Reply to “Of Cabbages and Kings: Some Considerations on Classifications, Diagnostic Schemes, Semiology, and Concepts”

We read with interest the contribution of Prof. Peter Wolf 1. We certainly agree with his view that any new proposal for classification of epileptic syndromes should be based on clear principles, and that the first task of the ILAE Committee should be the definition of the principles they will use to design a new classification system of epileptic seizures and epileptic syndromes. Indeed, in the recent report of the ILAE Task Force on Classification and Terminology, headed by Prof. Engel 2, a list of accepted epileptic seizures and epileptic syndromes is presented. However, there is no definition of what he means by accepted epileptic seizure or epileptic syndrome and it is also unclear what Committee or Task Force accepted these entities. We also agree with Prof. Wolf that the discussion of seizure types is confusing. In the recent report of the ILAE Task Force 2, they indicate that new scientific advances now “permit creation of a list of seizure types that represent diagnostic entities, as opposed to phenomenological descriptions, based on known or presumed common anatomy and pathophysiology. Such diagnostic entities, like syndromes would have etiological, therapeutic and prognostic implications, and could be used to supplement syndromatic diagnosis, or stand alone when syndromic diagnoses cannot be made.” In previous publications, we stressed the importance of clear differentiation between the classification of epileptic seizures and epileptic syndromes. We suggested that a clear distinction can be achieved by making a pure semiological (phenomenological) classification of the epileptic seizures (3–10) that contrasts clearly with a more pathophysiological classification of the epileptic syndromes based on the complete clinical and diagnostic picture. The semiological classification system has the advantage of placing the diagnostic emphasis on the epileptic syndrome. The semiological classification of seizures is similar to the systematic specification of a semiological neurological sign (for example, resting tremor), which is only one tool to define the underlying disease (for example, Parkinson's Disease). There is no room for confusion between the classification of the symptomatology (neurological signs or semiological type of epileptic seizures) and the underlying disease or syndrome. Independent classification of semiological seizure types and epileptic syndromes permits also systematic studies of the occurrence of different semiological seizure types in the various epileptic syndromes and vice versa of the different epileptic syndromes that can be associated with any given semiological seizure type. The 5-axis approach suggested by the Task Force adds unnecessary complexity and redundancy to the classification of the epilepsies. Axis 1 uses a “glossary” to replace the systematic semiological seizure classification that we have developed and tested in thousands of people with epilepsy in our and other centers. The proposed approach then adds a pathophysiological classification of seizures, whose terminology is very similar to the terminology used in the glossary. Additional complexity is added by a fourth syndromatic axis. This redundancy and complexity of this scheme make it extremely unlikely that such a system will eventually be used by non-epileptologists. For example, in the glossary the ILAE Committee defines the terms myoclonic, astatic and atonic 11. In axis 2, Prof. Engel lists distinctive seizure types labeled as myoclonic seizures, atonic seizures, and myoclonic atonic seizures 2. Then on the syndromatic level, Prof. Engel identifies an epilepsy syndrome with myoclonic-astatic seizures. 2 Would it not be preferable to adopt a system similar to the classification systems universally used in clinical neurology? At our institution, as at other epilepsy centers around the world, we successfully have used such an approach, to classify the patient with epilepsy by: 1. defining the seizures semiologically (equivalent to defining the neurological signs or symptoms a patient has); 2. determining the etiology of the epilepsy; and 3. identifying the location of the epileptogenic zone (equivalent to defining the location of the neurological lesion). This is then complemented by describing some essential clinical features (like mental retardation, etc.) and in selected cases by defining an “epileptic syndrome.” However, we believe that most of the currently described “epileptic syndromes” are redundant. These syndromes contain artificial data that are not useful beyond the clinical picture for making the diagnosis of epilepsy, or at the least are too complex. We need a critical review of all the epileptic syndromes, in order to crystallize those that are useful in everyday practice, to add meaningful information to the existing three-axis classification by semiological seizures, etiology, and location of the epileptogenic zone. We have started such a study and are testing the system by applying it retrospectively to thousands of patients seen at the Cleveland Clinic over the last twenty years. We agree with Prof. Wolf that the semiological seizure classification is a “gardener's” practical classification system. The three-axis classification system defined above is also primarily a practical “gardener's” classification that, however, can be used very effectively as a guide for treatment and prognosis. In most patients, epileptic seizures are only epiphenomenona to another disease specified by the etiology in the three-axis classification outlined above. Creating special “epileptic syndromes” for these conditions is highly artificial. On the other hand, effective management of the epileptic seizures in these patients requires the definition of the etiology, the semiological seizure type, as also the location of the epileptogenic zone. These practical need justify such a three-axis classification system. We appreciate Prof. Engel's intention to have a general debate on the new development of the ILAE classification system. However, we would encourage the Task Force to again critically evaluate the procedure taking into consideration some of the following guidelines: Try to define a classification system that is as simple as possible, avoiding any redundancy. A 5-axis classification is too complex for the use by general neurologists. Define a classification system that is useful and applicable by epileptologists in technologically advanced Epilepsy Centers, and also by nonspecialists who have limited access to modern diagnostic technology. In the semiological seizure classification, this is achieved by defining a limited number of broad categories that can be easily applied by just taking the clinical history of the patient (for example by just specifying that a patient has motor seizures). On the other extreme, the semiological classification also allows detailed definition of the ictal semiology essential in advanced epilepsy surgery centers (for example, by classifying a motor seizure as a right hand clonic seizure → generalized tonic-clonic seizure). Try to define a classification system that is close to the system used to classify other neurological disease. This approach would greatly facilitate its acceptance by general neurologists. Take a fresh look at all the epileptic syndromes, retaining only those that add useful information to the three-axis classification described above. We believe that only a very small proportion of patients with epilepsy have “epileptic syndromes” that require special definition and provide practically useful additional information. Do not mix glossaries and classification systems. Glossaries should be an addendum to a classification system, not an axis. As pointed out by Prof. Wolf, the Task Force on Classification and Terminology should define with precision the criteria they used to establish different categories in each axis, making sure that there is no overlap (redundancy) between the categories. We agree with Prof. Wolf that the ideal classification is the scientific “botanist” classification. However, it is difficult to understand why Prof. Wolf concludes in his comments that a practical “gardener's” classification is not a classification (he uses this argument to dismiss the semiological seizure classification as a “guide for the bedside” or just an “arrangement”). We would encourage the Task Force on Classification and Terminology to actually look for a classification system that complies with the points specified here, whether a practical or scientific classification, or is a mixture of each. Do not publish preliminary data that can be interpreted as an approved classification system. Define a classification system that has been discussed in detail by all members of the Classification Committee, and apply it to a large patient population to test if indeed it is a practically useful system. Once this has been accomplished and the Task Force has had an opportunity to modify the proposed classification system after testing it “in the field,” it is time to publish the results.