A glimpse into abnormal cortical development and epileptogenesis at epilepsy surgery

Generations of neurologists have approached the bedside with the conviction that clinical observations not only serve the needs of patient care but can also provide important insights about the human nervous system in health and disease. With the widespread use of surgery for the treatment of the epilepsies during the last decade, epileptologists had similarly high expectations for insights from epilepsy surgery. There was the hope that the pathogenesis of various epilepsies might be explained by direct study of surgically resected human epileptogenic brain tissue. This hope was expressed at international conferences, in workshops, and in clinical conferences where surgical treatment was being planned and evaluated.^1,2 It was anticipated that direct physiologic and anatomic analysis of resected human brain would provide insights into cellular mechanisms of seizure generation. Despite the problems posed by lack of suitable control tissue, the opportunity for study of resected human epileptic tissue had the immediate appeal of direct relevance to human epilepsy and thus a potential advantage over animal models. It was also hoped that molecular studies in resected brain might eventually provide a framework for classification of epileptic disorders at genetic and molecular levels. As is often the case in scientific inquiry, the insights have been more elusive than the expectations.

In this issue of Neurology, a clinical-pathologic analysis of surgically resected brain from three patients with epilepsy associated with cortical dysplasia suggests that study of surgically resected human brain tissue may indeed provide insights into abnormalities of cortical development and epileptogenesis. This series of three cases, and ongoing studies in other epilepsy centers,³ demonstrate how advances in MRI, surgical therapy, and methods for characterization of inhibitory and excitatory neurons now allow detailed study of epileptogenic circuity associated with cortical dysplasias.

High-resolution MRI has now shown that "cryptogenic" of "idiopathic" epilepsy is often associated with regions of focal cortical dysplasia.⁴ Two of the three cases demonstrated the benefit of MRI studies for evaluation and treatment of patients with epilepsy. However, as illustrated by the one patient in this series with widespread cortical dysplasia that was undetected by MRI, MRI may miss subtle cortical dysplasia. Such subtle cortical dysplasias may be even more common than currently recognized and could play a role in many cases of milder partial epilepsies that are not treated by surgical intervention.

Cortical dysplasia, which is also known as cortical microdysgenesis, was originally described by Taylor et al.⁵ and is typically associated with extremely active epileptogenic EEG abnormalities.⁶ The regions of dysplasia are focal areas of abnormal cortical cytoarchitecture with variable degrees of altered cortical lamination and aberrantly located large, round neurons that have been referred to as "balloon" cells. Cortical dysplasias vary in severity and are more subtle than cerebral malformations such as lissencephaly, pachygyria, polymicrogyria, and laminar or nodular heterotopias but share the common feature that all are consequences of abnormal cortical development. Although these malformations are commonly regarded as neuronal migration disorders because neurons in regions of cortical dysplasia are found in abnormal locations, it is likely that abnormalities in neurogenesis, cell survival, and other defects in normal cortical developmental processes may also contribute to cortical dysplasias.

In resected cortex from each of the three patients, there were consistent patterns of alterations in the location of neurons that were immunoreactive for a variety of GABAergic markers and the glutamatergic NMDAR1 and GluR2-3 receptor subunits. The observation of reorganized neurotransmitter expression in regions of cortical dysplasia is of interest and follows increasing recognition that many critical events in early cortical development are influenced by neurotransmitter signaling. Many neurotransmitters, including GABA and glutamate, the principal transmitters of the adult cortex, are present in proliferating cells of the embryonic cortex. GABA (an excitatory transmitter during postnatal days 1 to 8 in the rodent) and glutamate increase inward currents and Ca²⁺ influx in cells in the ventricular zone and decrease the number of cells synthesizing DNA in rat embryonic cortex.⁷ GABA and glutamate also regulate outgrowth of neurites, neuronal survival, growth cone pathfinding, neuroblast movement, migration on radial glial cells, and synapse elimination.⁸ It is reasonable to suggest that a variety of pathologic and environmental factors affecting the embryo could alter neurotransmitter signaling and interfere with normal developmental processes, which could modify the structural and functional organization of cortex in the adult. Such interference was recently demonstrated in developing hamsters in which activation of NMDA receptors in the developing neopallidum induced heterotopias and intra-cortical arrest of migrating neurons.⁹ A central question for epilepsy research is how the reorganization of neurotransmitter expression in dysplastic neurons and circuits, which was demonstrated in these three patients and may have been caused activity-dependent alterations in neurotransmitter signaling during development, may contribute to electrographic and behavioral seizures.

How do regions of cortical dysplasia generate paroxysmal, synchronous neuronal burst discharges, and what factors play a role in the interictal-ictal transition that results in the unpredictable emergence of brief, epileptic seizures? These questions remain unanswered despite the advances by studies of neural circuitry in a variety of chronic models of epilepsy induced by kainic acid, pilocarpine, and kindling. It may be too much of expect that studies directed toward these questions in human epileptic cortex will be any more revealing. The initial immunocytochemical observations in these three patients, however, demonstrated consistent patterns of abnormality that suggest how cellular alteractions could generate paroxysmal synchronous epileptic activity. The patterns of cellular abnormalities in this highly epileptogenic human cortex from a well-characterized epileptic focus may therefore have more general significance.

Caution is needed in making inferences from a few cases. The patients had different ictal manifestations, had abnormal intellectual function suggesting diffuse cortical dysfunction, and had MRI studies that revealed variable spatial extent of cortical abnormalities. These observations suggest that numerous widespread abnormalities may be contributing to seizure generation. Nevertheless, the regions of cortical dysplasia closely associated with the site of origin of recorded electrographic seizures in these cases revealed common features. There was an apparent reduction in putative GABAergic neurons identified by parvalbumin, calbinden, and calretinin, but caution is required before concluding that reduction in immunostaining implies neuron loss. Increased numbers of large pyramidal neurons and balloon cells were observed and were associated with alterations in the distribution of neurons with NMDAR1 and GluR2-3 subunits of glutamate receptors. Abnormal plexuses of parvalbumin reactive (putative GABAergic) terminals surrounded the large pyramidal and balloon cells. These cellular alterations in epileptogenic regions of dysplastic human cortex may be clues about changes in inhibitory and excitatory circuitry that play a role in the initiation of seizures in this particular epileptic condition.

The preliminary observations in these cases, which clearly require additional quantitative analysis and confirmation by other groups, are consistent with a reduction in inhibitory neurons, alterations in excitatory circuits, and compensatory changes in inhibitory interneurons. This pattern of alterations will come as no surprise to epilepsy researchers who have studied chronic models during the last decade. It is hoped that epilepsy surgery groups with access to human dysplastic cortex and other cortical structural pathologies associated with seizures will continue to pursue detailed anatomic and physiologic studies of the neuronal and circuit alterations. Observation of consistent patterns of circuit alterations in resected epileptogenic human tissue may provide an impetus and direction for more rigorously controlled experiments in chronic models of epilepsy. Earlier studies in the pilocarpine model^10-12 and recent emerging observations in the chronic models induced by kainic acid¹³ and kindling¹⁴ have revealed similar patterns of long-term alterations that include neuronal loss accompanied by reduced inhibition and alterations in excitatory circuitry. Although variability and heterogeneity have plagued epilepsy research for decades, the observation of consistent patterns of circuit alterations in epileptic human dysplastic cortex and emerging evidence of similar patterns of circuit alterations in a variety of chronic models of epilepsy suggest opportunities to define general principles and mechanisms that play a role in the common but heterogeneous clinical disorders that comprise the epilepsies.