Increased seizure susceptibility in the immature brain

Epilepsy is the third most common neurologic disorder and the one with the highest incidence of seizures occurring early in life (41). In experimental epilepsy, a seizure results from an imbalance between excitation and inhibition (5). A possible explanation is that the excitatory postsynaptic potentials (EPSPs) are dominant, leaving the countereffect of inhibitory postsynaptic potentials (IPSPs) relatively inoperable (60, 83, 84). Another possibility, suggested by Sloviter (84), is that inhibitory synaptic inputs are impeded or nonfunctional during a seizure because of dormancy of interneurons and their disconnection from glutamatergic afferents. However, it has since been directly demonstrated that GABAergic inhibition in hippocampal interneurons is operative and only partly reduced in two seizure models of temporal lobe epilepsy (20). Moreover, lack of inhibitory drive cannot be the only explanation for elicitation of asynchronous activity. This is because the period of maximal seizure susceptibility does not coincide with the first postnatal week, when inhibition is minimal or lacking, but instead occurs just prior to the second week and continuing on into the third week of postnatal life, when inhibitory events are reaching maturity (6, 15, 49, 51, 54, 93).

In contrast to the mature brain, the immature brain is highly susceptible to epileptiform activity (2, 62, 68) and status epilepticus (SE) (1, 13, 43, 86, 101, 104). Early studies in neonatal kittens showed that increased seizure susceptibility occurs during a critical window of development, because immature neocortical neurons have high input resistance and hence larger currents (68). Other factors that may lead to developmentally regulated changes in seizure susceptibility include (1) the extent of axonal myelination, which affects communication among cells (78); (2) the existence of electrotonic junctions or ephaptic influences to facilitate neuronal synchronization (78); and (3) the delayed maturation of glia, which may result in accumulation of potassium in the extracellular space and lead to general hyperexcitability (62).

Studies of the immature hippocampus show that expression of inhibitory events in CA1 neurons is delayed. During the first 2 weeks of life, both orthodromic and antidromic stimulation produce only EPSPs, whereas hyperpolarizing IPSPs first appear at 10–14 days of age (78, 79, 94, 96, 97). The apical dendrites of CA1 pyramidal neurons are most sensitive to N-methyl-d-aspartate (NMDA), expressed by large influxes of calcium, during the same critical window for epileptogenesis (second to third postnatal weeks) (39). As the brain matures, the sensitivity of CA1 apical dendrites to NMDA decreases and is similar to responses elicited by dendrites before this critical window. Lowering extracellular calcium levels in young animals can also induce spontaneous paroxysmal activity (2). In accordance with neurophysiologic observations, electron microscopy of the CA1 subregion in rabbit brain reveals few or no symmetric synapses, which have been associated with inhibition in adult animals, until the second and third weeks of development (80).

The developmental window of increased epileptogenesis is also evident within CA3 neurons. For example, IPSPs in CA3 develop by the end of the first postnatal week, 1 week earlier than in the CA1 area. Between the second and third postnatal weeks, CA3 is also more prone to epileptiform discharges relative to CA1 (78, 79, 96). Physiologic experiments in slices prepared from the immature hippocampus show that excitation is predominantly mediated by GABAA receptors (6, 15). In contrast, in adult slices, application of GABA antagonists (such as penicillin) produce prolonged interictal but not ictal activities (92, 93, 95). It has also been postulated that increased seizure susceptibility during the critical window reflects a transitory enhancement of excitation rather than a lack of inhibition (96). These in vitro studies support the conjecture that epileptic activity is developmentally regulated.

To determine whether brain damage early in life increases seizure susceptibility later in life, Sperber and colleagues produced bilateral CA3 lesions in 14-day-old rat pups by radiofrequency stimulation (unpublished observations). After rats matured, they were kindled using a rapid kindling procedure (stimulated at 20-minute intervals). This resulted in a higher kindling stage in lesioned rats than in nonlesioned rats. Similarly, longer afterdischarge durations were observed with increasing stimulation trials (88). A higher susceptibility to seizures and damage was also found in immature rats with hippocampal and cortical neuronal migration disorders, experimentally induced by prenatal exposure to methylazoxymethanol acetate (3, 36). Therefore, if brain damage occurs postnatally, either by direct injury or by in utero chemomanipulation, young rats are more prone to seizures and premature damage.