Neonatal seizures: a clinician's overview

Seizures are the most frequent neurological event in newborns (NBs), provoked often by noxae not apt to cause them in later life. This is because receptor families of excitatory amino acids (EAA) are overexpressed at this stage of brain ontogenesis, which is also why most neonatal seizures rapidly abate, even when neurological deficits persist. The brain's immaturities dictate distinct seizure phenotypes. A classification proposed in the late 1960s has been criticized, and a new one has been advocated, based on correlations between EEGs and behaviors, leading to a classification of seizures into 'epileptic' and 'non-epileptic'. The taxonomic pitfalls of these classifications are discussed, and the notion advanced that many seizures fail to fulfil the criteria to label them as epileptic. While etiological factors have changed in time, the striking dichotomy in outcome has persisted. Many etiologies, often multifactorial, are unique in NBs, and they are discussed with reference to diagnosis and therapies. Four syndromes of NB seizures, accepted into the International Classification of the Epilepsies, are critically analyzed, some appearing to rest on fragile grounds. Controversies persist whether seizures per se are injurious to the immature brain. Clinical studies suggest that neither duration in days or length of seizure phenotypes correlates with outcomes, the most valid prognostic indices being offered by etiologies and by patterns of EEG polygraphy. However, because most seizures are symptomatic, it may be difficult to distinguish morbidity due to underlying pathology from that possibly added by seizures. Animal experiments suggested that they are injurious. The theory of energy failure, postulated to cause a cascade of events leading to inhibitions of DNA, proteins, lipids and disrupted neuronal proliferation, synaptogenesis, myelination, has largely been disproved. Brains of immature animals have been shown to have the oxidative machinery needed to fulfill energy demands, even during status convulsivus. They are also capable of using anaerobic metabolism and require less ATP when aerobic energy production ceases. Recent explanations for the injurious consequences of hypoxic ischemia and of prolonged convulsions postulate that neuronal damage occurs from excessive release of EAA which, by binding to their ligand-gated ionic receptors, cause a large influx of Ca2+, resulting in cell death. Because of the overabundance of EAA receptors in early ontogenesis, the excitotoxic hypothesis would appear attractive, but some observations militate against it. Among these is the dissociation found between the focal neurotoxicities induced by EAA injected into the brain and their absence following the concomitant convulsions. The latter are not blocked by pretreatment with EAA antagonists, while these prevent injuries caused by the injected EAA. There is no convincing evidence that excessive release of EAA occurs during NBs' seizures. Even if it does occur, it has been shown that immature neurons have a better capacity to self-protect from increased Ca2+ influx, and also that direct application of glutamate to immature neurons leads to significantly lower Ca2+ influx. These data raise doubts about the postulated excitotoxicity caused by NBs' seizures, being consistent with the fact that no one, so far, has observed neuronal damage from drug-induced convulsive states in NBs. Lack of overt neuronal injuries does not preclude that long-term subtle changes might be induced by noxae apt to provoke transient ictal events. Thus models developed in our laboratories demonstrate that long-term epileptogenicity results following postnatal O2 deprivation without evidence of neuronal injuries or of long-term behavioral or electrophysiological alteration. However, both age at which hypoxia occurs and specific proconvulsant methods used strictly determine whether increased epileptogenicity will occur.