- Top of page
Summary: Purpose: Questioning the presence of any possible prognostic predictors, this study includes a long-term follow-up of clinical and EEG characteristics of 16 patients with idiopathic partial epilepsy (IPE) who subsequently developed epilepsy with electrical status epilepticus during slow sleep (ESES) spectrum disorders.
Methods: Epilepsy, cognitive and behavioral parameters, and waking and non–rapid eye movement (NREM) EEG data were evaluated and scored for initial stage (i.e., IPE stage), preESES, ESES, and ESES remission periods, individually, on a chronologic basis. Data from 25 healthy subjects who had had IPE at the appropriate ages served for comparison with the patients' data during the IPE stage.
Results: Results revealed a higher incidence in seizure frequency and variability in the ESES group and a resistance to a single antiepileptic drug (AED), as compared with controls, during the IPE stage. Mean age at onset of epilepsy was younger in the ESES group versus controls (5.5 and 7.3 years, respectively). At least one of the premonitory clinical features for development of ESES [an increase in the seizure frequency and/or addition of new types of seizures (93%), appearance of cognitive and/or behavioural changes (81.2%), or a progression in EEG abnormalities (66%)] was present in all patients. Epilepsy remitted in patients within the ESES spectrum at a similar age as in controls in 81.2%, as ESES findings in the EEG disappeared by age 13 years in 94%. Seizure prognosis proved to be the most favorable among the questioned parameters.
Conclusions: An increase in seizure frequency or development of new seizure types, a deviance in behavior or decrease in cognitive performance, or a spreading tendency of the previously focal abnormalities in control EEGs may be premonitory features of a developing ESES and necessitate close follow-ups with sleep EEGs in children with IPEs.
- Top of page
Starting with the introduction of the concept of continuous SW activity during NREM, in 1971 (6), growing interest in the nosology and syndrome-related characteristics of this phenomenon, its impact on neurologic and behavioral functions, and its pathophysiological aspects has been accumulated. The term ESES, used by Tassinari et al. (1), was proposed to be modified as CSWS in 1989 by the ILAE. The reason for the new proposal was that epilepsy is a clinical, not an electrographic, phenomenon, and that this activity is not always associated with epilepsy. However, although with some nuances, both terms are currently used. Leaving reserve for further nosologic considerations, the term ESES is used for only the EEG changes in this text. “ESES spectrum disorders” is used for all conditions associated with ESES in the EEG, including LKS. Although association of ESES with lesion-related epilepsies is a well-known condition (8), such cases were not included in this study because of the possibility that they might possess some varying individual characteristics in relation to the localizations and nature of the brain pathologies. Besides, the target of the study was the IPE stage; and the question was whether any signs suggesting a tendency to progress into an ESES spectrum disorder could exist early during their course. Some such changes, in fact, when compared with typical cases of IPE stage in the study, were present in the patient group. Those were a higher incidence in seizure frequency and in the variability of seizure types and resistance to a single AED (75%:16%, in the study/IPE groups). An EEG finding present in 25% of the patients but in none of the controls was a focal slowing in the EEG at the site of the predominant epileptogenic foci in both waking and NREM periods, starting from the very early stages of epilepsy. This finding was continuous rather than paroxysmal and could be easily misinterpreted as suggesting a destructive lesion rather than a functional change, especially in some epochs without any visible spikes interfering with the slow activity. Such a finding may be related to the severity and the stability of the neuronal firing, because it persisted in all the EEGs of those patients throughout the later stages of the disease and always in the same region as before. This slow activity served for the origin of bilateral discharges pertaining to the ESES stage (Fig. 1) later.
No significant relation between the bilaterality of epileptogenic foci and progression into ESES stage was found in this study. Those findings suggest that the possibility of an evolution into ESES may be considered in the IPE stage when a continuous slow-wave activity is present in the predominant epileptogenic locus in the EEG, when there are multiple seizure types, and when the seizures are resistant to conventional AEDs in patients with IPE.
Age at the onset of the first afebrile seizure in the study group was younger than the respective age in the controls (5.5 vs. 7.3 years), which was a finding with statistical significance.
Family history of epilepsy (including febrile seizures) was mentioned in ∼15% of patients with ESES (8). In our study, the patient population with epileptic members in the family was found to be ∼3 times higher in the study group (37.5%) than in the controls (12%). This difference may suggest an additional genetic tendency for evolution of ESES, in the setting of IPE, which is already known as a condition with genetic nature. No other family member with epilepsy, conversely, was diagnosed or suspected by us according to the patients' reports, to have any ESES spectrum disorder in our study group. This topic may require bigger patient populations.
Major head trauma in the history was remarkably high (43%) in the patient group. Only events with serious injury leading to necessity to be cared in emergency units but resulting in no clinical and radiologic abnormalities were included in this parameter. This surprisingly high incidence in the patient group may lead to questioning a past cranial trauma as a predisposing factor for ESES, in patients with IPE.
EEG changes starting from the IPE stage throughout evolution into the ESES stage very likely takes some time, possibly with interindividual variances. Precise duration of this process is almost always impossible to determine in a given patient, unless found at the time of routine follow-up recordings. The inclusion of an arbitrary period of 6 months as the pre-ESES stage was decided on the basis of such a necessity in this study. This stage was the time in the middle of two consecutive EEGs (because patients with IPE have sleep-waking EEGs taken at least once a year in our routine) when the latest EEG showed ESES and the previous one had findings typical for the IPE stage. Changes in seizure characteristics were the most consistent clinical stigmata during pre-ESES or at the onset of ESES stages, with an incidence of 93.5%. Increase in seizure frequency and addition of new seizures types at the time of ESES in epilepsy patients have been reported previously (9). New types of seizures developed in 13 (81.3%) patients in our study group, and six (38%) had an increase in the preexisting partial seizures. Seizures associated with ESES have been documented as absences, absence status, head drops, and falling attacks (8). Other paroxysmal phenomena less frequently seen were negative myoclonias (8,10), tonic contraction of facial muscles, myoclonic absences, epileptic spasms, and generalized convulsions (8,11). Tonic seizures were reported to be absent in those patients (8,12). ESES-associated seizures in our patient group were absences, head drops, drop attacks, rare myoclonic absences, facial myoclonias, or partial versive–tonic seizures during sleep and leading to awakening sometimes, and also an increase in the tendency for secondary generalization of the rolandic type of partial seizures.
A benign course and good response to treatment in seizures in ESES spectrum disorders have been reported previously (13). Independent of the severity of epilepsy, seizures are known to disappear by the termination of ESES (8). A benign course of epilepsy was a similar finding in our study, and the seizures disappeared almost at the same age as in the control subjects with IPE, only. The concordance rate for the remission of epilepsy and ESES in the EEG was also very high (84.6%), and in the majority of this group, seizures disappeared some months earlier. Epilepsy was active during the latest control in only two cases, and ESES was not totally abolished, but transformed to hemi-ESES in one.
Behavioral and/or cognitive involvement also was prominent (81.2%), but a less frequently complained of phenomenon than seizures, during the stages of pre-ESES and ESES in our patient group. As the feature with most serious consequences in the ESES spectrum disorders, it is reported to exist in varying degrees in nearly all cases (8,14–17). It may have a prompt onset in rare conditions, but a progressive course is the usual fashion. Very rarely, it may precede epilepsy. Mild to severe disturbances in cognitive and/or behavioral performances such as deterioration in language, temporospatial disorientation, distractibility, hyperactivity, aggressiveness, impulsivity, reduced attention span, and communication problems, rarely up to the degree of psychotic features, may be encountered (18,19). Major complaints pertaining to 10 (62.5%) patients in our study were speech problems (primarily, a decrease in the speed and fluency of the spoken language), hyperactivity, attention deficit, easy frustration, emotional outbursts, and a decline in memory and in school performance. A single patient had nocturnal enuresis by the onset of ESES, and this symptom showed a direct relation with the fluctuations of ESES in her EEGs. Severe cognitive decline was persistent in three cases. The single patient with LKS was resistant to treatment with major AEDs in various combinations and hydrocortisone. At his latest control, epilepsy and ESES in his EEG were present for 11.5 and 4 years, respectively. The remaining two patients had low IQ scores after 7 and 12 months, each, later than their last EEGs, which were free of ESES. Cognitive and neuropsychological impairments are known to contribute to ESES spectrum disorders as the most resistant feature to treatment (13). Therefore the significance of ESES in the EEG, which is always an age-limited phenomenon, is solely related to the cognitive and behavioral consequences that may lead to permanent deficits. Mental deterioration has been related to the duration of ESES in previous reports. A long-duration (>2 years) of ESES was found to be the major factor predicting a bad prognosis in cognitive and behavioral modalities (20,21). However, two of three cases with severe cognitive impairment in our study had ESES with durations of 4 months and 1 year. Conversely, two other patients with ESES with 2 and 3 years' duration had no more major cognitive or behavioral problems at the ESES remission period. This discrepancy suggests the possibility that variables other than the duration of ESES also may be responsible for the long-term cognitive consequences, at least in some cases.
EEG as a parameter signaling the development of ESES was in the third level after seizures and cognitive and behavioral changes. To our knowledge, EEG changes in the IPE stage suggesting an evolution into ESES have not been discussed before. Among nine patients with EEG recordings during the pre-ESES period, five (55.6%) showed abnormalities suggesting a progression into ESES. They all had secondary bilateralization of the previously focal epileptogenic activity. This picture was called near-ESES if the bilateral discharges covered ≥50% of NREM stage. Asymmetry in bilateral discharges of near-ESES was much more prominent than in those of ESES. The pre-ESES stage, in our opinion, may be a critical time to decide on more efficient treatment procedures to prevent possible future consequences of ESES. Similar suggestions regarding early interventions before the development of full SW index (i.e., 85%) have been proposed previously by others (4,22,23). The question whether a SW index of 85%/NREM be needed to recognize a condition as ESES may be subject to further considerations. Presently, it seems that when some clinical and/or EEG features in children with partial epilepsies predict a development toward ESES, early measures may provide more benefit than may vigorous treatment strategies after months or years of surviving ESES. Our observations on some newer patients not included in this study suggest that they are more responsive to modifications in the AED regimens during the pre-ESES period than in the ESES stage, the time when corticosteroids are frequently needed.