The system epilepsies: A pathophysiological hypothesis

Epilepsia, 53(5):771–778, 2012 doi: 10.1111/j.1528-1167.2012.03462.x

CRITICAL REVIEW AND INVITED COMMENTARY

The system epilepsies: A pathophysiological hypothesis *Giuliano Avanzini, yPaolo Manganotti, zStefano Meletti, xSolomon L. Moshe´, *Ferruccio Panzica, {Peter Wolf, and **Giuseppe Capovilla *Department of Neurophysiology, IRCCS Foundation Neurological Institute ‘‘Carlo Besta,’’ Milano, Italy; yDepartment of Neurological, Neuropsychological, Morphological and Movements Sciences, University of Verona, Verona, Italy; zDepartment of Neuroscience, University of Modena and Reggio Emilia, Modena, Italy; xSaul R. Korey Department of Neurology, Dominick P. Purpura Department of Neuroscience and Department of Pediatrics, Laboratory of Developmental Epilepsy, Montefiore/Einstein Epilepsy Management Center, Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, New York, U.S.A.; {The Danish Epilepsy Center, Dianalund, Denmark; and **Epilepsy Center, Department of Child Neuropsychiatry, C. Poma Hospital, Mantova, Italy

SUMMARY We postulate that ‘‘system epilepsies’’ (SystE) are due to an enduring propensity to generate seizures of functionally characterized brain systems. Data supporting this hypothesis—that some types of epilepsy depend on the dysfunction of specific neural systems—are

The SystE Concept We propose the hypothesis that some types of epilepsies—‘‘system epilepsies’’ (SystEs)—reflect the pathologic expression (ictogenesis) of an identifiable neural system made up of brain areas, the integrated activity of which subserve normal physiologic functions (Salek-Haddadi et al., 2009; Moeller et al., 2011). We suggest, too, that SystEs lead to functional results that cannot be obtained by pathologic activity within the individual elements alone. The phenomenology associated with a given SystE is determined by a contextual involvement of the contributing structures, and therefore predictable according to the functional specialization of the involved system. This notion makes SystEs different from the epilepsies resulting from the spread of a discharge originating in a more or less circumscribed region (the boundaries of which do not necessarily coincide with those of a functionally defined brain area) and propagating sequentially along one or more neural pathways to other brain areas (which do not necessarily belong to a unitary brain system). The concept of epilepsies due to the involvement of a brain system was discussed by the International League Against Epilepsy (ILAE) Task Force on Classification and Accepted February 23, 2012. Address correspondence to Giuliano Avanzini Department of Neurophysiology, Fondazione IRRCS, Istituto Neurologico Carlo Besta, Via Celoria 11, 20133 Milano, Italy. E-mail: [email protected] Wiley Periodicals, Inc. ª 2012 International League Against Epilepsy

reviewed. The SystE hypothesis may drive pathophysiologic and clinical studies that can advance our understanding of epilepsies and can open up new therapeutic perspectives. KEY WORDS: Generalized epilepsies, Focal epilepsies, Neural systems, Thalamo-cortical systems, Cortico-cortical connections.

Terminology (Engel, 2006) and formulated by Wolf (2006). It has been the subject of a number of ongoing discussions, many of which have been published (Bertram et al., 2008; Avanzini, 2009; Capovilla et al., 2009). The SystE hypothesis postulates that the ‘‘enduring propensity to generate seizures’’ (Fisher et al., 2005) of some epilepsies is due to the specific susceptibility of a system as a whole, although it may be possible to identify some trigger areas within the system. The neural system responsible for SystEs is revealed by the clinical/electroencephalography (EEG) semiology of the seizures, but the concept refers to the persistent susceptibility of the seizure-generating system, which is assumed to exist also in the interictal period. When evaluating the criteria for defining a neural system responsible for SystEs, it is worth considering a definition of a system used in engineering, mathematics, and information technology: ‘‘A system is a construct or collection of different elements that together produce results not obtainable by the elements alone. The elements, or parts, can include people, hardware, software, facilities, policies, and documents; that is, all things required to produce systemlevel results. The results produced by a system can include functions, behaviors, system-level qualities, properties, characteristics, and performances. The value added by the system as a whole, beyond that contributed independently by the parts, is primarily created by the relationship among the parts: that is, how they are interconnected’’ (Rechtin, 2000). Although this definition is concerned mainly with human activity and seems to be aimed at concrete solutions (see also the fellows’ consensus of the International Council on Systems Engineering at http://www.incose.org), it

771

772 G. Avanzini et al. contains the essence of a system: the presence of dynamic interactions among components and the overall behavior emerging from these interactions, which appears when all of the parts of a system work together. For example, a clock has the functional property of being a device for measuring time only once it has been assembled from its elements. Also relevant to the present discussion are the definitions of ‘‘simple’’ and ‘‘complex’’ systems. Simple systems have a small number of components that act in accordance with well-known laws (e.g., the simple pendulum, the behavior of which can be described using Newton’s equations of motion). Examples in neurology include the neuromuscular junction and the spinal sensorimotor reflex arc. Complex systems, in contrast, consist of several components interacting with each other in such a way that their collective behaviors, arising from the interaction of subsystems, cannot be expressed and are understood merely as a sum of the behaviors of their parts. Complex systems are characterized by nonlinear components with nonlinear interactions among these components, so that small perturbations or changes in their parameters or inputs may produce dramatic changes in their macroscopic behaviors (Lopes da Silva et al., 2003). Complex systems should not be confused with complicated systems in which, even if they have many components, the interactions are relatively simple and the results highly predictable (Amaral & Ottino, 2004). One example of a noncomplex, complicated system is an aircraft that is built from a very large number of components assembled on the basis of a complicated plan to perform a relatively limited number of well-defined and mainly predictable activities. Unlike complicated systems, the emergent properties or functions of a complex system: 1 Depend more on the interaction of its components than on their individual properties; 2 Are not due to the existence of a central controller or external organizing supervisors; 3 Postulate the existence of rules of interaction, which may not be explicitly defined, and are investigated upon the assumption that a knowledge of them would lead to a full understanding of system functioning. On the basis of the preceding, many neural systems can be viewed as being composed of a set of synaptically interconnected areas, the coordinated activities of which lead to functional results that cannot be obtained by the activity of the elements alone. Anatomic connections are necessary but not sufficient to define such a system, which emerges only when the different components actively participate in accomplishing system functions. Examples of activities depending on complex neural systems are the EEG rhythms generated by thalamocortical systems that are involved in the control of attention, vigilance, and sleep. A great deal of experimental evidence converges in supporting the primary role of the thalamic reticular nucleus, the thalamic nuclei projecting to the cortex, and the cortical areas and their interconnections in generating the 3 Hz spike and wave Epilepsia, 53(5):771–778, 2012 doi: 10.1111/j.1528-1167.2012.03462.x

(SW) activity associated with absence seizures. The primary role of one or another component of the circuit has long been debated, but the evidence supports the conclusion that all parts of the system are necessary to generate SW activity, and that no one component can do so in isolation (Avoli et al., 2001).

Investigational Approaches to Studying Neural Systems and Their Interactions A number of investigational tools can help us identify and characterize the neural systems involved in SysEs. The EEG and magnetoencephalography (MEG) studies have provided important insight into the organization of oscillating neural systems that generate various brain rhythms. Oscillatory brain activity may be one of the key mechanisms used by the brain to integrate the information processed in multiple specialized local brain areas (Varela et al., 2001). This synchronization-dependent integration has become known as functional or effective connectivity; as applied to a given neural system, ‘‘functional connectivity’’ identifies the causal interactions or the direction of information flow between brain regions. Functional connectivity has been studied in particular detail in the visual system, which includes a number of anatomically and physiologically distinct areas that are each specialized in processing a particular aspect of the visual scene, such as shape, motion, or color. The results of animal experiments indicate that the synchronization of neuronal activity in the visual cortex seems to be responsible for connecting different but related visual features so that a visual pattern can be recognized as a whole (Gray & Singer, 1989; Singer & Gray, 1995). Moreover, synchronization between areas of the visual and parietal cortex, and between areas of the parietal and motor cortex, has been observed during a visuomotor integration task in awake cat (Roelfsema et al., 1997). These findings suggest that studying the synchronization of oscillatory activities in different brain areas is a suitable means for identifying the different structures belonging to an integrated neural system, but its potential for studying candidate forms of SystE has not yet been explored. Once a neural system putatively responsible for a given SystE has been identified, electrical or magnetic stimulation can be used to test the excitability of its structures. The technique of signal analysis that is currently used for studying event-related potentials (ERPs) can be used to analyze farfield potentials expressing the activity of subcortical structures as well as early or late near-field potentials that more directly reflect cortical activation. Application of ERP analysis in studying a putative SystE is described in subsequent text, in the section devoted to focal epilepsies. Transcranial magnetic stimulation and electroencephalography (TMS-EEG) makes it possible to record the immediate reaction of a system (e.g., the thalamocortical system) to controlled perturbations of different cortical areas. In

773 System Epilepsies healthy subjects, Massimini et al. (2007) found that the TMS triggering of slow waves reveals intrinsic instability in thalamocortical networks during non–rapid eye movement sleep, and that evoked slow waves lead to a deepening of sleep and an increase in the EEG slow-wave activity (0.5– 4.5 Hz) that is thought to play a role in brain restoration and memory consolidation. Furthermore, TMS allows a selective analysis of the excitatory versus inhibitory components of a given neural system. For example, the paired-pulse method (Ziemann et al., 1996) has been used to investigate the reduction in cortical inhibition in various forms of epilepsy, including juvenile myoclonic epilepsy (JME) (Manganotti et al., 2000, 2004, 2006). Conventional neurophysiologic methods are still first-line techniques for investigating brain activities because of their unique ability to describe neural activities in real time. However, because their spatial resolution is limited, efforts have been made to develop recording systems based on a larger number of electrodes (high-density EEG) and to combine EEG with magnetic resonance imaging (MRI) and functional magnetic resonance imaging (fMRI) techniques. EEG–functional MRI (fMRI) combines MRI spatial resolution with EEG time resolution, and can therefore be profitably used to study the structure and function of the distributed neural systems putatively involved in SystE. EEG-fMRI corecording studies in patients with idiopathic generalized epilepsies (IGEs) (Aghakhani et al., 2004; Gotman et al., 2006) strongly support the epileptic condition as the expression of a hyperexcitable corticosubcortical network consisting of well-defined and discrete brain regions, rather than the expression of generalized brain dysfunction. Especially given that impaired consciousness is the hallmark of IGErelated seizures, it is interesting that the SW discharges associated with IGEs involve some brain regions that are part of the consciousness system (Blumenfeld, 2005). Reflex seizures occur in a proportion of SystEs and can provide unique insights into the ictogenesis of human epilepsies and the involved networks. Photosensitivity (PhS) is the most widely studied reflex epileptic phenomenon, found in cases of IGEs (Wolf & Goosses, 1986), which are here proposed as SystEs. The threshold response is a widespread bilateral, often symmetrical SW discharge, often with a posterior predominance, and is called the photoparoxysmal response (PPR). If it develops into a clinical seizure, the seizure may be a myoclonic, absence, or simple focal seizure of the occipital type with visual hallucinations or versive movements; the maximum possible response is a generalized tonic–clonic seizure (GTCS). It has been reported that increased excitability of the striate cortex is a precondition for PhS (Siniatchkin et al., 2007a). Parra et al. (2003) found that the synchronization of gamma oscillations was most prominent in the frontal and precentral regions when a myoclonic seizure was evoked, whereas gamma synchronization prevailed in parietal regions with reflex absences. Moeller et al. (2009b) evoked PPRs evolving to a GTCS; in this

phase, fMRI demonstrated the activation of the visual cortex, the thalamus, and the superior colliculi and lateral geniculate bodies, together with deactivation of the frontal and parietal cortex and the precuneus, suggesting a thalamocortical core system for the ictogenesis induced by PhS (and more generally IGEs). According to Moeller et al. (2011), the thalamocortical network between seizures shows no signs of functional deviation in patients with IGE, but Chavez et al. (2010) found increased connectivity with a clear modular structure in the brain networks of patients with absence epilepsy (AE).

Proposed Examples of SystEs We discuss absence epilepsy AE and JME as typical paradigms of SystE. We then consider benign childhood epilepsy with centrotemporal spikes (BCECTS) and West syndrome (WS) as potential SystE candidates. Absence epilepsy Since the earliest observations of Hans Berger (1933), investigators have been struck by the particular characteristics of the bilateral synchronous 3 Hz SW discharges associated with absence seizures, which were later considered by Penfield and Jasper (1954) to be the hallmark of centrencephalic epilepsies. The bilateral synchronous expression of SWs prompted investigators to speculate about a possible SW origin in a subcortical structure that could simultaneously recruit the cortical areas of both hemispheres. After the seminal paper of Jasper and Droogleveer-Fortuyn (1947), many other investigations supported the conclusion that the SW discharge is the result of epileptic activity generated within the cortico-thalamocortical circuitry. Penfield and Jasper (1954) included centrencephalic epilepsies in the category of ‘‘functional epilepsies,’’ that is, those in which no brain lesions are demonstrable and the origin of which cannot be attributed to any other cause but hereditary predisposition. AE thus complies with the main definition of SystE as a condition due to the persisting susceptibility of the thalamocortical system as a whole. However, during the following decades, the results of both animal experiments (using the rodent models Genetic Absence Epilepsy Rat from Strasbourg [GAERS] and Wistar Albino Glaxo Rats from Rijswijk [WAG/Rij]) and human studies modified and, to some extent, challenged the original views of AE. For example, the associated EEG discharges are more prevalent in the frontal regions in humans (Holmes et al., 2004) and in the parietal regions in GAERS and WAG/Rij. Moreover, using these same animal models, Meeren et al. (2002) and Polack et al. (2007) discovered that the SW discharges initiate in somatosensory cortical areas and immediately afterward invade the corticothalamic circuitry. We propose here that AE can be considered a good example of SystE in which the propensity to generate seizures is due to the specific susceptibility of the thalamocortical system, which is Epilepsia, 53(5):771–778, 2012 doi: 10.1111/j.1528-1167.2012.03462.x

774 G. Avanzini et al. endowed with oscillating properties. In rodent models, this susceptibility primarily affects the somatosensory component, with a trigger area located in the cortical representation of the face. The new experimental data are consistent with clinical observations of the regional prevalence of the EEG discharges associated with generalized seizures and their lateralized onset from a leading hemisphere that can be demonstrated by means of signal analysis. In the group discussions that took place in Antalya (Bertram et al., 2008), there was a general consensus that (1) the use of the term ‘‘focal’’ to describe the regional onset of SWs can be misleading, as it may create confusion with ‘‘true’’ focal seizures; (2) the available data support the idea of a trigger zone within a given thalamocortical system that has a particular genetically determined epileptogenic susceptibility; (3) the trigger area becomes a part of the oscillating network during absence seizures; and (4) the oscillations constitute an emergent property of the whole system (as in dynamic, nonlinear systems). The group that met in Antalya included both basic scientists and clinicians who agreed upon the concept of ‘‘system epilepsy’’ to designate epilepsies caused by the oscillating thalamocortical system that controls vigilance. This association is consistent with clinical and experimental observations of an increased probability of absence and SW occurrence during the transition from the waking state to sleep. This conclusion is also supported by experimental and computational models (see Lopes da Silva et al., 2003; Lytton, 2008) that provide further insights into the dynamics of such oscillatory processes. Simulations based on a dynamical nonlinear model could help in understanding critical parameters, variations in which can provoke sudden massive synchronous discharges. Future studies of rodents and humans, using correlation analysis of multichannel neurophysiologic recordings combined with imaging and biomolecular investigations, may answer such questions as the following: 1 Is there any evidence of the enduring propensity of the thalamocortical system to generate SW discharges? 2 Can a specific propensity of topographically organized thalamocortical subsystems be recognized? 3 Is there any further evidence of trigger areas within the involved thalamocortical subsystems? 4 Is it possible to identify changes in the excitable properties and underlying biologic mechanisms relevant to epileptogenesis in the involved section of thalamocortical system? 5 Are the biologic mechanisms responsible for thalamocortical dysfunction influenced by brain maturation, thus accounting for the age-dependence of AE? Juvenile myoclonic epilepsy and related conditions Like AE, JME is characterized by bilateral EEG discharges and classified as an IGE. The JME discharges Epilepsia, 53(5):771–778, 2012 doi: 10.1111/j.1528-1167.2012.03462.x

consist of SWs and polyspike-waves (PSWs), often associated with myoclonic jerks that may fuse into tonic–clonic seizures. PSWs and the associated myoclonus can be asymmetrical, occasionally unilateral, and occur predominantly on awakening. Absences and 3-Hz SWs are observed in one third of cases (Panayiotopoulos et al., 1989; Janz, 1994). Analyses of polygraph traces have shown that the myoclonus is consistently associated with a frontocentral short train of quasi-rhythmic 16–27 Hz polyspikes (Panzica et al., 2001) in the frequency range of movement-related fast central rhythms (Conway et al., 1995). Back-averaging analysis shows that the latency between the myoclonic jerk and the associated cortical spike is compatible with a cortical origin of myoclonus (Panzica et al., 2001). Analysis of the relationships between JME discharges and sleep rhythms shows a close relationship between 16 and 27 Hz cortical polyspike trains and spindles (Gigli et al., 1992; Avanzini et al., 2000), which suggests that cortical polyspikes are modulated by the thalamocortical loop responsible for spindle activity. The role of thalamocortical circuitry in polyspike generation is further supported by sleep microstructure analysis (Terzano et al., 1992) of patients with JME (Gigli et al., 1992; Avanzini et al., 2000). The above results suggest that a wide cortical area, including the motor and premotor cortices, is responsible for the PSW discharges, but that myoclonic events arise when the frequency of the polyspikes approaches that of central motor rhythm, and when paroxysmal activity reaches a critical threshold at which it activates the motor cortex. The premotor and motor cortical areas are the recipients of thalamic projections from thalamic nuclei that may play a critical role in generating thalamocortical rhythms. Furthermore, increased motor system connectivity has been demonstrated in patients with JME, indicating functional deviations within the motor system beyond the seizures (Vollmar et al., 2011). On the basis of the above evidences we hypothesize that JME is due to the susceptibility of a system that includes motor and premotor cortices and related subcortical areas. The oscillatory thalamic mechanisms may play a special role in triggering cortically generated multispikes associated with myoclonic jerks. Very intense and prolonged discharges generated in the system can lead to generalized tonic–clonic seizures. The frequent association of PhS provides an opportunity to reproduce the myoclonus and the associated discharges for pathophysiologic studies. A high proportion of patients with JME (30.5–42% according to Wolf & Goosses, 1986) show PhS, but whether a PhS response requires the involvement of oscillating thalamocortical circuitry is still a matter of debate (see above) and its electrophysiologic analysis can provide further insights into JME pathophysiology. Language-induced perioral reflex myocloni (LIPORMs) are single, small, lightning-like myoclonia in the lips, tongue, facial muscles, jaws, and throat that are the hallmarks of primary reading epilepsy and occur in a significant

775 System Epilepsies proportion of patients with JME (25–35% according to Mayer et al., 2006; and Guaranha et al., 2009). The ictal EEG discharge consists of single or very short groups of spikes or sharp waves that are unilateral in 30% of cases, bilaterally asymmetrical in 38%, and bilaterally symmetrical in 32%, with temporal dominance in 80% and frontal in 20% (Wolf, 1992). Unlike PhS, the LIPORM network does not respond rapidly to the stimulus but needs to be ‘‘warmed up’’ over a period of time that is intraindividually relatively stable. A network of the cortical areas involved in normal language, motor function and reading, together with associated subcortical structures (the striatum, thalamus and cerebellum), subserves the ictogenesis of LIPORMs (SalekHaddadi et al., 2009). Issues to be addressed regarding JME and related conditions include the following: 1 Is there any evidence that the thalamic nuclei projecting to motor and premotor cortical areas are part of the putative system responsible for JME? 2 What is the role of the basal ganglia in the JME-related system? 3 Does PhS or LIPORMs identify specific subtypes of JME-related system? 4 As in the case of AE, the enduring propensity of the system to generate seizures needs to be demonstrated. 5 Can brain maturation influence cerebral hyperexcitability and explain why JME and related conditions appear at a certain age? Benign childhood epilepsy with centrotemporal spikes Benign focal epilepsies are age-related epilepsies that present with ictal focal symptoms that are not associated with any other neurologic sign or symptom. The most common types, BCECTS and benign epilepsy with occipital paroxysms (the Gastaut type, and to some extent the Panayiotopoulos type) are related to the cortical representation of sensory systems, the effect of which on epileptic discharges provides some insight into their pathophysiology. Herein we discuss BCECTS, for which most of the information comes from human studies (there are no reliable animal models). Electrophysiologic analyses of centrotemporal spikes (often referred to as rolandic spikes) have provided some information concerning its generators. According to Kellaway (2000), rolandic spikes arise in the somatosensory cortex and are then transmitted to the motor cortex. The finding of changing sites of focal discharges in the ipsilateral or contralateral hemisphere suggests that the cortical dysfunction is distributed widely throughout the cortex. Further insights come from the analysis of focal spikes evoked by somatosensory stimulation, first reported by De Marco and Negrin (1973), in the form of contralateral foot tapping. Manganotti et al. (1998) demonstrated that the spikes evoked by tapping or electrical stimulation of the fingers of one hand were similar in morphology and scalp distribution

to spontaneously occurring spikes. In the same group of patients, somatosensory evoked potentials (SEPs) following first digit nerve stimulation were characterized by a late high-amplitude positive-negative-positive complex that resulted from the average of the evoked spikes, and could be differentiated from the late negative component (N60) of normal SEPs on the basis of its dipole configuration and dynamic range. Overall, the results of somatosensory stimulations are consistent with the hypothesis that the initial source of the evoked spikes is in the sensory cortex, with subsequent involvement of the motor and secondary sensory areas (Manganotti et al., 1998), which would also account for the observation that clinical seizures in some patients are heralded by paresthesia of the mouth and hand. The well-known activation of rolandic spikes during slow-wave synchronised EEG activity (non–rapid eye movement [REM] sleep) suggests that sleep-regulating thalamic nuclei may play a role in their generation. Nobili et al. (1999) found a high positive correlation between BCECTS discharges during sleep and delta (0.5–4.0 Hz) and spindle activity (12.0–16.0 Hz), with a much higher correlation coefficient for spindles than the delta band. These findings suggest that the thalamic networks that play an important role in pacing sleep spindles may also be involved in focal discharges responsible for rolandic spike generation (as they are known to be in the case of AE; see above). Of interest, Manganotti and Zanette (2000) demonstrated that motor evoked potentials (MEPs) induced by transcranial magnetic stimulation in patients with BCECTS were consistently facilitated by a repetitive conditioning electrical stimulation (of the thumb) in the spindle frequency range. In discussing the role of thalamic circuitry in generating rolandic spikes, Huguenard (2000) proposed that his studies support the possibility that local recurrent oscillatory activity can be generated within the thalamocortical circuitry. The above data suggest that subtle age-dependent dysfunctions of the somatosensory system (i.e., its thalamocortical part) may play an important role in the pathophysiology of BCECTS. Further studies can be designed to prove or disprove the hypothesis that BCECTS and other idiopathic focal epilepsies can be described as SystEs. Because some of the statements that appear in the preceding text are only indirectly supported by the available evidence, hypothesis testing would be greatly benefited from experiments in suitable experimental models, which sadly are not yet available. West syndrome WS is an epileptic encephalopathy characterized by infantile spasms and hypsarrhythmia, and often by psychomotor regression including autistic behaviors. The latter often appear later and may indicate a progression of the syndrome. In this respect, there may be evidence of progression based on a continuing interaction of underlying systems and epigenetic influences. The syndrome is a relatively rare Epilepsia, 53(5):771–778, 2012 doi: 10.1111/j.1528-1167.2012.03462.x

776 G. Avanzini et al. manifestation of common insults, including genetic mutations, brain dysplasias, perinatal asphyxia, and other perinatal traumatic events. It is not yet clear whether different mechanisms are responsible for hypsarrhythmia and infantile spasms and whether the psychomotor regression is a direct consequence of the hypsarrhythmic EEG pattern (Dulac, 2001; Dulac et al., 2010). To address these questions, the presumed substrates of WS need to be identified, and may encompass multiple channels of altered functions. The manifestations of WS include exaggerated motor phenomena that involve the pyramidal and extrapyramidal pathways as well as the assumed brainstem-originating tracts that are responsible for the expression of tonic seizures (spasms) in animal studies, intermittent dyscognitive states implicating thalamocortical or reticular formation involvement, and failure to acquire new developmental milestones and regression that may involve widespread networks organized as systems. In the recent past, two presumably competing brainstem and cortical hypotheses have been proposed to explain the pathophysiology of the spasms (Kellaway, 1959; Tucker & Solitare, 1963; Bignami et al., 1964; Branch & Dyken, 1979; Tominaga et al., 1986; Chugani et al., 1990; Panzica et al., 1999; Lado & Moshe, 2002; Frost & Hrachovy, 2003). The brainstem hypothesis postulates that a brainstem generator may be responsible, whereas the cortical hypothesis emphasizes cortical malfunction. It is not clear why all infants with similar underlying insults (i.e., dysplasias, genetic mutations, etc) will not develop WS. This heterogeneity suggests an acquired (epigenetic) influence. Contributing factors that may regulate the systems underlying the expression of the key features of WS include dysfunction of the hypothalamic-pituitary-adrenal axis (Nalin et al., 1985; Baram et al., 1992, 1995) and an immune-mediated disorder (Hrachovy & Frost, 1989). These factors may explain the efficacy of steroids in this condition, at least in terms of controlling the spasms and hypsarrhythmia (although not necessarily in altering the cognitive outcomes). The electrodecremental response and hypsarrhythmia each reflects diffuse or multifocal/system dysfunction. Electrodecremental seizures are hypothesized to arise from paroxysmal activity primarily in the cortex. Alternatively, electrodecremental seizures may result from increased activity in subcortical circuits projecting to cortex, leading to diffuse desynchronization and abnormal cortical electrical activation. Dysfunction in the arousal systems of the brainstem could alter cortical ‘‘tone’’ and result in an abnormal EEG background. Stimulation of small regions of the brainstem can produce global alterations in cortical EEG (Moruzzi & Magoun, 1949). Inactivation of the brainstem reticular activating system may produce widespread changes in cortical activity and impairment of consciousness (Plum & Posner, 1980). Paroxysmal activity in subcortical arousal systems might induce abrupt changes in cortical tone that could appear as electrodecremental Epilepsia, 53(5):771–778, 2012 doi: 10.1111/j.1528-1167.2012.03462.x

responses. The arousal system of the upper pons colocalizes with reticulospinal projecting neurons projecting caudally and may mediate the ‘‘startle-like’’ movements associated with the spasms (Magoun, 1963; Vining, 1990). It has been proposed that hypsarrhythmia may represent ongoing seizure activity, and that infantile spasms and electrodecremental events result from activation of subcortical circuits attempting to control cortical seizure activity (see in Lado & Moshe, 2002). Recent EEG-fMRI studies of WS have demonstrated that epileptiform discharges in hypsarrhythmia are associated with hemodynamic and metabolic changes in the cerebral cortex, and that high-voltage slow waves correlate with blood oxygen–level dependent (BOLD) changes in cortical and subcortical structures (Siniatchkin et al., 2007b). Lado and Moshe (2002) have emphasized the importance of the interaction between cortex and subcortical regions in WS, and the requirement that both regions contribute to WS. They proposed that infantile spasms originate in the abnormal interaction of cortical and subcortical circuits rather than in either region alone, and that this abnormal interaction between cortical and subcortical circuits may be further augmented by a delay in the maturation of white matter connections between cortical and subcortical regions. The data show that a dysfunction of a single brain structure cannot be responsible for such a complex manifestation as WS, and that the typical electroclinical picture requires the active participation of a pathologic system in which different brain areas (the cortex, thalamic nuclei and brainstem) work (or do not work) together. When some of these stations do not work, different electroclinical phenotypes develop. WS can therefore be proposed for further investigations that might define the structures of the involved system and their interconnections. The renewed interest in developing ‘‘realistic’’ animal models (Scantlebury et al., 2010; Chachua et al., 2011; Swann & Moshe, 2012) provide the means to test the hypothesis that WS may be a prototype of SystE, with presentation of symptoms and signs depending on the properties of the system at given time, as a function of possible underlying pathology, and the developmental stage of the brain (necessary for the exquisitely narrow window in which West syndrome occurs). Taking into account the possible role of cortical and subcortical structures, as well the maturation of myelinization and immune dysfunction, Scantlebury et al. (2010) created a model of symptomatic infantile spasms that may provide insights into the system underlying the expression of hypsarrhythmia and spasms.

Conclusions The data reviewed in this article support the hypothesis that some types of epilepsy may depend on the dysfunction of a specific brain system, and go beyond the dichotomy between focal and generalized epilepsy. The SystE

777 System Epilepsies hypothesis may be a useful means of driving pathophysiologic studies that can open up new therapeutic perspectives. The hypothesis postulates that SystE depends on a specific susceptibility of a given neural system to epileptogenic factors. Some of the epilepsies that we propose as examples of SystEs are thought to be genetically determined, which raises the question of the possible genetic mechanism responsible for the elective system. Another key question is how developmental variables may account for the age-dependency of a given SystE. Many other questions can be generated by the SystE hypothesis, and we believe that its ability to generate testable hypotheses is one of its main merits.

Acknowledgments We gratefully acknowledge the contributions of the following participants in the workshop on system epilepsy (Monreale, Italy. April 22–24, 2011): F. Beccaria, G. Bertini, R. Caraballo, A. Covanis, N. Fejerman, E. Ferlazzo, C. Ozkara, A. Romeo, N. Specchio, G. Stranci, P. Striano, W. van Emde Boas, F. Vigevano, F. Zara. G. Avanzini has received research support from the European Union through the FP6 LIFESCIHEALTH, Project ‘‘Epicure,’’ under contract no. LSHM-CT-2006-037315. S.L. Mosh has received research support from NIH: R01 NS20253 (PI), R01-NS43209 (Investigator), 2UO1-NS45911 (Investigator), and the Heffer Family Foundation.

Disclosure S.L. Mosh serves on the editorial boards of Neurobiology of Disease, Epileptic Disorders, Brain and Development, and Physiological Research. He has received a consultancy fee from Eisai and a speaker’s fee from GSK. The remaining authors have no conflicts of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

References Aghakhani Y, Bgshaw AP, Benar CG, Hawco C, Andermann F, Dubeau F, Gotman J. (2004) fMRI activation during spike and wave discharges in idiopathic generalized epilepsy. Brain 127:1127–1144. Amaral LAN, Ottino JM. (2004) Complex networks. Eur Phys J B 38:147– 162. Avanzini G. (2009) System epilepsy: reply to Lders et al. Epilepsia 50:2326. Avanzini G, Panzica F, de Curtis M. (2000) The role of the thalamus in vigilance and epileptogenic mechanisms. Clin Neurophysiol 111(Suppl. 2): 19–26. Avoli M, Rogawski MA, Avanzini G. (2001) Generalized epileptic disorders: an update. Epilepsia 42:445–457. Baram TZ, Mitchell WG, Snead OC 3rd, Horton EJ, Saito M. (1992) Brainadrenal axis hormones are altered in the CSF of infants with massive infantile spasms. Neurology 42:1171–1175. Baram TZ, Mitchell WG, Hanson RA, Snead OC 3rd, Horton EJ. (1995) Cerebrospinal fluid corticotropin and cortisol are reduced in infantile spasms. Pediatr Neurol 13:108–110. Berger H. (1933) Uber das Elektrenkephalogram des Menschen. Sieben Mitteilung. Arch Psychiat Nervenkr 100:310–320. Bertram EH, Onat FY, Ozkara C, Mosh SL, Avanzini G. (2008) Workshop on idiopathic generalized epilepsies. Bridging basic science and clinical research (October 3–6, 2007 Antalya, Turkey). Epilepsia 49(Suppl. 11): 1969–1972. Bignami A, Zappella M, Benedetti P. (1964) Infantile spasms with hypsarrhythmia: a pathological study. Helv Paediatr Acta 19:326–342.

Blumenfeld H. (2005) Consciousness and epilepsy: why are patients with absence seizures absent? Prog Brain Res 150:271–286. Branch CE, Dyken PR. (1979) Choroid plexus papilloma and infantile spasms. Ann Neurol 5:302–304. Capovilla G, Berg AT, Cross JH, Moshe SL, Vigevano F, Wolf P, Avanzini G. (2009) Conceptual dichotomies in classifying epilepsies: partial versus generalized and idiopathic versus symptomatic (April 18–20, 2008, Monreale, Italy). Epilepsia 50:1645–1650. Chachua T, Yum MS, VelÐkov J, VelÐek L. (2011) Validation of the rat model of cryptogenic infantile spasms. Epilepsia 52:1666–1677. Chavez M, Valencia M, Navarro V, Latora V, Martinerie J. (2010) Functional modularity of background activities in normal and epileptic brain networks. Phys Rev Lett 104:118701.1–118701.4. Chugani HT, Shields WD, Shewmon DA, Olson DM, Phelps ME, Peacock WJ. (1990) Infantile spasms: I. PET identifies focal cortical dysgenesis in cryptogenic cases for surgical treatment. Ann Neurol 27:406–413. Conway BA, Halliday DM, Farmer SF, Shahani U, Maas P, Weir AI, Rosenberg JR. (1995) Synchronization between motor cortex and spinal motoneuronal pool during the performance of a maintained motor task in man. J Physiol 489:917–924. De Marco P, Negrin P. (1973) Parietal focal spikes evoked by contralateral tactile somatotopic stimulation in four non-epileptic subjects. Electroencephalogr Clin Neurophysiol 34:308–312. Dulac O. (2001) What is West syndrome? Brain Dev 23:447–452. Dulac O, Bast T, Dalla Bernardina B, Gaily E, Neville B. (2010) Infantile spasms: toward a selective diagnostic and therapeutic approach. Epilepsia 51:2218–2219. Engel J Jr. (2006) Report of the ILAE classification core group. Epilepsia 47:1558–1568. Fisher RS, van Emde Boas W, Blume W, Elger C, Genton P, Lee P, Engel J Jr. (2005) Epileptic seizures and epilepsy. Definitions proposed by the International League against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epilepsia 46:470–472. Frost JD Jr, Hrachovy RA. (2003) Infantile spasms: diagnosis, management and prognosis. Kluver Academic Publishers, Boston/Dordrecht/London. Gigli GL, Calia E, Marciani MG, Mazza S, Mennuni G, Diomedi M, Terzano MG, Janz D. (1992) Sleep microstructure and EEG epileptiform activity in patients with juvenile myoclonic epilepsy. Epilepsia 33:799–804. Gotman J, Kobayashi E, Bagshaw AP, Bnar CG, Dubeau F. (2006) Combining EEG and fMRI: a multimodal tool for epilepsy research. J Magn Reson Imaging 23:906–920. Gray CM, Singer W. (1989) Stimulus-specific neuronal oscillations in orientation columns of cat visual cortex. Proc Natl Acad Sci U S A 86:1698–1702. Guaranha MSB, da Silva Sousa P, de Araffljo-Filho GM, Lin K, Ferreira Guilhoto LMF, Caboclo LOSF, Yacubian EMT. (2009) Provocative and inhibitory effects of a video-EEG neuropsychologic protocol in juvenile myoclonic epilepsy. Epilepsia 50:2446–2455. Holmes MD, Brown M, Tucker DM. (2004) Are ‘‘generalized’’ seizures truly generalized? Evidence of localized mesial frontal and frontopolar discharges in absence Epilepsia 45:1568–1579. Hrachovy RA, Frost JD Jr. (1989) Infantile spasms: a disorder of the developing nervous system. In Kellaway P, Noebels JL (Eds) Problems and concepts in developmental neurophysiology. John Hopkins University Press, Baltimore, pp. 131–147. Huguenard JR. (2000) Circuit mechanisms of spike-wave discharge: are there similar underpinnings for centrotemporal spikes? Epilepsia 41:1076–1077. Janz D. (1994) Juvenile myoclonic epilepsy with typical absences. In Duncan JS, Panayiotopoulos CP (Eds) Typical absences and related syndromes. Churchill, London, pp. 174–183. Jasper H, Droogleveer-Fortuyn J. (1947) Experimental studies of the functional anatomy of the petit mal epilepsy. Assoc Res Nerv Ment Dis Proc 26:272–298. Kellaway P. (1959) Neurologic status of patients with hypsarrhythmia. In Gibbs FA (Ed.) Molecules and mental health. Lippincott, Philadelphia, pp. 134–149. Kellaway P. (2000) The electroencephalographic features of benign centrotemporal (rolandic) epilepsy of childhood. Epilepsia 41:1053–1056. Lado FA, Moshe SL. (2002) Role of subcortical structures in the pathogenesis of infantile spasms: what are possible subcortical mediators? Int Rev Neurobiol 49:115–140. Epilepsia, 53(5):771–778, 2012 doi: 10.1111/j.1528-1167.2012.03462.x

778 G. Avanzini et al. Lopes da Silva F, Blanes W, Kalitzin SN, Parra J, Suffczynski P, Velis DN. (2003) Epilepsies as dynamical diseases of brain systems: basic models of the transition between normal and epileptic activity. Epilepsia 44(Suppl. 12):72–83. Lytton WW. (2008) Computer modelling of epilepsy. Nat Rev Neurosci 8:626–637. Magoun H. (1963) The waking brain. Charles C Thomas, Springfield, IL. Manganotti P, Zanette G. (2000) Contribution of motor cortex in generation of evoked spikes in patients with benign rolandic epilepsy. Clin Neurophysiol 111:964–974. Manganotti P, Miniussi C, Santorum E, Tinazzi M, Bonato C, Polo A, Marzi CA, Fiaschi A, Dalla Bernardina B, Zanette G. (1998) Scalp topography and source analysis of interictal spontaneous spikes and evoked spikes by digital stimulation in benign rolandic epilepsy. Electroencephalogr Clin Neurophysiol 107:18–26. Manganotti P, Bongiovanni LG, Zanette G, Fiaschi A. (2000) Early and late intracortical inhibition in juvenile myoclonic epilepsy. Epilepsia 41(Suppl. 9):1129–1138. Manganotti P, Tamburin S, Bongiovanni LG, Zanette G, Fiaschi A. (2004) Motor responses to afferent stimulation in juvenile myoclonic epilepsy. Epilepsia 45(Suppl. 1):77–80. Manganotti P, Bongiovanni LG, Fuggetta G, Zanette G, Fiaschi A. (2006) Effects of sleep deprivation on cortical excitability in patients affected by juvenile myoclonic epilepsy: a combined transcranial magnetic stimulation and EEG study. J Neurol Neurosurg Psychiatry 77:56–60. Massimini M, Ferrarelli F, Esser SK, Riedner BA, Huber R, Murphy M, Peterson MJ, Tononi G. (2007) Triggering sleep slow waves by transcranial magnetic stimulation. Proc Natl Acad Sci U S A 104:8496– 8501. Mayer TA, Schroeder F, May TW, Wolf PT. (2006) Perioral reflex myoclonias: a controlled study in patients with JME and focal epilepsies. Epilepsia 47:1059–1067. Meeren HK, Pijn JP, Van Luijtelaar EL, Coenen AM, Lopes da Silva FH. (2002) Cortical focus drives widespread corticothalamic networks during spontaneous absence seizures in rats. J Neurosci 22: 1480–1495. Moeller F, Siebner HR, Wolff S, Muhle H, Granert O, Jansen O, Stephani U, Siniatchkin M. (2009b) Mapping brain activity on the verge of a photically induced generalized tonic–clonic seizure. Epilepsia 50:1632–1637. Moeller F, Maneshi M, Pittau F, Gholipour T, Bellec P, Dubeau F, Grova C, Gotman J. (2011) Functional connectivity in patients with idiopathic generalized epilepsy. Epilepsia 52:515–522. Moruzzi G, Magoun H. (1949) Brain stem reticular formation and activation of the EEG. EEG Clin Neurophysiol 1:455–473. Nalin A, Facchinetti F, Galli V, Petraglia F, Storchi R, Genazzani AR. (1985) Reduced ACTH content in cerebrospinal fluid of children affected by cryptogenic infantile spasms with hypsarrhythmia. Epilepsia 26(Suppl. 5):446–449. Nobili L, Baglietto MG, Beelke M, De Carli F, De Negri E, Rosadini G, De Negri M, Ferrillo F. (1999) Modulation of sleep interictal epileptiform discharges in partial epilepsy of childhood. Clin Neurophysiol 110:839–845. Panayiotopoulos CP, Obeid T, Waheed G. (1989) Absences in juvenile myoclonic epilepsy: a clinical video-EEG study. Ann Neurol 25:391– 397. Panzica F, Franceschetti S, Binelli S, Canafoglia L, Granata T, Avanzini G. (1999) Spectral properties of EEG fast activity ictal discharges associated with infantile spasms. Clin Neurophysiol 110:593–603. Panzica F, Rubboli G, Franceschetti S, Avanzini G, Meletti S, Pozzi A, Tassinari CA. (2001) Cortical myoclonus in Janz syndrome. Clin Neurophysiol 112:1803–1809. Parra J, Kalitzin SN, Iriarte J, Blanes W, Velis DN, Lopes da Silva FH. (2003) Gamma-band phase clustering and photosensitivity: is there an

Epilepsia, 53(5):771–778, 2012 doi: 10.1111/j.1528-1167.2012.03462.x

underlying mechanism common to photosensitive epilepsy and visual perception? Brain 126:1164–1172. Penfield W, Jasper HH. (1954) Epilepsy and the functional anatomy of the human brain. Little, Boston. Plum F, Posner J. (1980) The diagnosis of stupor and coma. Davis, Philadelphia. Polack PO, Guillemain I, Hu E, Deransart C, Depaulis A, Charpier S. (2007) Deep layer somatosensory cortical neurons initiate spike-andwave discharges in a genetic model of absence seizures. J Neurosci 27:6590–6599. Rechtin E. (2000) Systems architecting of organizations: why eagles can’t swim. CRC Press, Boca Raton, FL. Roelfsema PR, Engel AK, Kçnig P, Singer W. (1997) Visuomotor integration is associated with zero time-lag synchronization among cortical areas. Nature 385:157–161. Salek-Haddadi A, Mayer T, Hamandi K, Symms M, Josephs O, Fluegel D, Woermann F, Richardson MP, Noppeney U, Wolf P, Koepp MJ. (2009) Imaging seizure activity: a combined EEG/EMG-fMRI study in reading epilepsy. Epilepsia 50:256–264. Scantlebury MH, Galanopoulou AS, Chudomelova L, Raffo E, Betancourth D, Mosh SL. (2010) A model of symptomatic infantile spasm syndrome. Neurobiol Dis 37:604–612. Singer W, Gray CM. (1995) Visual feature integration and the temporal correlation hypothesis. Ann Rev Neurosci 18:555–586. Siniatchkin M, Groppa S, Jerosch B, Muhle H, Kurth C, Shepherd AJ, Siebner H, Stephani U. (2007a) Spreading photoparoxysmal EEG response is associated with an abnormal cortical excitability pattern. Brain 130:78–87. Siniatchkin M, van Baalen A, Jacobs J, Moeller F, Moehering J, Boor R, Wollf S, Jansen O, Stephani U. (2007b) Different neuronal networks are associated with spikes and slow activity in hypsarrhythmia. Epilepsia 48(Suppl. 12):2312–2321. Swann JW, Moshe SL. (2012) On the basic mechanisms of infantile spasms. In Noebels JL, Avoli M, Rogawski MA, Olsen RW, DelgadoEscueta AV (Eds) Jasper’s basic mechanisms of the epilepsies. Oxford University Press, New York, pp. 272–285. Terzano MG, Parrino L, Anelli S, Boselli M, Clemens B. (1992) Effects of generalized interictal EEG discharges on sleep stability: assessment by means of cyclic alternating pattern. Epilepsia 33:317–326. Tominaga I, Yanai K, Kashima H, Kato Y, Sekiyama S, Yokochi A, Miura I. (1986) An anatomo-clinical case of sequelae of acute encephalopathy. Infantile spasm with hypsarrhythmia. Rev Neurol 142:524–529. Tucker JS, Solitare GB. (1963) Infantile myoclonic spasms. Clinical, electrographic and neuropathologic observations. Epilepsia 4:45–59. Varela F, Lachaux JP, Rodriguez E, Martinerie J. (2001) The brainweb: phase synchronization and large-scale integration. Nat Rev Neurosci 2:229–239. Vining EP. (1990) Chaos, balance and development: thoughts on selected childhood epilepsy syndromes. Epilepsia 31(Suppl. 3):30–36. Vollmar C, O’Muircheartaigh J, Barker GJ, Symms MR, Thompson P, Kumari V, Duncan JS, Janz D, Richardson MP, Koepp MJ. (2011) Motor system hyperconnectivity in juvenile myoclonic epilepsy: a cognitive functional magnetic resonance imaging study. Brain 134:1710–1719. Wolf P. (1992) Reading epilepsy. In Roger J, Bureau M, Dravet Ch, Dreifuss FE, Perret A, Wolf P (Eds) Epileptic syndromes in infancy, childhood and adolescence. 2nd ed. J. Libbey, London, pp. 281–298. Wolf P. (2006) Basic principles of the ILAE syndrome classification. Epilepsy Res 70(Suppl. 1):20–26. Wolf P, Goosses R. (1986) Relation of photosensitivity to epileptic syndromes. J Neurol Neurosurg Psychiatr 49:1386–1391. Ziemann U, Rothwell JC, Ridding MC. (1996) Interaction between intracortical inhibition and facilitation in human motor cortex. J Physiol 496:873–881.