Address correspondence to Hans O. Lüders, Epilepsy Center, Neurological Institute, Case Medical Center University Hospitals, 11100 Euclid Avenue, Cleveland, OH 44106, U.S.A. E-mail: Hans.Luders@UHhospitals.org
In this commentary we discuss the basic concept of an epileptogenicity level, which is variable in different brain regions and is a function of multiple factors including the basic epileptogenicity level, routine environmental or internal stimuli, and various triggering and causative factors. This concept blurs the distinction between focal versus generalized and between idiopathic versus symptomatic epilepsies. We suggest dropping the dichotomy idiopathic versus symptomatic and to instead simply list the different etiologic factors that increase the epileptogenicity level. On the other hand, even if there is a continuum between focal and generalized epilepsies, most epilepsies are either predominantly focal or predominantly generalized. It is useful to maintain this distinction (even if somewhat artificial) because focal epilepsies can be treated with epilepsy surgery, and all focal epilepsies tend to respond to the same type of anticonvulsants. Generalized epilepsies cannot be treated surgically and respond to different anticonvulsants depending on the semiologic type of seizures.
For more than a century epileptologists have subdivided epilepsies into two broad dichotomies, namely, generalized versus focal epilepsies and again into idiopathic and symptomatic epilepsies (Jackson, 1870). Initially this subdivision was based primarily on clinical observations (ictal semiology, clinically evident brain pathology), and lately the aide of sophisticated diagnostic techniques has greatly improved our ability to accurately classify the epilepsies according to these dichotomies (Commission on Classification and Terminology of the International League Against Epilepsy, 1989).
At the same time, more accurate diagnosis led epileptologists to the understanding that instead of dichotomies there is a continuum for the variables subdivided strictly in dichotomies in older classifications (Lüders et al., 1984).
In this commentary we present first our conceptual understanding of epilepsies and then critically analyze two manuscripts published in this issue of Epilepsia, which discuss the aforementioned dichotomies (Capovilla et al., 2009 [Correction added after online publication 14 May 2009: Avanzini, 2009 changed to Capovilla et al., 2009]; Rodin, 2009).
Epilepsies: Essential Concepts
Any classification of the epilepsies must be based on some essentials concepts. Figs. 1–6 illustrate the basic facts and concepts underlying the classification scheme we explain below.
1Epileptic seizures are a response of the brain to a large variety of stimuli. Epileptic seizures can be elicited in any brain, including apparently “normal brains,” as long as a sufficiently strong “epileptogenic” stimulus is applied. The typical example is epileptic seizures produced by electroshock, which is invariably successful in all patients as long as a sufficiently strong current is delivered.
2Therefore, the occurrence of an epileptic seizure is not necessarily a pathologic condition. The pathologic condition is the ease with which in any given individual seizures can be elicited.
3In this commentary we conceive epilepsy, defined as an abnormal tendency to produce epileptic seizures by one or more regions of cortex, as the interplay of several factors (see Figs. 1–6). We assume first that each region of the brain has a certain epileptogenicity level that is determined by the following factors:
aBasic epileptogenicity level determined by the inherited tendency to generate seizures. This is a function of the genetic constitution of an individual, with a number of genes playing a crucial role. We expect that in the future some of these crucial genes can be identified, giving us a more objective concept of the basic epileptogenicity level.
bThe basic epileptogenicity level is changing continuously, influenced by a variety of “routine” environmental factors, which either increase of decrease the basic epileptogenicity level. This is illustrated in the figures by representing the epileptogenicity level of any given brain region as a band, assuming that the epileptogenicity level is continuingly changing following a “Gaussian function over time.”
cIn addition, the epileptogenicity level may be altered greatly by triggering factors, such as intermittent photic stimulation, hyperventilation, injection of Metrazol, and electroshock. These are environmental factors that tend to occur infrequently or sporadic, but may greatly influence the epileptogenicity level. These factors can be easily identified and not infrequently may trigger clinical seizures.
dFinally, we have also causative factors. These are genetic factors or environmental factors that produce a lasting change of the epileptogenicity level of one or more regions of the brain, leading to “chronic” epilepsy. These include factors such as brain tumors, mesial temporal sclerosis, malformations of cortical development, and monogenetic epilepsies.
eThe diagrams shown below (Figs. 1–6) do not include the concept of “seizure spread.” However, most electroencephalographic seizures do not become symptomatic unless significant seizure spread occurs. At the same time, it is clear that different epilepsies have a different tendency to spread, with generalized epilepsies or epilepsies with frequent secondary generalization having a much higher tendency for seizure spread. In other words, it appears that the tendency of any focal seizure discharge to spread to another brain region is a function of the “strength of the seizure discharge,” but also of the epileptogenicity level of that brain region into which the seizure spreads (Lüders et al., 1984).
4Fig. 1 shows the epileptogenicity level for two normal individuals (one with a relatively low level of epileptogenicity for all brain regions and another individual with a relatively high level of epileptogenicity for all brain regions) and a patient with generalized epilepsy. Notice that in the normal individuals the epileptogenicity level never reaches the threshold necessary to trigger an epileptic seizure. On the other hand, in individuals with generalized epilepsy the epileptogenicity level can exceed the epilepsy threshold in several brain regions and, therefore, will trigger a seizure from time to time. This figure also illustrates the well-recognized fact that in generalized epilepsies different brain regions may trigger generalized seizures at different times (Lüders et al., 1984). The observation in patients with generalized epilepsy of short, relatively focal epileptiform discharges arising from different brain regions at different times supports this conclusion (Lüders et al., 1984). It is also obvious from this diagram that an epileptogenic trigger (for example the ingestion of Wellbutrin) can trigger an epileptic seizure in the “normal” individual with a relatively high level of epileptogenicity but not in the normal individual with a relatively low level of epileptogenicity.
5Fig. 2 shows the epileptogenicity level of an individual with left occipital focal epilepsy and of an individual with generalized epilepsy. The individual with generalized epilepsy has a relatively high epileptogenicity level in all brain regions and exceeds the seizure threshold in several brain regions. On the other hand, the individual with focal epilepsy has a relatively low level of epileptogenicity in all brain regions except the left occipital area.
6Fig. 3 shows an individual before receiving electroshock (normal epileptogenic levels at all brain areas) and during electroshock applied to both frontal regions (epileptogenic level exceeds seizure threshold in the frontal areas eliciting a clinical seizure). The graph for the patient with focal epilepsy is very similar, but in that case the focal epilepsy produces a permanent rise of the level of epileptogenicity in a specific area of the brain.
7Fig. 4 illustrates a patient with a focal epileptic status. Notice that in that case the entire broad band of epileptogenicity is above the seizure threshold level, implying that the patient is going to have continuous seizures. In this case the seizure frequency is simply a function of the duration of the postictal depression, for example, the longer the postictal depression the lower the seizure frequency.
8Fig. 5 illustrates the interplay of the “basic epileptogenicity level” with “triggering factors.” In these examples, there are two individuals who have the same tumor, which increases the epileptogenicity level by approximately the same amount. However, only the individual with a relatively higher “basic epileptogenicity level” will have seizures.
9Fig. 6 shows the changing epileptogenicity level in a patient with benign epilepsy of childhood. This illustrated the changing level of epileptogenicity as a maturational factor. Note that the epileptogenicity level exceeds the seizure threshold in the occipital region at age 2 years. At age 8 years the epileptogenicity level exceeds the seizure threshold in the central area, and finally at age 20 years the epileptogenicity level is normal in all brain areas.
Essential Concepts and the Dichotomies
The manner in which epilepsy is conceived in the previous paragraphs has a direct impact on our understanding of the basic two dichotomies discussed in this issue of Epilepsia.
1Generalized epilepsies are characterized by a relatively high level of “basic epileptogenicity,” which affects all brain regions to approximately the same degree. Focal epilepsy, at the other extreme, is characterized by a relatively low “basic epileptogenicity” in all brain regions and a “causative factor” that increases the epileptogenicity selectively in a given brain region. Between these two extremes, however, we can find all gradations. In other words, this is not a dichotomy but a continuum between the two extremes, which are defined as the “ideal” generalized and the “ideal” focal epilepsy.
2All epilepsies are determined by a multiplicity of causes including numerous genetic factors, several triggering factors, and a variable number of causative factors. Again the subdivision idiopathic versus symptomatic epilepsies is an abstract concept in which “idiopathic” refers to epilepsies caused primarily by genetic factors and “symptomatic” points to epilepsies that have clear “causative factors.” However, as noted earlier, all epilepsies actually include a mixture of genetic, triggering, and causative factors. Modern epileptology allows us now to elucidate with progressively more precision what these factors are.
Commentary on Use of Dichotomies in Modern Epileptology
The preceding discussion supports the notion that instead of dichotomies we actually have a continuum between generalized and focal and between idiopathic and symptomatic epilepsies, and that each individual is placed at a different position within both of these continua. How do these concepts affect the use of dichotomies discussed in the articles included in this issue of Epilepsia? (Capovilla et al., 2009 [Correction added after online publication 14 May 2009: Avanzini, 2009 changed to Capovilla et al., 2009]; Rodin, 2009)
1Fortunately an analysis of seizure semiology and interictal and/or ictal electroencephalography (EEG) usually allows clear distinction between generalized and focal epilepsies. Only a small minority of patients fall in the gray zone between focal and generalized epilepsies. Therefore, we feel it would be useful to maintain these subdivisions because the classification in generalized versus focal epilepsy has a direct impact on management decisions:
aPatients with focal epilepsy are frequently good candidates for epilepsy surgery. Patients with generalized epilepsy, however, cannot be managed with excisional epilepsy surgery.
bAll patients with focal epilepsy are treated with the same type of antiepileptics. Patients with generalized epilepsy, however, respond differently to different antiepileptics, mainly as a function of the type of seizures (for example, patients with generalized myoclonic seizures respond to different antiepileptics than patients with “absence” seizures).
2On the other hand, the dichotomy idiopathic versus symptomatic is of very limited practical use. We know that in most cases the main etiologic or modifying epilepsy factors are multifactorial. A subdivision of these epilepsy etiologic or modifying factors in four types, as suggested by Dr. Berg and collaborators, is of no practical use (Capovilla et al., 2009 [Correction added after online publication 14 May 2009: Avanzini, 2009 changed to Capovilla et al., 2009]). That is the reason that we propose to simply list the known genetic factors, the triggering factors, and the “causative factors.” We expect that in the near future our understanding of the influence of genetic factors on the epileptogenicity level in different brain regions will greatly expand. With that expansion of knowledge it will become clear that all patients with epilepsy have multifactorial factors that lead to epileptic seizures. In the future we expect also that we will be able to counteract many of the genetic factors that contribute to the generation of seizures, more or less in the same way as today we can take measures against the many “causative factors” (which usually are elucidated by high-resolution neuroimaging). We are at the start of a revolution and we should have the foresight to give up old concepts and replace them with new approaches that accommodate these new ideas.
Generalized versus Focal Epilepsies
The preceding discussion of essential concepts clarifies that there is no absolute distinction between generalized and focal epilepsies, but rather a continuum. The main distinction, as shown in Figs. 1 and 2 is that in generalized epilepsies many brain regions have a tendency to reach an epileptogenicity level that exceeds the seizure threshold. Implicit in this concept is also the increased tendency of generalized epilepsies to secondarily generalize, that is, generalized epilepsies may start from many cortical regions and have a markedly increased tendency to secondarily generalize. On the other hand, in focal epilepsies only one brain region has a tendency to exceed the seizure threshold and the tendency to secondary generalization is significantly less, that is, the initial seizure symptomatology remains “focal” and not infrequently there is no secondary generalization. This concept is also discussed in the article of Dr. Rodin, who also feels that seizure spread is an essential difference (Rodin, 2009).
Avancini et al. suggest the expression of “system epilepsy” for the benign focal epilepsies. It is generally accepted that the benign focal epilepsies occupy a special position because the epileptogenicity level in different brain regions varies with neuronal maturation, that is, the epileptogenicity level exceeds the seizure threshold level at different times at different brain regions (Fig. 6). We do not see, however, any need to assume that these epilepsies are different from other focal epilepsies with respect to their involvement of subcortical “systems.” All epileptic discharges, generalized or focal, will recruit subcortical systems. For example, the majority of focal spikes and all epileptic seizures in patients with “focal” frontal epilepsy are reflected in the subthalamic nucleus (Dinner et al., 2001). There is no doubt that many other subcortical systems—in the thalamus, the basal ganglia, the brainstem, and the spinal cord—will participate in interictal and particularly in ictal discharges. In that sense all the epilepsies are “system epilepsies.” Surgery of epilepsy, however, has taught us that despite this extensive “system” involvement, most focal epilepsies can be eradicated if we simply remove sufficient cortical tissue. In other words, the essential brain tissue to sustain the epilepsy tendency is the abnormal cortical “focus” and not the participating subcortical structures. The peculiarity of the benign focal epilepsies is not the involvement of subcortical structures (common to all focal epilepsies) but that the epileptogenic focus “migrates” with maturation.
It is also clear that generalized epilepsies can have different phenotypic expressions, ranging from focal seizures to different types of generalized seizures. It is also true that in the different generalized seizures not all the cortical neurons are involved and there are variable contributions of many cortical and subcortical neuronal pools. That is also correct for focal epilepsies in which different discharges may involve different set of neurons. Almost never all the neurons in a given region involved with the seizure discharge. Furthermore, the different discharges may affect most neurons, but in only a fraction of the neurons is it likely that the discharge is of sufficient strength to reach threshold and generate an axon potential.
We certainly need a better understanding of the pathophysiology that leads to different generalized epileptic seizure types and to different types of focal epilepsies, even when the initial discharge originates from the same location. However, simply coining a term and talking about system epilepsies will not be necessarily of any help. Besides, talking about system epilepsies would imply that one or another system (motor system, sensory system, autonomic system, and so on) is almost exclusively involved in any given seizure type. It is very unlikely that that is actually true. During most epileptic seizures many “systems” will be involved simultaneously even if the main symptomatology suggests the preferential involvement of only one “system.”
Classification of Epilepsies and Epileptic Syndromes
The new ILAE proposal of classification of the epilepsies emphasized the identification of numerous epileptic syndromes (Dinner et al., 2001; Engel, 2001). Proponents of a detailed syndromatic classification of the epilepsies argue that such a classification is useful in the management of the epilepsies (predicts which anticonvulsants will be most effective) and helps in the prognosis. In this issue of Epilepsia, Dr. Rodin points out that severity of illness (as specified by duration of illness, frequency of seizures, number of seizure types, amount of EEG abnormalities, and so on) is the key prognostic factor and that the current syndromes that carry a bad prognosis are actually those that include variables that point to a higher degree of illness severity (Rodin, 2009). We certainly agree with Dr. Rodin’s approach. There now is a tendency to subdivide the epilepsies into syndromes that at best are a set of symptoms, signs, and test results that tend to occur together. As a second step we investigate the antiepileptics that have the greatest beneficial effect in a given syndrome. In addition, the prognosis for any given syndrome is evaluated. The results of these studies greatly depend on the definitions of the different syndromes, which as we mentioned earlier are defined by relatively arbitrary criteria. On the other hand, Dr Rodin’s approach is scientifically sounder. He isolates numerous variables and then establishes statistical correlations with prognosis.
Epileptology has made great advances over the last decades. EEG and video-EEG evaluations allow us now to define with precision the seizures a patient is experiencing. In many case we can differentiate between clearly focal seizures and generalized seizures, an essential subdivision to decide if a patient is a surgical candidate and to assess the type of anticonvulsants that may be most effective. We also understand that there is a continuum between focal and generalized epilepsies and that in some cases patients fall in a gray zone in the transition from focal to generalized epilepsy.
Modern epileptology also understands that epileptic seizures are the consequences of a basic epileptogenetic tendency (genes), various more or less obvious triggering factors, and major causative factors. For any given patient it is essential that we clearly identify these factors. This will help us in the management. However, there is absolutely no value in classifying patients according to different types depending on the “type” of etiologic factors responsible for the epileptic seizures (Capovilla et al., 2009 [Correction added after online publication 14 May 2009: Avanzini, 2009 changed to Capovilla et al., 2009]).
The concepts illustrated in this commentary form the theoretical backbone for the epilepsy classification we have proposed in previous publications (Lüders, 2008).
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.
Disclosure: None of the authors has any conflict of interest to disclose.