Introduction

Over the past few years, we have gone from having no animal models of acquired epilepsy (defined as an illness characterized by the occurrence of spontaneous epileptic seizures) to having a plethora of animal models with a clinical course and a neuropathology that often closely mimic some forms of the human illness. In most of those models, chronic epilepsy is the long-term sequela of a bout of status epilepticus (SE) and is associated with hippocampal neuronal loss and hippocampal synaptic reorganization. Thus, SE-induced epileptogenesis is a widespread phenomenon induced by many different types of SE in several animal species. Human evidence of the same phenomenon, however, is surprisingly scarce and indirect. In this chapter we briefly review the phenomenon of SE-induced epileptogenesis and make a few points derived from our own research observations using the Sloviter model of perforant path stimulation (PPS) under anesthesia, which has the unique advantage that its lesions are restricted to the hippocampus.

Cortical lesions are often found in epilepsy patients, and there is considerable circumstantial evidence of their epileptogenicity (1, 51, 102). Palecortical foci, such as those associated with hippocampal lesions, are associated with one of the most common types of focal (localization-related) epilepsy (40, 50). Hippocampal lesions show the strongest association with intractable epilepsy and have been extensively studied. In the hippocampus of patients with temporal lobe epilepsy, intractable seizures and/or SE are associated with selective injury to Sommer's sector (prosubiculum and CA1) and CA3 (43); somatostatin- and neuropeptide Y-immunoreactive (ir) neurons are selectively lost in the dentate hilus (44, 69); and GAD-immunoreactive (ir) basket cells are relatively preserved (10, 69). Neuronal loss in the dentate hilus has been postulated to induce the aberrant mossy fiber sprouting also observed in these patients (9, 55, 66, 68, 124). In addition to these morphologic alterations, surgically resected hippocampi from patients with intractable epilepsy show enhanced N-methyl-d-aspartate (NMDA) receptor-mediated responses (57, 68, 138) and sometimes decreased γ-aminobutyric acid (GABA)-mediated inhibition (139). Consequently, these hippocampal lesions exhibit enhanced excitability, which may play a role in their epileptogenicity. Additional evidence for the epileptogenicity of hippocampal lesions is provided by human cases of domoic acid intoxication, which induced hippocampal sclerosis and caused temporal lobe epilepsy in humans (28), just as kainic acid does in experimental animals.

Several animal models of SE show neuronal damage in the hippocampus similar to that seen in patients with temporal lobe epilepsy, and also show spontaneous epileptic seizures after SE. Chemical convulsant–induced SE models are easy to induce and have been the most extensively studied. However, in these models there are many extrahippocampal lesions. Cholinergic drugs such as pilocarpine induce widespread brain damage (31, 47, 90) that is sometimes greater in neocortex than in hippocampus (31), and kainic acid, despite some selectivity for the hippocampus (13), still produces extensive extrahippocampal lesions (14, 107). Therefore, in these models we have no assurance that the seizures come from the hippocampus.

Sustained stimulation of the perforant path induced neuronal injury morphologically similar to that caused by kainic acid (92, 111, 126). Light microscopic analysis showed a highly reproducible pattern of hippocampal damage to a selected population of ipsilateral dentate hilar interneurons (110, 112, 114) and bilateral CA1 and CA3 pyramidal neurons (111). Presumably because the 24-hour stimulation is performed under urethane anesthesia (109), this model shows more restricted lesions than other types of experimental SE, and it is free from the direct effects of convulsant drugs on neurons. We carried out a detailed study of epileptogenicity in this model, where the lack of extrahippocampal lesions is a major asset, in the PPS model in awake rats, which is a good model of self-sustaining SE, and in the lithium-pilocarpine model, which is convenient for pharmacotherapy.