Evidence of Subtle Structural Abnormalities in Idiopathic Generalized Epilepsy

abstract & commentary

Source: Woermann FG, et al. Quantitative MRI in patients with idiopathic generalized epilepsy—Evidence of widespread cerebral structural changes. Brain 1998;121:1661-1667.

In contrast to focal epilepsies that are associ-ated with structural brain lesions, such as areas of cortical dysplasia, scarring or hippocampal sclerosis, idiopathic generalized epilepsies (IGE) are believed to arise from largely unclear genetic factors. In fact, the International League Against Epilepsy (ILAE) definition of idiopathic generalized epilepsies specifies that neuroradiological abnormalities should be absent in these patients (Commission of Classification and Terminology of the ILAE, Epilepsia 1985;26:268-278). Histopathological studies of patients with IGE, though rare, have identified microscopic structural changes termed microdysgenesis that consists mainly of excess neurons in the molecular layer of cortex (Meencke HJ. Epilepsia 1985;26:250-254). Increased numbers of ectopic neurons in the subpial space and in the subcortical white matter have been described (Meecke HJ, Janz D. Epilepsia 1984;25:8-21). The number of neuropathological studies in patients with IGE is small, however, and the finding of microdysgenesis associated with IGE is not without controversy (Lyon G, Gastaut H. Epilepsia 1985;26:365-367).

Woermann and colleagues bring quantitative MRI to bear with the question of whether subtle structural changes are present in the brains of patients with IGE. With the ability to image a volume less than a millimeter cubed, Woermann et al applied software methods to the analysis of the MRI images to measure cortical gray matter volume separately from the combined subcortical white and gray matter volume. Comparing a population of 45 patients with IGE, which was comprised of 20 patients with juvenile myoclonic epilepsy, 10 patients with childhood absence epilepsy, 10 patients with juvenile absence epilepsy, and five patients with tonic-clonic seizures on awakening, Woermann et al found mean total cerebral volume and the mean normalized cortical volume (expressed as a percentage of total brain volume) did not differ significantly from a population of 30 normal controls.

When regional differences of cortical and subcortical volumes were compared, however, structurally abnormal brains were detected in 15 of the 45 patients. To accomplish the identification of regional abnormalities, Woermann et al divided the cerebral volume of patients and controls into 10 slices of equal thickness arranged from anterior to posterior. Each slice consisted of equivalent portions of the right and left hemispheres, so that a total of 20 volumes of interest, 10 slices times two hemispheres were used. Volumes of interest were all normalized as a percentage of the total cerebral volume to permit comparison between individuals. The regional cortical and subcortical normalized volumes were extracted and used to compute several additional parameters that identified subtle variations in anatomy between patients and controls. A total of 80 different regional measurements were calculated for each patient or control. These measurements consisted of the normalized cortical and subcortical volumes in each volume of interest, the ratios of ipsilateral cortical to subcortical matter in volumes of interest, and the ratios of ipsilateral to contralateral volumes of homologous volumes of interest (e.g., ipsilateral cortical volume to contralateral cortical volume). The normal range was defined as the range of values within three standard deviations of the mean value of each measurement in the control population. Three of the 30 individuals in the control group had one abnormal measurement, while the rest of the control group had no abnormalities. Since three individuals in the control group had one abnormal measurement, the presence of two or more abnormal measurements was required to identify a structurally abnormal brain.

In the group of 15 IGE patients who met the criteria for structurally abnormal brains, there were an average of 5.4 abnormal measurements per patient. By epilepsy subtypes, eight of the identified patients had juvenile myoclonic epilepsy, four had juvenile absence epilepsy, one had childhood absence epilepsy, and two had tonic-clonic seizures on awakening. The majority of abnormalties were found in the central volumes of interest. Half of the abnormal measurements resulted from an increase in the ratio of the regional cortical volume to the ipsilateral subcortical volume. The distribution of regional asymmetries accounting for the remaining abnormal measurements is not reported.

Woermann et al interpret the increase in the regional ratio of cortical volume to subcortical volume by suggesting altered interneuronal connectivity in the cortex of patients with IGE. They argue that an area of abnormal cortex may contain increased neurons, neuronal volumes, or neuropil, or may project fewer or smaller axons to the subcortical volume. Either scenario, it is argued, will result in altered connections between neurons. Cortical hyperexcitability has been demonstrated in patients with IGE (see citations given by Woermann et al). Woermann et al go on to suggest that microdysgenesis may be the origin of the regional structural asymmetries detected by their methods and may in turn result in cortical hyperexcitability.


This study brings forward several interesting findings. If one takes the view that cortical development is a largely radially symmetrical process guided by global genetic influences, the presence of regional heterogenerities and asymmetries in the cortical and subcortical matter detected by MRI is somewhat unexpected. The observation that the majority of abnormalities were found in the central cortical and subcortical area may help identify the genetic trigger responsible for IGE. For example, it is possible that the abnormality in IGE may originate in the genetic factors guiding local cortical differentiation, such as the creation of the Betz cells in layer V of the motor cortex. It is also interesting that 10% of normals had one regional abnormality. While it is possible that the detection of abnormalities in normals results from statistical noise, it is provocative to consider that IGE might represent the extreme of a spectrum of genetic traits.

The findings described by Woermann et al are likely to fuel the controversy between proponents and opponents of the microdysgenesis etiology of idiopathic generalized epilepsies. There are results to support both points of view. The presence of regional abnormalities in one-third of patients with IGE supports the tenet that genetic miswiring of the cortex may be a factor in idiopathic generalized epilepsy. However, the presence of regional abnormalities in 10% of the controls, and the failure to detect a structurally abnormal brain in two-thirds of patients casts doubt on the possibility that microdysgenesis is the major cause of IGE. Of course, by altering their definition of the normal range, here taken to be values that lay within three standard deviations of the mean, it may be possible to detect more abnormalities in the patient population. It is likely, however, that a more stringent definition of the normal range would increase the detection of abnormalities in the normal controls.

If one accepts the proposition that microdysgenesis is a cause of IGE, structurally abnormal cortex or subcortical regions should also cause abnormal cortical and subcortical function and changes in the EEG background of IGE patients. Abnormal background activity of EEG is the expected result of dysfunction of the cortex or subcortical white or gray matter. One of the hallmarks, however, of idiopathic generalized epilepsies is that the background EEG is normal. (Daly, Pedley. Current Practice of Clinical Electroencephalography, New York: Raven Press; 1990). The presence of regional slowing of the background, as might be expected from a region of cortical microdysplasia, is not seen in patients with IGE. Generalized background slowing, a feature of dysfunction in the subcortical gray matter, is also not typically seen in patients with IGE. Clinically, the patients with IGE also usually develop normal intellect and neurologic function (Wyllie. The Treatment of Epilepsy, Baltimore: Williams and Wilkins; 1997), which might not be expected in patients with a congenitally dysfunctional cortex.

As with any thoughtful work, the results of Woermann et al raise more questions than they answer. Nevertheless, in bringing the modern tools of MRI and computer analysis to the problems of neuropathology, Woermann et al have devised an alternative to, though not a substitute for, autopsy studies. One may hope that in the future, methods will be developed that may provide better radiographic analysis of the brain. Nevertheless, it is equally clear that there is an ongoing need for autopsy specimens for modern pathologic study. It is also premature to settle on microdysgenesis as the cause of the idiopathic generalized epilepsies. The etiology of idiopathic generalized epilepsies remains open and will likely require the contributions of electrophysiology, neuroradiology, and neuropathology before it can be satisfactorily answered. (Dr. Lado is an EEG Fellow, Department of Neurology, Montefiore Medical Center-Albert Einstein College of Medicine, Bronx, NY. Dr. Moshe is Professor and Director, Pediatric Neurology and Clinical Neurophysiology, Department of Neurology, Montefiore Medical Center-Albert Einstein College of Medicine, Bronx, NY.)