Epileptic Disorders


Cognitive impairment and behavioral disorders in Encephalopathy related to Status Epilepticus during slow Sleep: diagnostic assessment and outcome Volume 21, supplement 1, June 2019


Studies describing quantitative aspects of epileptiform abnormalities on EEG are overrepresented in the literature, whereas those evaluating evolution of cognition and behavior are undervalued. The large variation in the design of the studies, the tests used, the age ranges evaluated and the variability in outcome measures preclude any reasonable comparison. Consequently, studies attributing improvement exclusively to epileptiform discharges should be interpreted with caution (Sanchez Fernandez et al., 2015). Probably the only relatively strong data available suggests that severity and duration of initial regression are the most important risk factors of cognitive impairment in the long term (Van Bogaert, 2013).

Outcome and predictors of outcome

Electrical status epilepticus of slow sleep is by definition what could be classified as a true epileptic encephalopathy. It is believed that cognitive and behavioural difficulties are the direct result of the underlying epileptic activity and any targeted treatments. However, we need to review the evidence on outcome; specifically, the relationship with resolution of the epileptiform activity, the influence of aetiology and indeed the role of treatment. Further, do we have evidence that we can actually influence outcome?

Like other epilepsy syndromes, Encephalopathy related to Status Epilepticus during slow Sleep (ESES) is described on the basis of electroclinical features but may have many aetiologies - specifically lesional and non-lesional. Presentation of cognitive deficit has been suggested to be related to the geographical prominence of spike wave activity - in Landau-Kleffner, this is seen with temporal prominence, whereas more global problems are associated with a frontocentral focus. However, there does not appear to be any evidence of a predictor of outcome dependent on localisation of epileptiform discharges.

Many have demonstrated ESES to be an age-related phenomenon; illustrated by both lesional and nonlesional cases. It has been well demonstrated that children with hemipolymicrogyria may present with focal seizures but can subsequently evolve with development of atonic attacks associated with the finding of electrical status epilepticus of slow sleep. In longitudinal studies, such attacks resolve along with the ESES (Guerrini et al., 1998; Caraballo et al., 2013) - however, there are no cognitive or behaviour measures included in the studies reported, so the overall impact on this aspect is unknown. It is now known that such phenomena occur in “acquired” cases or those with developmental pathology and may occur with unilateral or bilateral disease. In the case of unilateral structural abnormalities, surgery has been demonstrated to resolve the ESES but definitive improved cognitive outcome, although suspected as being likely, has not been proven (Loddenkemper et al., 2009).

There are many methodological problems with long-term studies reported to date within the literature. Not least, series often include both lesional and non-esional cases, with very small numbers owing to the rarity of the condition. Further, there is significant heterogeneity in the treatment utilised even in a single centre with no structured protocols, which applies also to reporting of outcome and time period (van den Munckhof et al., 2015). The natural history is such that electrical status epilepticus has a good prognosis and resolves with age - it is difficult to know therefore the impact of any intervention over time. This is well illustrated by the surgical Landau Kleffner series of Morrell; the initial series of 14/54 children who underwent surgery (multiple subpial transection) after full assessment although 6 had normal speech postoperatively and five improved (Morrell et al., 1995; Grote et al., 1999); a later follow-up study demonstrated that the extent of improvement was related to time from surgery. Indeed, a recently reported study showed no difference in outcome as to whether surgery was undertaken or not (Downes et al., 2015).

Praline et al., (2003) reported on seven adults who had experienced CSWS or Landau Kleffner Syndrome in childhood. At the time of review, they were aged 16-26 years; five had CSWS syndrome and two Landau Kleffner. They confirmed the epilepsy to have a good prognosis in the long-term; only one had continuing active epilepsy. However, 3/5 who had CSWS in childhood remained significantly cognitively impaired (Praline et al., 2003).

Possible predictors of outcome include age at onset, duration of ESES, response to treatment, aetiology and predominant location of the interictal focus. An older age at onset and shorter duration of ESES are correlated with a better cognitive outcome although this does not appear to be absolute (Scholtes et al., 2005). This aside, in a further study of 10 patients with non-lesional ESES followed for 15.6 years, patients with prolonged global intellectual regression had the worst outcome whereas those with more specific and short-lived deficits showed the best recovery (Seegmuller et al., 2012). Further, there was no correlation between outcome an age at onset or age at cessation of ESES, but the three most severely affected individuals had the longest duration of ESES. Kramer et al. reported on 30 patients from four clinics who had been determined to have ESES; 20 had an associated intellectual regression, having previously been normal, 3 of whom did not respond to treatment. They found a significant correlation between residual cognitive deficits and the total ESES period (Kramer et al., 2009). However, careful longitudinal review has illustrated that although such correlations are evident, prognosis of an individual is highly variable.

Pera et al. (2013) reviewed 25 children with a mean follow-up of 13.5 years; they suggested five patterns of clinical course within this otherwise small series with not necessarily a clear correlation between clinical course and EEG response to treatment. Group 1 demonstrated the classical regression with EEG abnormality, and improvement with EEG response to treatment; Group 2 had predominantly motor deficit with minimal cognitive change; Group 3 had cognitive deficits that persisted despite improvement in the EEG; Group 4 had associated cerebral lesions, with little change in cognition over time; and Group 5 included two patients with progressive deterioration based on neuropsychological tests without temporal correlation with ESES duration (less than seven months), and no association with clinical and electroencephalographic relapses. In total, 44% of the whole group had permanent cognitive impairment in the long term (Pera et al., 2013).

Diagnostic assessment of cognition and behavior in ESES

Because ESES is a highly heterogeneous and a relatively rare entity (even when atypical forms are included), multicentric studies usually fail in reaching a statistically significant number of participants; long-term follow-up is usually not possible, often because of lack of funding. However, improving our knowledge on outcome requires a method to measure it. Lessons from the past should lead the epileptological community to a more pragmatic approach and avoidance of circular statements; instead of indefinitely concluding that “controlled multicentric studies are needed...” we could agree upon a basic diagnostic assessment protocol to be followed by all major paediatric epilepsy centers.

A common-to-all, basic diagnostic assessment protocol

All children with ESES, independent of underlying etiology, should from onset be referred to a pediatric neurology center specialized in epilepsy. A commonly agreed, user-friendly software could be used to collect relevant data (that one day, if needed, could be merged into a common database) to include:

  • main neurological and somatic examination findings at onset;
  • family history, including familial cognitive and behavioral problems;
  • a minimum 24-hour video-EEG, preferably before administration of any treatment or within the first 8 weeks; agreement could be made upon a set of minimum technical requirements for polygraphic recordings (table 1);
  • the realization of a predefined, commonly agreed, core battery of neurocognitive tests and evaluation scales (see below) covering the most crucial parameters, sufficiently straightforward to be reproduced in different clinical settings;
  • high-resolution MRI performed according to standard epilepsy protocols (Gaillard et al., 2009);
  • genetic evaluation (to be enriched on the basis of progress made in the field).

Practices that may need a consensual attitude

  • An agreement on the first three AEDs to be used, including dose per kg, duration of each trial and minimal efficacy criteria. Lack of results from controlled trials should oblige us all to follow a consensual approach, referring to existing open studies, rather than a dogmatic position based on personal beliefs and impressions;
  • An agreement upon a core battery of neurocognitive tests, to be repeated at regular intervals (once a year?). Depending on the ages and abilities of the patients, tests should cover the major domains of cognition (intelligence, language, memory, attention, visuospatial functions, executive functions);
  • An agreement upon a basic scale evaluating AED side effects (Morita et al., 2012) to be repeated at regular intervals, particularly after a change in AED therapy; This could be systematically coupled to a sensitive and time-efficient screening tool for attention and executive functions, such as EpiTrack Junior (Kadish et al., 2013).

Practices that probably can only be decided arbitrarily!

An agreement upon intervals for 24H VEEG. Variability in current practices complicates comparison of findings across various studies, and limits the possibility of generating evidence-based guidelines for patient follow-up (Jehi et al., 2015). The reason that such an agreement between centres is hard to obtain probably reflects a disagreement upon primary efficacy criteria. One, rather classical, approach would be the “percentage of amelioration of the sleep-EEG ESES pattern”, reproducing what we are currently applying in the majority of centres.

However, another option, probably better serving our primary aim (see above) could be to use the results from core battery neuropsychological tests as an efficacy criterion. For example, at least for a period of 12 months, the AED prescribed would be replaced only in case of signs of cognitive deterioration (validated by a minimum of tests) or in case of clinically identifiable epileptic seizures. EEG variations would still be recorded without automatically leading to modifications of treatment.

If an agreement on the second option suggested above proves impossible to attend, we could at least agree that each individual centre would accept to systematically determine its treatment attitudes on the basis of the same criterion: either EEG variations or the results of the neuropsychological follow-up. Such an approach would at least allow the constitution of two comparable groups, in terms of global evolution in the long-run, of children treated for ESES.

In summary, the fluctuating clinical and EEG courses of ESES complicate the diagnostic process and evaluation of treatment. We need to validate a feasible screening method to be used by clinicians of all major paediatric epilepsy centres. It should be in accordance with the main lines of current clinical practices and preferably under the authority of a well-defined endorsed Task Force such as within the ILAE. The policy of publishing small series of patients per centre has clearly failed in reaching meaningful clinical conclusions. The natural course of the disorder(s) remains poorly known and multiple underlying mechanisms link the EEG patterns to developmental outcomes. A consensus on a homogeneous approach should be considered a priority.


None of the authors have any conflict of interest to declare.