JLE

Epileptic Disorders

MENU

Rolandic epilepsy: neuropsychology of the active epilepsy phase Volume 2, supplément 4, Supplément 1, Décembre 2000

There are few, but concordant studies indicating that children with typical rolandic epilepsy (RE) (simple partial sensory-motor seizures with centrotemporal spikes on EEG) as a group have more scholastic and neuropsychological problems as compared to controls, but the nature of these problems appears very variable [1-5].

It is difficult to tell from cross-sectional data, whether the deficits are due to the basic brain dysfunction responsible for the epilepsy or to other factors such as, for example, the effects of ongoing epileptic activity on developing cognitive functions.

Detailed neuropsychological studies of children with RE show that they have normal intelligence but often have limited weaknesses in various domains such as language, visuospatial abilities or isolated attentional problems. It is important to distinguish the cognitive deficits or learning problems which can be attributed to a localized epileptic dysfunction in the perisylvian region and involve "only" a specific instrumental capacity (i.e graphomotor skill, verbal expression, language comprehension, working memory) from those which are due to involvement of more general abilities such as sustained attention, processing efficiency, response inhibition and planning [6].

Two published, longitudinal studies [7, 8] and a personal one [9], suggest that transient cognitive disorders can occur in typical RE in the active phase of the disease. They seem to correlate with an increase in epileptic activity during this period. Attempts to correlate these deficits with the frequency of clinical seizures and with side of discharges (left versus right; unilateral versus bilateral foci) have given conflicting results. Short-term transient cognitive impairments (TCI) in direct time correlation with the density of paroxysmal EEG discharges have also been shown in a few children with RE, but it is surprising that these findings have not been more widely documented [10, 11].

Long-lasting, fluctuating, but fully reversible oromotor deficits have been described in longitudinal studies of individual cases of typical RE [12] which correlate with the abundance of paroxysmal activity and bilateralization of discharges in the "interictal phase" [13].

We recently studied an 11 year-old child with RE who had an isolated, prolonged, reversible disturbance of graphomotor skills documented by quantitative digitalized analysis of various components of the graphomotor act. [9]. This case was a good example showing that a persistent, extremely discrete, focal cortical epileptic dysfunction correlating closely with the paroxysmal activity on the EEG in the sylvian region, can occur in some cases of RE. The next question is whether the often subtle and transient cognitive impairments described in RE are due to the same epileptic dysfunction, but in other locations. It can be supposed, for example that if the main epileptic focus is situated in or spreads more posteriorly to the perisylvian region (for example the superior temporal area), difficulties with the decoding of sounds can occur, leading to auditory agnosia, that is the so-called Landau-Kleffner syndrome. We recently saw a child who after recovering from LKS, could tell us about his unnoticed sleep-related sylvian seizures! Evolution from typical RE to Landau-Kleffner syndrome has also been repeatedly described [14, 15].

The direct role of the epileptic discharges in causing these transient cognitive deficits in RE could be confirmed if it could be shown that effectively suppressing them with antiepileptic drugs improves the cognitive function concerned. Unfortunately, antiepileptic drugs in RE, with the possible exception of Sulthiame can sometimes markedly aggravate the situation (especially Carbamazepine) and mainly increase the frequency of epileptic discharges and produce temporary cognitive symptoms [15]. Such cases illustrate, in a dramatic fashion, what can also rarely happen spontaneously in very active cases of RE. At the moment, only Sulthiame appears to be able to both control seizures and suppress EEG discharges in some cases of RE (and possibly also improve cognitive function), but precise data on the latter are very limited.

Another major problem is posed by young children who present with developmental retardation or older children who have a learning disability and who on EEG are found to have focal sharp waves in the rolandic area and/or in other locations, whether or not they had had or have clinical seizures. Are the EEG abnormalities only non-specific markers of an underlying encephalopathy (genetic) or do they play a more direct role in causing clinical problems [16]? Knowing that children with RE have normal (and sometimes superior) cognitive abilities or, as discussed before, have only transient cognitive disturbances, one cannot a priori blame the basic brain dysfunction causing the RE to explain their cognitive disturbances. One can speculate that early and prolonged focal epileptic activity affecting neuronal networks in the process of specialization at a time of active learning could lead to permanent "damage" (i.e. abnormal connectivity) of the involved structures. Thus, the same child could have developemental retardation for these reasons but at a later age show stagnation or regression during a new phase of activity of the epilepsy as a more direct consequence of ongoing epileptic discharges. It should also be remembered that when a child is diagnosed as having RE, he may have had paroxysmal EEG discharges for a long time and possibly also, previous unrecognized seizures (typically during sleep). This means that the epileptic disorder could have interfered with developing cognitive functions before the diagnosis was made, i.e. during the "active" epilepsy phase.

Within the spectrum of "genetically determined partial epilepsies with focal sharp waves" [16], it can be hypothesized that children with typical RE (i.e. with sylvian seizures) are those at least risk of cognitive dysfunction because the epileptic disorder is mainly located in a zone with early maturation and no direct involvement in cognition. The differences in cognitive profiles and abilities seen in children with an apparently equally benign clinical epilepsy could be explained by the location(s) and spread of the focal epileptic activity (unilaterally and to the other hemisphere), interindividual variability in cerebral organization, age at onset of the epilepsy, its duration and severity. Some of these factors could explain the familial observations of siblings who all have RE, but quite different congitive problems.

In summary, it has now been shown that transient cognitive disturbances do occur in some children during the course of RE and that these are probably closely related to the activity of the epilepsy. They can be subtle, but possibly persistent or recurrent and sufficient to interfere with progress at school and, if unrecognized, lead to inapproriate expectations and demands and to secondary behavioral problems. More longitudinal studies are needed to define the range, type and evolution of neuropsychological disturbances in RE and its variants, looking both at specific deficits and at other cognitve capacities subsumed under the term "executive functions". The same applies to children with developmental retardation or learning disabilities who are found to have rolandic discharges but have no clinical epilepsy. It remains to be shown if an antiepileptic therapy which effectively suppresses the EEG discharges can improve cognitive efficiency in children with sectorial deficits, or accelerate the dynamics of learning. Detailed longitudinal studies of children with RE in very active stages of the syndrome and in whom a specific, quantifiable aspect of cognitive dysfunction can be closely monitored are likely to be the most informative at this stage.

Discussion at the meeting

It was pointed out that the location of the paroxysmal activity in RE is much more variable than usually acknowledged, and can change over time in the same hemisphere and from one side to the other [17]. Consequently, one must expect a great variability of neuropsychological problems, if the discharges in themselves do indeed play a direct, negative role. The best way of proving that EEG discharges do have a negative role in cognition is to demonstrate transient cognitive impairment (TCI) during periods with, as compared to periods without discharges, but data on this are still limited. The Strasbourg group (Metz-Lutz, Hirsch, et al.) has an ongoing study involving this topic and we should have their results soon. Another way of finding out the exact function of the zone involved in the discharge and its importance for cognitive processes is to document an electroclinical seizure and see if an ictal or postictal impairment is present, but this is very exceptional in RE and its variants.

If it is accepted that it is the EEG discharges and not the clinical seizures which are relevant from the neuropsychological point of view, the possible efficacy of drug treatment must be evaluated for its effect on the EEG. There are limited, but suggestive data indicating that Sulthiame can suppress EEG discharges in RE in a certain percentage of cases and further studies on this point are necessary.

The importance of developing adequate neuropsychological tools to document and in particular follow the evolution of cognitive deficits which are not massive is emphasized. However, one can not a priori know in a given child which deficit to monitor as a priority at the onset and one must tailor the evaluation to the individual situation.