[Relevance of basic research to clinical data: Good answers, wrong questions!]

What conclusions can be derived from experimental data on human epilepsies? This review discusses these issues, notably concerning human temporal lobe epilepsies (TLEs) and infantile epilepsies, where important advances have been achieved in both theory and the comprehension of epileptogenic mechanisms. A wide spectrum of human and animal data converge to show that the naive network transforms to one that generates seizures spontaneously. Thus, in TLE, experimental and human data suggest that the inaugurating status generates a sequence of events that lead to the sprouting of fibers and the formation of novel excitatory synapses. This reactive plasticity constitutes a basis for the generation of novel seizures by the epileptic network. Similarly, in vitro studies indicate that in immature hippocampal formation, the propagation of high- but not low-frequency seizures can transform a naive network into one that generates further seizures, thereby, giving an indication as to the types of seizure that are epileptogenic. In conclusion, it is suggested that although animal data cannot mimic human seizures in all their complex and variable etiologies, it provides essential indications on the mechanisms that enable seizure generation.