From rolandic epilepsy to continuous spike-and-waves during sleep and Landau-Kleffner syndromes: Insights into possible genetic factors


Address correspondence to Gabrielle Rudolf, PhD, Service de Neurologie, Hôpitaux Universitaires de Strasbourg, 1 place de l’Hôpital BP 426, 67091 Strasbourg cedex France. E-mail:


Epilepsy is a frequent neurologic disease in childhood, characterized by recurrent seizures and sometimes with major effects on social, behavioral, and cognitive development. Childhood focal epilepsies particularly are age-related diseases mainly occurring during developmental critical periods. A complex interplay between brain development and maturation processes and susceptibility genes may contribute to the development of various childhood epileptic syndromes associated with language and cognitive deficits. Indeed, the Landau-Kleffner syndrome (LKS), the continuous spike-and-waves during sleep syndrome (CSWS), and the benign childhood epilepsy with centrotemporal spikes (BCECTS) or benign rolandic epilepsy, are different entities that are considered as part of a single continuous spectrum of disorders. Genetic predisposition with simple to complex modes of inheritance has long been suspected for this wide group of childhood focal epilepsies. Recent reports on the involvement of the SRPX2 and ELP4 genes with possible roles in cell motility, migration, and adhesion have provided first insights into the complex molecular bases of childhood focal epilepsies.

Childhood epileptic syndromes have relatively specific age-dependent onset and display typically electroencephalography (EEG) features such as focal sharp waves in the centrotemporal area, but the causes remain unknown. It is well established that genetic predisposition plays a major role in the etiology of idiopathic epilepsies, whether generalized or focal. Evidence for this genetic contribution has come from twin studies showing significantly higher concordance rates in monozygotic than in dizygotic pairs, with focal or even more with generalized epilepsy (Berkovic et al., 1998), and from familial linkage studies that in turn have contributed to the identification of specific genes (Robinson & Gardiner, 2000; Guerrini et al., 2003; Berkovic et al., 2006; Taylor et al., 2008). Although rare monogenic forms of epilepsies are now well recognized, there is evidence for complex inheritance due to multiple susceptibility genes in most idiopathic epilepsies. The aim of this article is to provide further proof, from a review of the literature and personal data, of an insight into possible genetic factors underlying some forms of idiopathic focal epilepsies of childhood (IFEs).

Idiopathic Focal Epilepsies of Childhood and the Endophenotype of Centrotemporal Spikes (CTS)

IFEs include a broad spectrum of age-related epileptic syndromes characterized by recurrent seizures with or without major effects on social, behavioral, and cognitive development. The most frequent and well-known IFEs are benign occipital epilepsies of childhood (Panayiotopoulos and Gastaut subtypes) and benign childhood epilepsy with centrotemporal spikes also known as rolandic epilepsy. In addition to these two syndromes, the continuous spike-and-waves during sleep syndrome and the Landau-Kleffner syndrome of acquired epileptic aphasia represent more severe and less frequent forms of IFE. These two latter syndromes are considered different clinical expressions of the same pathologic entity because both are characterized by the association of seizures with paroxysmal EEG discharges activated during drowsiness and sleep and with neuropsychological deficits (Doose et al., 1996).

Characteristic diphasic focal sharp centrotemporal spikes (CTS) often followed by slow wave or rolandic spikes are the EEG hallmark of focal epilepsy of childhood. Given that in focal epilepsy of childhood, clinical manifestations alone show complicated or even controversial genetic contribution, the CTS trait has been used as an endophenotype (Neubauer et al., 1998; Strug et al., 2009). Endophenotype represents simpler cues to genetic underpinnings than the disease syndrome itself and can result in a more straightforward genetic analysis (Gottesman & Gould, 2003). CTS as a subclinical marker not only seems to be under strong genetic influence (Doose et al., 1997) but—in contrast to its clinical phenotype counterpart—may also be inherited as a monogenic trait (Degen & Degen, 1990; Neubauer, 2000; Bali et al., 2007). Indeed, up to half of the children with CTS will not show any clinical manifestations, suggesting that additional genetic and environmental factors sustain the emergence of actual clinical symptoms; a given CTS genetic locus would act in combination with other genetic and nongenetic factors to produce the phenotype of BCECTS (Doose et al., 1997; Callenbach et al., 2005; Clarke et al., 2007). The association of rolandic epilepsy or of CTS with reading disability (dyslexia) and with speech impairment such as dysphasia has often been suggested (Doose et al., 1997; Staden et al., 1998; Michelucci et al., 2008). A complex interplay between brain development and maturation and various susceptibility genes may sustain these related disorders that very probably have molecular mechanisms in common.

Genetics of IFE

Among the broad spectrum of IFEs, strong genetic susceptibility has been suggested in BCECTS (Bali et al., 2005). By considering the CTS endophenotype as the affected trait, evidence for linkage had been proposed at chromosome 15q14 (Neubauer et al., 1998); since then, it has been shown that the segregation ratio of CTS in BCECTS families is consistent with a highly penetrant autosomal dominant inheritance (Bali et al., 2007). Studies showing no co-twin concordance (Vadlamudi et al., 2004, 2006) have suggested that conventional genetic influences in BCECTS may be considerably less than for idiopathic generalized epilepsies, and hence that other mechanisms such as environmental or epigenetic factors may well play important roles and need to be explored. However, previous studies had clearly shown evidence for genetic contribution in BCECTS; several explanations including methodologic issues may account for this apparent discrepancy (Greenberg & Pal, 2007). Very recently, a genome-wide study demonstrated linkage of the centrotemporal sharp wave EEG trait to 11p13 and several polymorphic markers in the ELP4 (Elongator Protein Complex 4) gene showed association with the phenotype (Strug et al., 2009). This study not only and obviously argued in favor of genetic contribution, but also provided a first insight into the possible molecular bases of common BCECTS. The Elongator complex is involved in gene transcription and in tRNA modification, and members of the Elongator complex have been implicated in cell motility and in cell migration, and of the neurons during cortical development particularly (Creppe et al., 2009).

In contrast with most canonical forms of BCECTS, a rare syndrome combining the features of BCECTS and of speech and language disorders (rolandic epilepsy with speech dyspraxia, autosomal dominant: RESDAD) has shown monogenic inheritance. RESDAD was described in an Australian family (Scheffer et al., 1995) not large enough to enable successful genetic linkage analysis. A French family presenting with a phenotype highly similar to RESDAD but with X-linkage (RESDX syndrome) was then reported; as in the former Australian family, affected members had rolandic seizures associated with speech dyspraxia and variable degrees of mental retardation. Genetic study led to the identification of the RESDX-causing gene, SRPX2 (Roll et al., 2006). Another SRPX2 mutation was found in bilateral perisylvian polymicrogyria (BPP) (Roll et al., 2006) and argued further in favor of a link between (benign) rolandic epilepsy and (severe) BPP. Generally and together with the more recent ELP4 example, the SRPX2 case makes it elusive the distinction between so-called “nondevelopmental” versus “developmental” disorders (Ben-Ari, 2008), since RESDX syndrome could represent a still-undetectable and hence yet-ignored developmental disorder as well. It was reported that SRPX2 is involved in cellular migration and adhesion in cancer cells (Tanaka et al., 2009). Recently also, interaction of the secreted protein SRPX2 with a membrane-anchored receptor, uPAR (plasminogen activator, urokinase receptor) as well as with other partners such as cathepsin B, was demonstrated (Royer-Zemmour et al., 2008). Because uPAR plays an important and well-known role in cell proliferation, migration, and adhesion, the SRPX2/uPAR association might provide a possible molecular explanation for the pathophysiologic mechanisms underlying IFEs and BPP. The existence of several SRPX2 partners in the central nervous system would be consistent with a pleiotropic role for SRPX2 in the brain, especially if the partners are not all expressed in the same territories, and at the same time during brain development and maturation.

Other families with monogenic inheritance of a syndrome with clinical features reminiscent, if not identical, to RESDAD and RESDX, have been described more recently (Kugler et al., 2008; Michelucci et al., 2008). In these families, inheritance patterns (excluding SRPX2 locus at Xq22) and/or linkage analyses have excluded loci previously associated with various related phenotypes. Comorbidity between BCECTS and altered development of speech motor control (also known as speech sound disorder) has also been further demonstrated in a recent case-control study (Clarke et al., 2007). Another BCECTS-associated genetic syndrome has been reported in a consanguineous family with rolandic epilepsy, paroxysmal exercise-induced dystonia, and writer’s cramp. The syndrome is inherited as a recessive trait and linkage was found at 16p12-p11 (Guerrini et al., 1999), within the “infantile convulsions and choreoathetosis (ICCA)” critical region (Rochette et al., 2008).

It is now considered that the syndrome of continuous spike-and-waves during sleep and the Landau-Kleffner syndrome constitute more severe edges of a continuum with BCECTS at the other and more benign end. An impairment of brain development and/or maturation with genetic predisposition has also been questioned in CSWS and in LKS. Various neuroimaging abnormalities including pachygyria and perisylvian polymicrogyria have been described in CSWS (Van Hirtum-Das et al., 2006); in LKS, structural brain abnormality is uncommon. In a recent study, data reported a volume reduction in the bilateral superior temporal areas. However, it is unclear whether this reduction is the cause or the consequence of LKS (Takeoka et al., 2004) and first association with perisylvian polymicrogyria was reported recently (Huppke et al., 2005). The first pair of monozygotic twins both with CSWS was reported only a dozen years ago (Blennow & Ors, 1995). More recently, the first clinical, EEG and cerebral imaging data about two families combining BECTS and cryptogenic epilepsy with CSWS in first-degree relatives, was reported (De Tiège et al., 2006). These data suggested the existence of a common genetic basis shared by BCECTS on the one hand and cryptogenic epilepsy with CSWS on the other hand. The more severe clinical expression in such familial cases with CSWS could be due to other genetic (modifying) or to acquired factors, and may be reminiscent of the rare occurrence of patients with Dravet syndrome in families with inherited GEFS+ (generalized epilepsy and febrile seizures plus) syndrome (Singh et al., 2001). Similarly, the genetic predisposition to LKS is one of the many other pathophysiologic mechanisms that have been considered (Nieuwenhuis & Nicolai, 2006). Although more than 300 LKS cases have been described worldwide, only two sets of siblings with LKS have been reported (Landau & Kleffner, 1957; Nakato et al., 1989), and the one pair of monozygotic twins reported did not show concordance: one twin had LKS, and the other did not (Feekery et al., 1993).


The LKS, the CSWS syndrome, and BCECTS or rolandic epilepsy are different syndromes that are considered as part of a single continuous spectrum of disorders. Although identifying a genetic component in those syndromes has long appeared elusive, the genetic study of the rare and monogenic RESDAD and RESDX syndromes that associate rolandic seizures with verbal dyspraxia, and the genetic study of the centrotemporal sharp wave EEG endophenotype, have both yielded considerable progress in recent years. The SRPX2 gene is mutated in the RESDX syndrome as well as in bilateral perisylvian polymicrogyria, a developmental disorder of the brain that can also be found in patients with CSWS and LKS, whereas ELP4 variants associated with the EEG trait of BCECTS might partially interfere with the Elongator function during development. The regulatory and interactive pathways associated with SRPX2 and with ELP4 provide first and crucial clues to start explaining some of the pathologic aspects of age-related neurodevelopmental disorders associated with epilepsy and with cognitive impairments.


This work was supported by FRC (Fédération pour la Recherche sur le Cerveau) and by ANR/MRAR EPICOGN (Agence Nationale de la Recherche/Maladies Rares).

Disclosure: None of the authors has any conflicts of interest to disclose.