Disease modification in partial epilepsy

With the growth in antiepileptic drug treatment, the question arises as to what extent we are merely treating the symptom (i.e. the seizures) rather than the underlying disease process (i.e. epileptogenesis). Epileptogenicity can be considered as the process whereby structural and functional changes occur following an insult that in some cases result in epilepsy. Epileptogenicity also describes some of the changes and processes that contribute to the progression observed in some epilepsies. These processes have been modelled in animals mostly by the kindling model of epilepsy, in which repetition of subconvulsive stimuli results in a progressive epileptic state and eventually leads to spontaneous seizures. However, it is not clear that kindling has a human correlate, so models in which an initial insult (status epilepticus, hyperthermia, hypoxia, trauma) is followed by the development of lowered seizure threshold and, in some instances, spontaneous seizures have been used. These models seem to support the "second hit" hypothesis, in which there is an initial insult resulting in lowered seizure threshold, and then a later insult, the "second hit", that results in the expression of epilepsy. These models also support the concept of a latent period during which there could be targeted therapies to prevent the epileptogenic process. Although the occurrence of neuronal damage is one such target, neuronal damage is not necessary for epileptogenesis, and other mechanisms are at play. At the present time, it is not known whether targeted therapies may also affect compensatory processes, such as brain repair. Clearly, this would be a potential risk of such strategies. Epidemiological evidence and trials indicate that our present antiepileptic drugs are not effective in preventing epileptogenesis; antiepileptic drugs were, however, not designed for this purpose. Data from animal experiments suggest that treatment of non-convulsive status epilepticus following specific insults may prevent epileptogenesis. The relevance of this for the human condition remains uncertain, but non-convulsive status epilepticus is probably an under-recognized and probably under-treated condition. Perhaps one of the most salutary findings has been the observation of decreased childhood epilepsy with improved neonatal care. This highlights the importance of medical care at the time of an insult, and of prevention of the insults. This review discusses the data that support the concepts underlying epileptogenesis and the model systems that are presumed to reflect the human condition. Particular attention is paid to the potential for interrupting the processes underlying epileptogenesis.