 100 Hz and digitized at the sampling rate of 1024 Hz with the average reference. We calculated interhemispheric time differences by determining the latency between the negative peaks of 60 bilateral spikes, of which 20 were taken at the beginning, 20 at the center, and 20 at the end of the spike-and-wave discharges.

Seizures and epilepsy were classified according to the recommendations of the International League Against Epilepsy.^12,13 All patients underwent clinical examinations before, during, and after the ESES period. Most who could be tested underwent IQ assessment with various scales (Wechsler intelligence scales, Terman Merrill Scale, Leiter International Performance Scale, developmental test of visual-motor integration) or assessment of adaptive behavioral criteria¹⁴ during and after ESES. Mental retardation was classified as mild, moderate, severe, or profound.

Results. General clinical findings, seizure patterns, and results of cognitive evaluations are shown in the table. For all patients, MRI showed irregular infolding of thickened gyri with multilobar distribution. For two patients (5 and 9), the cortical abnormality was bilateral, involving the perisylvian regions and extending anteriorly to the posterior third of the inferior frontal lobes, a morphologic pattern corresponding to the bilateral perisylvian syndrome¹⁵(figure 1, A and B). For the other seven patients (1 through 4, 6 through 8) (figure 1, C through F), the abnormal cortex was limited to one cerebral hemisphere, centering on the sylvian fissure and variably involving the surrounding lobes. The affected hemisphere was mildly to severely reduced in size compared with the contralateral one. For all patients, close examination of the malformed cortex revealed irregular digitations at the junction between the gray and the white matter overlaid by small, fused gyri, and appearance characteristic of polymicrogyria. The underlying white matter appeared mildly reduced in volume, and the overlying subarachnoid spaces were slightly enlarged. The signal intensity of the white matter and the appearance of the remainder of the cerebral cortex, corpus callosum, and cerebellum were normal. In three patients in whom the malformed hemisphere was markedly reduced in size, the homolateral cerebral peduncle was smaller than the contralateral one.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Axial MR images of Patients 5 (A), 9(B), 1 (C), 4 (D), 6 (E), and 8 (F). (A), Fast spoiled gradient recalled steady-state sequence (FSPGR), 1.5 T. (B) T1-weighted, 1 T. (C) T2-weighted, 0.5 T. (D) T1-weighted, 0.5 T. (E) T1-weighted, 1 T. (F) Inversion recovery weighted (IRW), 0.5 T. Right is on reader's left for all images. Patients 5 and 9 show bilateral perisylvian polymicrogyria with fairly symmetric distribution. Unilateral polymicrogyria is shown for Patients 1 (right hemisphere), 4 (left hemisphere), 6 (right hemisphere) and 8 (left hemisphere), in whom the affected hemisphere is markedly reduced in size and shows a diffusely abnormal sulcal pattern. The normal digitations between gray and white matter are reduced and have been replaced by many small gyri and by areas of irregular cortical thickening. Patient 1 (C) also showed bilateral atrophic changes probably from a recent trial of adrenocorticotropic hormone.

Three patients had a first-degree relative with epilepsy or febrile seizures. No patient had a family history of congenital anomalies or mental retardation. No patient had known exposure to teratogens during pregnancy. However, the pregnancy of the mother of Patient 4 was complicated by severe head injury with fractured skull at the 16th week of gestation and that of the mother of Patient 5 by peritonitis treated surgically at the 10th week of gestation.

Epilepsy began between the ages of 14 months and 5 years with infrequent focal motor seizures for all patients. ESES was first detected between the ages of 2 and 5 years. It coincided with the onset of intractable atonic absences for eight patients (Patients 1 through 6, 8, and 9)(figure 2). One patient (7) showed lateralized ESES without absences. ESES lasted from 1 to 10 years and fluctuated in severity. Visual quantification of EEG abnormalities during sleep, performed for all patients, revealed a spike-and-wave index (percentage of the tracing occupied by spike-and-wave activity) ranging from 35% to 85% during non-REM sleep and alternating with disappearance of generalized abnormalities during REM sleep(figure 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Temporal relation among frequency of drop attack (atonic absence) seizures, ESES, and age for Patients 1 through 6, 8, and 9. The solid curve represents the mean frequency of drop attack seizures, and the shaded lines standard deviation. The entire length of the curve corresponds to duration of the follow-up period. The horizontal lines on the bottom correspond to the period in which drop attack seizures and ESES were present for each patient. Pat = patient; ESES = electrical status epilepticus during sleep.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. EEG recording for Patient 3. (Left) Patient awake. Repetitive spikes over the left hemisphere with temporo-occipital predominance. (Center) Continuous diffuse spike-and-wave activity at about 2.5 Hz during slow sleep. (Right) During REM sleep. Disappearance of continuous spike-and-wave activity and presence of repetitive left temporo-occipital spikes, left central slow-wave activity, and independent spike activity in the right frontal region.

Interhemispheric time differences for Patients 5 and 8 varied from 10 to 20 ms (Patients 5 mean ± SD, 12.1 ± 2.7; Patient 8 mean± SD, 18.0 ± 3.4) with the leading side constant. There were no significant changes in interhemispheric time differences at the beginning or end of discharges. In addition to the diffuse spike-and-wave activity during slow-wave sleep, all patients showed focal or multi-focal spikes while awake, clearly predominant over the centrotemporal area of the hemisphere from which focal motor seizures originated.

Drop attack seizures were recorded by means of video-EEG polygraphy for all except Patient 7. The seizures were found to be caused by loss of muscle tone during slow spike-and-wave discharges (figure 4). Severity was extremely variable from one patient to another and frequently from one episode to another for the same patient. Thus drop attacks could be extremely rapid and without postural reaction, resulting in serious injury, or could be slower and stepwise. Focal motor seizures were recorded during sleep for two patients (5 and 9). The seizures produced mild unilateral facial jerking for 30 to 60 seconds followed by awakening(figure 5).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4. EEG recording for Patient 1. Interictal spike over the right frontocentral region followed by an ictal discharge of generalized spike-and-wave complexes that lasted 7 seconds accompanied by an atonic drop attack. During the ictal discharge, each slow wave was accompanied by cessation of muscle activity. The seizure was followed by slow-wave activity over the left hemisphere.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5. EEG recording for Patient 5. The child is asleep. Continuous generalized spike-and-wave complexes are replaced by focal ictal activity characterized by a build-up of rhythmic spikes that predominate over the left centroparietal region and are progressively replaced by rhythmic spike-and-wave complexes. The focal seizure activity ceases abruptly after 24 seconds. The patient has mild jerking of the right periororal muscles and awakens at the end of the seizure.

Three patients (1 through 3) had been treated with often-inappropriate antiepileptic drugs (phenobarbital, phenytoin, and carbamazepine) before ESES was diagnosed with poor results. Patients 1 and 2 also had undergone short-term trials of adrenocorticotropic hormone, obtaining only transient benefit. These three patients were treated with the late addition of valproate, ethosuximide, and benzodiazepines with unsatisfactory response. Immediate and stable cessation of atonic absences and ESES was observed for Patient 5 immediately after ethosuximide was added to valproate, for Patient 6 after lamotrigine was added to clobazam and vigabatrin, and for Patients 4 and 9 immediately after institution of valproate combined with clobazam. For Patient 8, who because of close EEG follow-up had the diagnosis of ESES confirmed a few months after it started, valproate in addition to phenobarbital led to immediate cessation of drop attacks. Patient 7, who had only rare partial motor seizures, was successfully treated with carbamazepine monotherapy. Anterior callosotomy, performed on Patient 1 at 6 years of age, did not modify clinical and electrographic findings and did not the course.

For eight patients (1 through 6, 8, and 9) ESES and generalized seizures ceased between the ages of 5 and 12 years (mean, 8 years). At the last visit they were 8 to 18 years of age (mean, 14.5 years) and were either seizure free or had very infrequent focal motor seizures during sleep. The patient with asymmetric ESES (Patient 7) at 23 years of age had been seizure free since 14 years of age. At the end of the study, three patients were free from use of antiepileptic drugs. No patient had definite cognitive deterioration after ESES compared with earlier evaluations. However, assessment was carried out with different methods and at different centers. Furthermore, for most patients, in-depth evaluation was precluded because of low cognitive level.

Discussion. The course of epilepsy among these nine patients followed a stereotyped pattern. ESES and atonic absences were preceded by earlier-onset focal EEG abnormalities and focal motor seizures. After an active period of variable duration, both absences and ESES disappeared. Rare partial motor seizures persisted for a while before likewise vanishing progressively. Although the long-term course of epilepsy was ultimately favorable, ESES duration was quite variable from one patient to another, and a trend toward longer ESES duration among patients with earlier onset and shorter ESES among patients with later onset could be detected (see figure 2). The effectiveness of treatment with antiepileptic drugs is difficult to establish because of the different combinations used. However, drugs active against spike-and-wave-related epilepsy, above all valproate and ethosuximide, appear to be effective, especially if administered promptly after onset of ESES. The seizure outcome was similar to that of patients with nonlesional ESES (2,3). The age range of patients with ESES and related seizure manifestations corresponded to that of patients with absence and those with rolandic epilepsy.¹³ This overlapping of age-related expression was paralleled by some clinical and electrographic similarities, such as the generalized spike-and-wave trait and focal motor seizures that resembled rolandic seizures (see figure 5).

Although behavioral disorders usually vanish with the disappearance of ESES, the cognitive impairment that may have been seen among most but not all patients^2,16,17 may persist.⁴ We were unable to appreciate any lasting cognitive impairment that could conclusively be attributed to ESES, but this may have been caused by limitations of the ascertainment procedures. For the following reasons it is unclear to what extent cognitive impairment detectable after the end of ESES should be attributed to the limitations of ascertainment. First, to our knowledge, in no studies have children been examined systematically before and after ESES. Second, examination of small children is usually less accurate and does not allow detection of the specific areas of neuropsychological dysfunction that are more clearly manifested at a later age. Third, assessment of severe brain damage among children is complex, and it is difficult to establish to what extent impairment is attributable to structural brain damage or to epilepsy and attendant EEG abnormalities.

Because the course of ESES appears to be stereo-typed and independent of brain damage and its cause, varying in duration but always ceasing before adolescence, age-related factors must play a role in its genesis. One possible factor is age-related SBS.^18,19 Kobayashi et al.^19,20 observed two patterns of intraburst small interhemispheric time difference variations in SBS. In Type I SBS, time differences remained unchanged throughout spike-and-wave bursts, whereas in Type II SBS time differences disappeared during the latter part of the discharges. Kobayashi et al. hypothesized two mechanisms of SBS. In the first, SBS likely occurs through the corpus callosum. In the second, after an initial diffusion of discharges through the corpus callosum, the corticothalamic system might play a role in generalization of the discharges in their final stage. Measurements of small time differences suggested that in ESES, SBS is Type I.^19,20 Analogous conclusions may be drawn from the results of the methohexital suppression test and electrical intracarotid amobarbital test among patients with continuous diffuse spike-and-wave activity.²¹ The interhemispheric time difference values we observed for two patients confirm that Type I SBS operates in polymicrogyria-related ESES, a finding analogous to the findings of Kobayashi et al.²⁰ among patients with nonlesional ESES. However, the ineffectiveness of callosotomy performed at 6 years of age on Patient 1 might indicate that other, not purely cortical mechanisms may play a role in the development of age-related spread of spike-and-wave activity. A slight effect of corpus callosum section on secondary bilaterally synchronous interictal EEG discharges also was observed by Spencer et al.,²² who hypothesized that SBS should not be imputed simply to callosal spread mechanisms.

ESES and associated atonic absences appear in the same age range as idiopathic absence epilepsy, the prototype of epilepsy with spike-and-wave-related cortical inhibition.²³ In this age range, idiopathic generalized epilepsy with convulsive seizures is rare,²⁴ confirming that epileptogenesis is subject to strongly age-related inhibitory influences. Age-related enhanced inhibition may contribute to the shift observed among our patients from initial disinhibitory ictal activity to subsequent hypersynchronous spike-and-wave discharges²⁵ that predominate during ESES. However, although redundant spike-and-wave discharges prevent disinhibitory seizure activity,²⁶ they also increase spike-and-wave-related inhibition of muscle tone,²⁷ producing atonic drop attacks.²⁸

Polymicrogyria is a term used to describe an excessive number of small convolutions separated by shallow and enlarged sulci.²⁹ In this malformation, horizontal cortical lamination often is spared²⁹ and the basic cytoarchitectonic abnormality is midcortical laminar necrosis predominating in Layer V.³⁰ Although these findings are microscopic, at macroscopic examination the abnormal convolutions prove to be packed and merged, producing an abnormal gyral pattern, which is recognizable with MRI.⁷ Our patients had no significant prenatal or family history that could provide evidence for a specific cause. The cortical abnormality could therefore have been caused either by an unrecognized intrauterine event between the 13th and 24th weeks of gestation³¹ or by a mutation in regionally expressed developmental genes.³²

Why can ESES be found among patients with polymicrogyria but not those with other epileptogenic cortical malformations, such as focal cortical dysplasia with abnormal proliferation of neurons and glia,³³ hemimegalencephaly,³⁴ schizencephaly,³⁵ or the agyria-pachygyria band heterotopia spectrum?^36,37 Selective damage to cortical Layer V³⁰ can result in an imbalance between the excitatory activity of intrinsically bursting pyramidal neurons, situated only in Layers IV and V³⁸ and consequently reduced in number, and the inhibitory activity of γ-aminobutyric acid-ergic interneurons, which distributed throughout the cortex would therefore be spared to some extent. In such circumstances, enhanced excitation arising from the malformed cortex would lead to an overwhelming inhibitory response, triggering hypersynchronous neuronal discharges.²⁶ Because horizontal neuronal lamination is spared,²⁹ spike-and-wave activity may then easily spread to become bilateral and synchronous.^21,39 Such a model would imply that the more severe the disruption of horizontal cortical lamination in the epileptogenic cortex, the lower are the changes of development of SBS. For example, in focal cortical dysplasia, in which disruption of cortical organization within the epileptogenic area is complete,^29,33 only focal electroclinical manifestations occur^7,37,40-42 despite high epileptogenicity.^42,43 The site and extent of polymicrogyria might also be relevant in producing ESES. To our knowledge it has never been described among patients with small areas of polymicrogyric cortex not involving the frontal lobes,^37,44 the participation of which in SBS mechanisms is well established.^45,46

Patients with malformations of cortical development and intractable seizures are candidates for resective surgery or callosotomy, depending on seizure type, location, and extent of abnormal cortex.^4-6,33,41 Some authors recommend early operation.⁴⁷ The role of resective surgery in epilepsy with ESES has not been addressed specifically, but it has been hypothesized that surgery may be effective when a focal abnormality is identified.⁴⁸ However, the good prognosis of epilepsy and the doubtful association with an acquired neuropsychological deficit should discourage early surgical procedures in the treatment of patients with ESES and polymicrogyria. Multiple subpial transections^21,49 with selective interruption of intracortical horizontal fibers might represent a rational option for patients with early onset ESES and evidence of incipient cognitive deterioration.

Although ESES is generally rare, not exceeding 0.5% in an unselected population of persons with epilepsy,⁵⁰ it is not uncommonly associated with polymicrogyria, reaching a frequency of 18% in our series. Sleep EEG should therefore be performed on patients with polymicrogyria, especially in the ESES age range and in the presence of drop attacks, low attention span, and hyperactivity.