JLE

Epileptic Disorders

MENU

Cognitive function in adolescents and young adults in complete remission from benign childhood epilepsy with centro-temporal spikes Volume 3, issue 4, December 2001

Figures

See all figures

Epilepsy is often associated with cognitive disorders, particularly in children [1-6]. It constitutes a model for observing the possibilities and limits of cerebral plasticity. The relationship between epilepsy and the cognitive functions is multifactorial [7]. Benign childhood epilepsy with centrotemporal spikes (BECTS) also called benign rolandic epilepsy was described for the first time by Nayrac and Beaussart [8]. BECTS is classified according to the International Classification of Seizures and Epilepsies [9] under idiopathic localization-related epilepsy, which include age-related childhood epilepsies, without anatomic lesions and are subject to spontaneous remission. BECTS is considered benign with regard to seizure outcome and control with antiepileptic drugs [10]. BECTS is the most common form of epilepsy in childhood, accounting for 14 to 17% of children with epilepsy under 16 years of age [11-14]. Classically without anatomic lesions, modern neuroimaging procedures may reveal abnormalities such as hippocampal atrophy [15] or hippocampal asymmetries and high signal intensities on T2-weighted images beneath the cortex-white matter junction, in the frontal and temporal lobes [16], in rare patients. This epilepsy constitutes a particularly interesting opportunity for evaluating a homogeneous group of epileptic patients, and for differentiating the neuropsychological consequences of a cerebral lesion and its subsequent treatment from the effects directly related to the epileptic discharge itself. The prognosis of BECTS is good since seizures disappear between the ages of 15-18 years [10, 17-22] except for a few rare exceptions [23]. The meta-analysis of Bouma et al. [24] concluded that 99.8% are in remission by18 years of age. Normal neurological and cognitive findings have been considered a prerequisite for the diagnosis of BECTS. Nevertheless, neuropsychological problems have been described [25, 27] and the long-term outcome has not been fully evaluated [28]. Difficulties in maintaining attention [29], in language [30-32] and visuo-motor disorders [23] have been reported. Concerning long term outcome, d'Allessandro et al. [28, 30] reported normalization of attention, visuo-motor performances and semantic fluency after remission of the epilepsy.

Explaining the neuropsychological disorders in BECTS is not easy. Transient cognitive impairment (TCI) is a well identified clinical phenomenon resulting in discrete neuropsychological deficits during infra-clinical epileptic discharges [33], which are responsible for impaired psycho-social functioning in daily life [1, 34-36]. Binnie [33, 37] studied children with rolandic spikes with or without the clinical typical pattern of BECTS and reported a high discharge rate accompanied by TCI (errors during short-term spatial memory tests), suggesting a direct link between epileptic activity and cognitive functions. Another approach to demonstrate a possible relationship between paroxysmal anomalies and neuropsychological disorders in BECTS is to study the relation between lateralized hemispheric function such as language, and the topography of the paroxysmal anomalies. Weglage et al. [38] reported that verbal IQs were the same in BECTS patients and in controls. Staden et al. [32] found disorders in auditory discrimination, auditory-verbal learning, expressive grammar, reading and spelling in a population of BECTS children. Piccirilli et al. [39] tested the lateralization of language functions in BECTS using a time-sharing paradigm [40] in right-handed children who had a right or left rolandic focus by comparing them to a paired control group. The performances of children with a right rolandic focus were similar to those found in controls, thus suggesting left hemispheric lateralization in language, whereas children with a left focus demonstrated a different pattern, implying a bilateral representation of language functions. It was suggested that focal activity may alter cerebral mechanisms underlying cognitive activity.

In this study, we evaluated the neuropsychological outcomes of adolescents and young adults in remission from BECTS by searching for persistent neuropsychological disorders and by investigating a possible link between the initial epileptogenic focus and the functional cerebral organization of language. We compared the performances of a group of patients in remission from BECTS who had an initial right or left focus to a control group and to a group of adults with complete recovery from childhood absence epilepsy (CAE), defined by the ILAE [9] as a generalized idiopathic epilepsy occurring in school-aged children. BECTS and CAE are idiopathic epilepsies with common features: appearance in normally developed children, family history suggesting hereditary transmission, good seizure control and remission in adolescence [13, 41-44]. We used classical tests to measure lateral hemispheric function: dichotic listening task, dual-task procedure. Our hypothesis was that cerebral organization patterns would be different according to the initial right or left topography of the rolandic focus.

Patients and methods

Population

Adolescents and young adults in total remission from BECTS or CAE were recruited using data obtained from the specific disease codes from the Neuropaediatric and Paediatric units and the EEG laboratories of both the Tours and Reims Neuropaediatric Departments. All of the enrolled patients had been followed in either the neuropaediatric or paediatric units of Tours or Reims Hospital. In each case, the diagnosis of BECTS or CAE had been made by a child neurologist using classical criteria [9]. Each patient was contacted by letter and informed of the aims of the study. The study protocol was approved by the ethics committee of Tours. Written informed consent was obtained from all patients and from the parents of underage children. None of the children had a personal history of brain damage and in each case, the neurological examination was within normal limits.

The BECTS group was composed of adults in complete remission from BECTS, fulfilling the ILAE criteria [9]: onset between 3 and 13 years-old, normal psychomotor development, seizure activity related to sleep, brief, partial hemifacial motor or generalized nocturnal seizures, with a typical EEG pattern (rolandic paroxysms defined as spikes or spike-waves simultaneously recorded over the posterior frontal, central, mid-temporal and occasionally the parietal regions with a maximum amplitude over the centro-temporal region) [10, 45]. The criterion for remission was no seizure activity or abnormal EEG tracing for at least one year. Fourteen patients were lost to follow-up, fourteen refused to participate in the study from the onset and six failed to show up for their appointment. The final sample was thus comprised of twenty three patients (men: 14, women: 9, mean age: 20 years 2 months [5, 16-28]), 18 of whom were right-handed and 5 who were left-handed [46]. The initial epileptic focus was determined using an awake and sleeping (nap) EEG, combined with polysomnography in the majority of cases, at the age of the first seizure. The side of the initial focus was established from EEG reports when the EEG showed that it was localized to one cerebral hemisphere without diffusion to the contralateral hemisphere or secondary generalization. Data from patients history are summarized in tables I and II.

The CAE group comprised adults in remission from benign partial epilepsy as defined by the ILAE criteria [9]: onset between 3 and 8 years, absence of prior brain damage, multiple daily absences associated with bilateral, synchronous, symmetrical 3 c/s spike discharges and an otherwise normal tracing. The patients selected were considered to be in complete remission, were receiving no treatment and had had normal EEGs over at least the previous year. Fifteen patients were lost to follow-up, eight refused to participate in the study and two were not considered to be in complete remission. Consequently, the final group contained ten patients (men: 3, women: 7, mean age: 21 years [8, 16-25]). Data from patients histories are summarized in table III.

The control group consisted of thirty three patients, (mean age: 20 years 2 months [16-27,5]) with no significant neurological history and who were paired according to age (using 5 year intervals), sex, handedness, educational and social level. Subjects who had a family history of epilepsy were deliberately excluded since it has been shown that healthy siblings of epileptics may have benign focal spike-waves and could therefore have potential infra-clinical seizure activity [47]. The control volunteers were recruited by letter or the placement of adverts.

Materials and procedures

Evaluation of global cognitive functions

Overall intellectual capacity was evaluated using the revised Wechsler Intelligence Scale for Adults (WAIS r) [48]. Verbal and non-verbal memory performance was evaluated with Wechsler's Memory Scale (WMS r) [49]. Language functions were tested by using the Binois Pichot vocabulary test [50] and a formal lexical and semantic evocation test [51].

Executive functions were studied by the classic tests: Trail-Making Test [52], Stroop [53] and the modified Wisconsin Card Sorting Test (WCST) [54]. They were scored according to the number of tries, the number of correct answers, the number of errors (perseverative and non-perseverative).

The two epileptic groups and the control group underwent all of these tests except for WAIS r and WMS R which were not proposed to the control group because of sufficient normative data.

Evaluation of lateralized hemispheric functions

We used a word dichotic listening task [55] and a dual-task procedure [39] evaluated in a previous study [56]. The recording was composed of fifty four items: word pairs and short sentences [55]. Correct answer scores were calculated for each ear (right ear: RE and left ear: LE), and a laterality index was computed: LI=RE-LE/RE+LE [57-60].

The dual-task procedure introduced by Kinsbourne and Cook [40] in order to evaluate the hemispheric lateralization of language, is based on the notion of "functional cerebral distance": two independent activities, which depend on neighbouring cerebral structures, that are simultaneously performed. They interfere with one another and lead to poorer performance in one of the two tasks [61, 62]. In right-handed subjects, speech interferes more with right-hand movements than left-hand tasks [40, 62]. The subject was sitting facing a MacIntosh® computer screen and asked to tap with the index and middle finger of one hand [39] as fast as possible on the B and N keys of the keyboard for 10 s sequences. The time-sharing verbal task consisted of repeating three animal names as many times as possible for the duration of the sequence [56]. This produced four different circumstances: isolated right-hand activity (RH), isolated left-hand activity (LH), right-hand activity associated with a verbal time-sharing task (RV) and finally, left-hand activity associated with a verbal time-sharing task (LV). The performance of each subject during the time-sharing task was converted to percentage change from baseline (with baseline representing performance in the hand showing no concurrent activity) according to the formula [39, 61]: Right percentage change (Right %) = RH-VH/RH X 100.

Statistical analysis

In each group, we calculated the medians and the ranges of all scores obtained. Comparisons between groups were accomplished using the Wilcoxon and Kruskal Wallis non-parametric tests. The SAS computer programme (SAS Institute, Cary, NC, USA) was used for the statistical analysis.

Results

Evaluation of global cognitive functions

The scores obtained during the global cognitive tests are shown in table IV. The Global IQ (GIQ) was normal in the two adult groups in complete remission, although it was lower in the CAE group [92] than in the group of BECTS patients in complete remission [100] (P<0.05). Verbal IQ (VIQ) was the same in the two groups. The Performance IQ (PIQ) was lower [85] in the group of CAE patients in complete remission (PIQ CAE=85; PIQ BECTS =101, P = 0.02). The Picture Arrangement and Block-design subtest scores were significantly lower in CAE as opposed to BECTS patients (respectively, P<0.01 and P<0.05). Verbal and non-verbal memory performance was the same in both groups. No significant differences in the Binois-Pichot vocabulary test or the various tasks used to evaluate the executive functions were found (table V).

Evaluation of lateralized hemispheric functions

Only the medians and ranges of right-handed patients are reported (tables VI and VII). There was no significant difference among the three groups. During the dual-task procedure, the Right %-Left % index also showed no significant difference among the three groups. Right % was different in the three populations without significance (P<0.06).

We found a significant difference between BECTS patients and the CAE patients in the Right % change respectively 5.5 and 11.6 (P = 0.04). The Right % change was less in BECTS patients, suggesting that the interference with right hand activity, while concurrently performing a verbal task, had less effect on right hand activity. In contrast, the Right % change was not significantly different between CAE patients (Right % = 11.6) in remission and controls (Right % = 9.5 (table VII). Right % change in BECTS patients was lower (Right % = 5.5) than controls (Right % = 9.5) but the difference was not significant (P = 0.08).

We evaluated the Right and Left % change in each population as a function of the initial seizure focus (right or left) in the sub-group of the eighteen, right-handed BECTS patients (nine of whom who had an initial right focus and six an initial left focus). The three patients who had an initial bilateral focus were excluded from the analysis. The resulting comparison of BECTS patients, according to the topography of the initial focus, showed that the Right % change pattern was not significantly different (P<0.2). However, qualitative analysis was interesting in term of distribution of the Right % and Left % (figures 1 and 2). In fact, the pattern was different according to the initial right [9] or left [6] focus. All patients except one with an initial left focus had a Right % lower than the Left %. The pattern was opposite in the group with the initial right focus [9] for all subjects except two. One subject in this group had a negative and similar Right and Left %. This qualitative analysis suggested that the Right % pattern seemed to be different between BECTS with initial right focus and BECTS with initial left focus.

Discussion

Our results indicate some subtle qualitative differences in cognitive functioning in BECTS and CAE patients.

The study of intelligence and memory testing showed normal results for patients with BECTS. The intellectual performances of CAE patients in complete remission were a little lower than those found in BECTS patients. In particular, the Picture Arrangement and Block-design sub-test scores were significantly lower in CAE patients. This lower score can be interpreted as being due to a delay in treatment or slowing of motor skills or also as reflecting an impairment in visuospatial processing. We failed to demonstrate any significant differences between the three populations with respect to executive functioning.

Using the dichotic listening task, we did not find any significant differences between the three groups, because of a ceiling effect. However, during the dual-task procedure, verbal activity seemed to influence right hand activity in BECTS, but results were not significant. A qualitative analysis of the % change as a function of the topography of the initial focus suggested that topography plays a role. Indeed, BECTS patients in remission following an initial right seizure focus demonstrated the same pattern as controls, whereas BECTS patients with an initial left focus, showed a reverse pattern.

Evaluation of the effects of epilepsies on cognitive functions in adults initially, involved epilepsies with a lateralized focus, whether they were related to a lesion or not [63]. In some childhood epileptic syndromes, a correlation between the type of syndrome, the topography of the paroxysmal activity and the cognitive pattern has been described [64, 65]. Focal epileptic discharges during development can modify the structure and function of the area involved, as well as the construction of certain connections [66]. Riva et al. [67] described a loss of specialization of the epileptic hemisphere independently of the existence or absence of a discernible lesion on CT scan in twenty four children with a unilateral epileptic focus without the classical pattern of BECTS.

The impact of epilepsy on language development was investigated by Cohen and Lenormand [68]. They studied six right-handed children with partial epilepsy (left frontal focus), between the ages of 3 and 9 years-old and reported impairment in language. Comprehension progressively increased over time, achieving a normal level around the age of 7 years, but expression remained impaired.

Normal cognitive findings are part of the definition of BECTS [9]. In rare patients, neuroimaging has revealed some abnormalities [15, 16]. Recent studies have described neuropsychological disorders in the active phase of BECTS, involving global cognitive and lateralized hemispheric functioning. Weglage et al. [38] confirmed neuropsychological and behavioural disorders in a group of untreated children with a rolandic focus, with or without seizures, when compared with a paired control group. Global IQ was within normal limits and the PIQ was lower in BECTS patients, while the VIQ was the same in both groups. Staden et al. [32] reported selective language impairment in twelve out of twenty children with BECTS, although their IQs were normal without correlation with the topography of the focus because the majority of the patients studied had a bilateral focus. Piccirilli et al. [39] evaluated the lateralization of language functions in BECTS using a dual-task procedure in twenty-two right-handed children with a right or left rolandic focus by comparing them to a paired control group.The performances of children with a right rolandic focus were similar to those found in the control group suggesting that language was lateralized to the left hemisphere whereas those with a left focus demonstrated a different pattern, suggesting a bilateral representation of language functions.

Our results confirm that global intellectual function was normal in BECTS. Some subtle deficits in lateralized hemispheric function may be suspected, after the disappearance of seizures in adults in total remission from BECTS. CAE patients had normal, although lower, global IQ and Performence IQ, suggesting a deficit in visuospatial functioning or in cognitive speed. We found no difference between populations in tasks evaluating the executive functions. There is thus no evidence of frontal dysfunction in CAE patients in remission contrary to what has been described in another idiopathic epilepsy: juvenile myoclonic epilepsy [69-71]. We were unable to demonstrate any difference among the three populations during the dichotic listening task since it was limited by a ceiling effect.

Our results are preliminary and the sample was small. Qualitative analysis during the dual-task procedure, according to the pattern of Right and Left % suggested, some reorganization in language functions, according to the initial topography of the focus as found by Piccirilli [39]. None of the tests used to evaluate language functions (WAIS R verbal sub-tests, verbal fluency, Binois Pichot) were abnormal. A more detailed evaluation of language functions would be useful to evaluate the consequences of focus localization

There is no single mechanism explaining a suspected disorder of functional organization. One interpretation is that there is an interference between paroxysmal abnormalities and the cognitive functions, as described in the Landau-Kleffner [73] or continuous spike-waves during sleep syndromes (CSWS, 74-77). Similarly, Binnie [33, 37] postulated a relationship between paroxysmal rolandic abnormalities and cognitive functions. Impairment of brain maturation was also suggested to be responsible for both the seizure activity and the neuropsychological defects [78-81]. However, the exact nature and cause of this hypothetical impairment of brain maturation is still unknown [80]. According to Doose and Baier [78], the disorder in cerebral development may precede the electroencephalograhic abnormalities. In fact, epilepsy appears after the acquisition of oral language but the time of onset of the epileptic dysfunction remains unknown [82]. In our study, groups became too small for statistical evaluation of the effect of age at onset of epilepsy on cognitive functions. Functional imaging studies in BECTS have supplied very little additional information [31, 83] failing to demonstrate any right/left asymmetry in glucose distribution as a function of the topography of the epileptogenic focus.

BECTS has a good prognosis in terms of epilepsy. However, we suspect that a selective disorder into cerebral functional organization should persist into adulthood in term of reorganization of language function. This study is a preliminary one and results must be confirmed with a larger sample. Detailed longitudinal studies are needed in order to evaluate lateralized hemispheric functioning and the language functions in adulthood.

CONCLUSION

Acknowledgements:

The authors would like to thank Mrs Christine Clerc for her assistance and the Clinical Research Center of the Tours Medical School for their help with the statistical tests.

Received May 14, 2001 Accepted November 9, 2001