The electrical discharge of neurons associated with seizure activity stimulates a marked rise in cerebral metabolic activity. Estimates from animal experiments indicate that energy utilization during seizures increases by more than 200%, while tissue adenosine triphosphate (ATP) levels remain at more than 95% of control, even during prolonged status epilepticus (1,2). Cerebral metabolic activity can be measured using 2-deoxy-2-[¹⁸F]-fluoro-D-glucose positron emission tomography (FDG PET) (3), and magnetic resonance spectroscopy (MRS) can be used to measure the turnover of glutamate, the major excitatory neurotransmitter (4). Comparison of FDG PET and MRS in patients with seizures indicates that glucose metabolism and the turnover of glutamate are tightly coupled (3). This is because the transporters that remove glutamate as well as gamma-aminobutyric acid (GABA), the major inhibitory neurotransmitter, from the synaptic cleft require substantial energy (5). These transporters are predominantly located on astroglia that surround synapses, and they remove glutamate and GABA soon after their release from presynaptic nerve terminals to terminate their activity. These transporters are also located on excitatory neuron terminals and are responsible for synaptosomal glutamate uptake (6). The transporters are linked to a Na⁺/K⁺-ATPase, which in turn is powered by anaerobic metabolism of glucose (4). Additionally, chloride cotransporters that maintain the transmembrane Cl^– gradient are also energy dependent (7). The brain generally withstands the metabolic challenge of seizures quite well because enhanced cerebral blood flow delivers additional oxygen and glucose. Mild to moderate degrees of hypoxemia that commonly accompany seizures are usually harmless. In addition, the immature brain is less vulnerable to injury from seizures than the adult brain (8,9). However, severe seizures (ie, lasting >5 min) and status epilepticus (>30 min) can sometimes produce an imbalance between metabolic demands and cerebral perfusion, especially if severe hypotension or hypoglycemia is present. Injury can also occur through a direct effect on synapses known as excitotoxicity (10,11).

Excitotoxicity refers to a process of injury triggered by excessive activation of neuronal receptors for the excitatory amino acid (EAA) neurotransmitters glutamate and aspartate. Under normal circumstances, activation of EAA receptors mediates more than half of the neuronal communication in the brain (12). Seizures initiate release of EAA neurotransmitters from nerve terminals into the synaptic cleft, where they bind to specific neuronal receptors and cause postsynaptic excitation. In the developing brain, EAA receptors perform an additional role as regulators of neuronal development (13). During a prolonged seizure, if the EAA neurotransmitter release continues unabated to raise the synaptic concentration of glutamate, excessive EAA receptor stimulation may damage and eventually kill neurons.

An important component of the excitotoxicity theory holds that EAAs cause intracellular calcium overload, which intoxicates neurons (14–16). Calcium and sodium are allowed to enter neurons through pores or ionophores controlled by EAA receptors. The concentration of calcium within the neuron is normally 10,000 times lower than in the extracellular fluid, and calcium entry is tightly controlled. The neuron has several mechanisms for protecting itself from excessive calcium influx, including sequestration of calcium within mitochondria and endoplasmic reticulum, and energy-dependent pumps that move calcium outward across the cell membrane. Continued bombardment of the neuron by EAA neurotransmitters can overwhelm these protective mechanisms, allowing calcium to poison the metabolic machinery of the neuron (17). The EAA theory proposes that excitatory neurotransmitters are a major direct link between excessive seizure activity and neuronal injury; however, metabolic derangements such as hypoxia may play a contributing destructive role by reducing the efficiency of energy-dependent protective mechanisms and hastening EAA-stimulated metabolic exhaustion (18).

PHARMACOLOGY OF EXCITATORY AMINO ACID RECEPTORS

Glutamate is released from presynaptic nerve terminals, but the synaptic concentration is quickly reduced by an active reuptake process mediated by its specific, energy-dependent reuptake pump. Impairment in the activity of the EAA reuptake pump may play an important role in EAA neurotoxicity in disease states. In experimental hypoxia–ischemia, synaptosomal uptake of glutamate is transiently reduced, contributing to marked elevations in extracellular glutamate concentrations, which may reach the micromolar range (19). Metabolic disorders may contribute to EAA neurotoxicity by this mechanism.

The excitatory responses to glutamate are mediated by several receptor subtypes that may be classified into two broad categories as either N-methyl-D-aspartate (NMDA)-type or non-NMDA-type EAA receptors (20,21) (Figures 4.1 and 4.2). Glutamate is a relatively flexible amino acid molecule that can assume several different conformations, and NMDA is a rigid glutamate analog that represents one of them (22). NMDA maximally activates a glutamate receptor that is linked to an ion channel. Together, the receptor and its channel are referred to as the NMDA receptor/channel complex (see Figure 4.1). Molecular cloning of NMDA receptors showed that their diverse characteristics are related to variations in specific combinations of subunits (23). The NMDA channel is permeable to cations such as sodium and calcium, and at normal membrane potential the channel is blocked by magnesium ions (24). Partial depolarization of the membrane potential is necessary for the magnesium block to be removed. As a result, the NMDA receptor/channel complex is not generally involved in ordinary rapid impulse flow.

Several antagonists of the NMDA receptor/channel complex have been developed, some of which block the NMDA receptor (competitive NMDA antagonists) whereas others block the associated cation channel (noncompetitive antagonists). Drugs such as 3-(2-carboxylpiperazin-4-yl)-propyl-1-phosphonic acid (CPP), CGS-19755, and 2-amino-5-phosphonovaleric acid (AP5) are competitive blockers of the NMDA receptor (25). The channel-binding site is sometimes referred to as the phencyclidine receptor. Two potent preclinical noncompetitive channel-blocking drugs are MK-801 and phencyclidine (26,27). The clinically approved anesthetic ketamine (28) and the antitussive dextromethorphan (29), as well as the anticonvulsant felbamate (30), are also NMDA channel blockers. Memantine, an “un-competitive” NMDA channel blocker, has recently been approved by the Food and Drug Administration for use in Alzheimer disease. It differs from noncompetitive NMDA blockers in that it binds more avidly to open channels, rather than indiscriminately blocking all channel conformations (31). This mechanism may be associated with fewer psychotomimetic side effects than are associated with more complete channel blockade.

The NMDA receptor/channel complex also includes several regulatory sites. One receptor is specific for the simple amino acid glycine (32,33). Unlike the inhibitory glycine site in the spinal cord, this site is not sensitive to strychnine, and activation facilitates opening of the NMDA channel by glutamate. Recognition sites for zinc and polyamines also appear to regulate NMDA receptor/channel complex activity (13). Drugs that block the glycine site on the NMDA receptor are being developed to attempt to avoid the psychotomimetic side effects of noncompetitive NMDA antagonists (34).

FIGURE 4.1 Current understanding of the components of the NMDA receptor/channel complex, based on biochemical and electro-physiologic evidence, is outlined in this schematic diagram. The NMDA recognition site is coupled to a cationic channel that is permeable to Ca²⁺ and Na⁺. NMDA receptor activation is modulated by several regulatory sites. Activation of a closely associated glycine recognition site is required for channel activation and markedly enhances responses to NMDA. Physiologic concentrations of Mg²⁺ block the NMDA-associated ionophore, and membrane depolarization relieves this blockade. Thus, NMDA receptor/channel activation requires activation of both NMDA and glycine receptors and concurrent membrane depolarization. NMDA-receptor channel activation also can be modulated by Zn²⁺ and polyamines (not illustrated). Noncompetitive NMDA receptor antagonists, such as phencyclidine and its thienyl derivative TCP, block NMDA responses by binding within the NMDA-operated ionophore. Cerebral events that impair the mechanisms that regulate Ca²⁺ homeostasis, can produce a prolonged elevation of intracellular Ca²⁺ concentration and cytotoxicity.

FIGURE 4.2 This schematic diagram outlines the synaptic components that contribute to EAA-mediated synaptic transmission, second-messenger generation, and calcium homeostasis. (1) EAAs such as glutamate are released from presynaptic terminals in a calcium-dependent process by presynaptic depolarization. (2–4) Glutamate in the synaptic cleft depolarizes the postsynaptic membrane by binding to at least three subsets of EAA receptors. Activation of (2) the NMDA receptor/channel complex, (3) AMPA receptors, and (4) kainate receptors produced calcium (Ca²⁺) and sodium (Na⁺) entry through receptor-associated ionophores. Furthermore, activation of a subset of metabotropic receptors, which are linked to phospholipase C, produces phosphoinositol hydrolysis and generation of the second messengers inositol triphosphate and diacylglycerol. (5) The excitatory action of synaptically released EAA is terminated by a presynaptic, high-affinity, energy-dependent transport process. (6–9) Mechanisms that regulate post-synaptic Ca²⁺ homeostasis.

Research investigating differences in function between synaptic and extrasynaptic NMDA receptors have found significant differences (35). Activity-dependent neuronal firing patterns, neuromodulators, and cerebral metabolism modulate the synaptic release of glutamate and therefore extrasynaptic glutamate concentrations. Extrasynaptic glutamate, in turn, can affect neural function and can be neurotoxic on its own. Synaptic and extrasynaptic NMDA receptors differ in their subunit composition, downstream transducing cascades, and effects on cell survival (36). Extrasynaptic NMDAR-dependent neuronal death cascades are mediated by several transduction pathways; many of them antagonize those triggered by synaptic NMDARs (Figure 4.3). Advances in the understanding of selective pharmacology and function of NMDAR subunits will advance the field of new therapeutic strategies given their differential effects on synaptic and extrasynaptic receptors and, therefore, on cell survival (37).

FIGURE 4.3 Transduction pathways involved in cell survival or cell death triggered by synaptic or extrasynaptic NMDA receptors (NMDARs). According to a recent model, synaptic NMDARs, composed primarily of NR1/NR2A subunits, trigger neuroprotective transduction pathways, whereas extrasynaptic NMDARs, composed predominantly by NR1/NR2B subunits, initiate cell death pathways. Extrasynaptic NMDARs are present at perisynaptic sites distributed on dendrites and in some cases adjacent to glia-like processes or axons. Synaptic NMDAR-dependent neuroprotection involves transduction cascades that elicit induction of survival genes, suppression of death genes, or protection against oxidative stress. Calcium influx via synaptic NMDARs triggers Ca2⁺/calmodulin (CaM) kinase IV and extracellular receptor kinase (ERK)1/2 mediated phosphorylation and activation of CREB (cyclic-AMP response element binding protein) and its coactivator CREB binding protein (CBP); calcineurin elicits dephosphorylation and nuclear translocation of the transducer of regulated CREB-activity (TORC), a key step in CREB activation. CREB promotes transcription of activity-regulated inhibitors of death (AID) and brain-derived neurotrophic factor (BDNF) genes, which provide broad-spectrum neuroprotective effects. Synaptic NMDAR activity also triggers transcriptional suppression of core components of the intrinsic apoptosis cascade, in part via suppressing activity forehead box protein O (FOXO)–dependent expression of proapoptotic genes. This antiapoptotic effect results from Akt-mediated phosphorylation and nuclear export of FOXO. Synaptic NMDAR activation also enhances thioredoxin activity, which elicits antioxidant effects (not shown). In contrast, extrasynaptic NMDARs trigger cell death via several transduction pathways; many of them antagonize those triggered by synaptic NMDARs. These include CREB dephosphorylation by Jacob (juxtasynaptic attractor of caldendrin on dendritic spines) protein, inhibition of the ERK1/2 pathway, promotion of nuclear import of FOXO, and activation of calpain.
Bim, bcl2-interacting mediator of cell death; PI3K, phosphatidyl inositol 3’kinase; Txnp, thioredoxin interacting protein.

Non-NMDA receptors were initially identified because they were activated preferentially by the glutamate analogs quisqualic acid and kainic acid (19). Kainic acid is a very potent neurotoxine and convulsant agent in the adult brain (38,39). Currently, two major types of non-NMDA receptors are recognized: one that is linked to an ion channel that fluxes sodium and, to a lesser extent, calcium and another that is linked to stimulation of phosphoinositide (PPI) turnover or downregulation of cyclic-AMP production (20). The so-called ionotropic non-NMDA receptors are now divided into kainate receptors and AMPA (alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate) receptors. Molecular studies indicate that these two receptors have similar molecular characteristics, with AMPA receptors made up from combinations of GluR1 to GluR4 receptor sub-units, and kainate receptors made up of GluR5 to GluR7 subunits. AMPA receptors mediate most of the fast excitatory activity in the brain, and they include most of the receptors previously referred to as quisqualate receptors (20). Perampanel is the selective noncompetitive AMPA-antagonist (40). The anticonvulsant topiramate is known to have several cellular targets for its therapeutic actions that, in addition to AMPA receptors, also include the GluR5 subunit of the kainate-receptor (41), voltage-gated sodium channels (42), and specific GABA-A receptor isoforms (43). Kainate receptors stimulate presynaptic release of glutamate as well as activating postsynaptic receptors. The glutamate receptors linked to second messengers are called metabotropic receptors. A marked increase in glutamate release, which occurs during a prolonged seizure, is likely to result in the activation of all types of glutamate receptors.

DEVELOPMENT OF EXCITATORY AND INHIBITORY SYNAPTIC MARKERS

Synaptic markers for EAA neurotransmitters undergo marked changes during postnatal development (44–47). Studies of the ontogeny of NMDA and AMPA-sensitive receptor subtypes during postnatal development indicate that NMDA-type and non-NMDA-type glutamate receptors have a unique ontogenetic profile in the hippocampal formation (47). When receptors for NMDA-sensitive ³H-glutamate binding, strychnine-insensitive glycine binding, and ³H-N-(1[2-thienyl]cyclohexyl)3, 4-piperidine (TCP) binding to the channel were examined independently, all were found to be overexpressed in the neonatal period relative to the adult. NMDA-sensitive binding exceeded adult levels by 50% to 120% in all regions of the hippocampus examined, with peak densities occurring between postnatal days 10 and 28. The ontogenetic profiles of the glycine modulatory site and the phencyclidine receptor were similar to each other but were delayed with respect to the timing of the NMDA recognition site. The physiologic relevance of overexpression of NMDA recognition sites relative to other components of the NMDA receptor/channel complex is unclear. However, the temporal expression of NMDA recognition sites correlates with the molecular changes in receptor subunits and synaptogenesis (48–50). Studies of NMDA receptors in human cerebral cortex in early life have also demonstrated enhanced levels and different characteristics compared to adults (48,51).

The ontogeny of NMDA receptor binding also appears to correlate with physiologic changes. Long-term potentiation, a synaptic correlate of memory, is enhanced during the neonatal period in rats, at the same time that NMDA-mediated neurotransmission predominates over AMPA-mediated activity (52). Postsynaptic excitation appears to predominate over inhibition during the first few postnatal weeks of hippocampal development, as does enhanced seizure susceptibility (53). In the CA3 region of the hippocampus, superfusion with NMDA elicits recurrent synchronized burst activity, and the epileptogenic effect of NMDA increases from postnatal days 1 to 10 (54). Also, the susceptibility to epileptiform activity increases from postnatal days 4 to 6 to postnatal day 14 (55). This activity is blocked by specific NMDA receptor antagonists (56). The development of NMDA receptors tends to lag behind the development of sensitivity to NMDA neurotoxicity, which transiently peaks on day seven.

AMPA receptors also undergo developmental changes that make them more calcium permeable and more likely to mediate seizures and excitotoxicity in infancy and childhood (57). AMPA receptors that lack GluR2 subunits are calcium permeable, while AMPA receptors that include GluR2 subunits flux only sodium. Jensen et al have studied the development of AMPA receptor subunits in rats and human postmortem tissue and found a correspondence between the expression of GluR2-lacking receptors and regional susceptibility to brain injury and seizures (58–60). GluR2-lacking AMPA receptors appear first in radial glia and oligodendrocytes in the periventricular white matter at 25 to 32 weeks gestation in humans (59). Their expression is delayed to around 40 weeks gestation in pyramidal neurons and nonpyramidal neurons of the cerebral cortex, persisting into early infancy and beyond in selected regions (59). These ontogenetic changes in AMPA receptors are consistent with experimental evidence that AMPA-receptor antagonists provide protection against seizures and damage at these stages in development (61).

The development of markers for GABAergic inhibitory synapses also undergoes major changes in the postnatal period and tends to lag behind development of excitatory synapses (62–64). GABA_A receptor antagonists cause prolonged seizure-like discharges in the hippocampus of 2- to 3-week-old rats (53). However, during early life, extending to the end of the first week postnatally in rats, stimulation of GABA_A receptors paradoxically produces excitation rather than inhibition as a result of a lag in the development of the chloride KCC2 transporter, which maintains low intracellular Cl^– levels (65–68). Without this transporter, high concentrations of Cl^– accumulate within the dendrite, resulting in depolarization when the Cl^– channel is opened by GABA. This is consistent with the theme that excitatory neurotransmission is enhanced early in development to support activity-dependent neuronal development and plasticity and becomes more balanced with inhibitory neurotransmission later in life (69). These developmental programs contribute to the higher incidence of seizures in infants and children.

DEVELOPMENTAL CHANGES IN GLUTAMATE NEUROTOXICITY

The immature brain is quite sensitive to neurotoxicity from glutamate and certain of its analogs (39,70). Olney discovered a number of years ago that feeding monosodium glutamate to neonatal mice produced characteristic lesions in the hypothalamus (39). More recently, it has become clear that the receptors that mediate glutamate neurotoxicity respond differently in the immature brain compared to the adult brain. These observations have important implications for the EAA neurotransmitter hypothesis of epileptic brain injury in children.

In adults, kainic acid is the most potent neurotoxic analog of glutamate, but in neonatal rodents NMDA is the most toxic analog (70–72). Although kainic acid produces seizures in the immature brain, it produces little cytotoxicity (73). In contrast, NMDA injected into the 7-day-old rat brain produces prolonged seizures and extensive neurotoxicity. At 7 days of age, the relative neurotoxicity of several glutamate analogs is NMDA >>> AMPA > kainic = zero, whereas in the adult kainic acid >> NMDA ≥ AMPA. The severity of brain injury produced by direct injection of NMDA into the 7-day-old rat brain is approximately 60 times greater than in the adult (70). There is a relatively sharp peak of sensitivity to NMDA neurotoxicity at 7 days of age in the rat, with less sensitivity at times before and after, suggesting a potentially important effect of age on sensitivity to EAA-mediated injury. The shift in sensitivity to NMDA neurotoxicity parallels developmental changes in genetic expression of NMDA subunits that make up the receptor channel complex (50,74). These changes appear to be programmed to mediate activity-dependent neuronal plasticity during early development (13). NMDA receptor activity is needed to promote normal development, and excessive blockade of these receptors leads to neuronal apoptosis (75). The NMDA receptor is a potential Achilles heel for the developing brain: both overstimulation and blockade can produce damage under certain circumstances (13).

The susceptibility of the developing brain to NMDA-induced brain injury parallels the susceptibility to hypoxic-ischemic brain injury. Olney’s studies suggest that the acute neurohistologic picture of NMDA-induced injury in the 7-day-old rat brain is virtually identical to that from a hypoxic–ischemic injury at the same age (76). NMDA may induce metabolic changes similar to hypoxia–ischemia. Studies using (34) P MRS in brain slices indicate that NMDA produces a rapid exhaustion of phosphocreatine stores and nucleotide triphosphate levels following its application (77). The noncompetitive antagonist MK-801 protects against hypoxic–ischemic damage in the 7-day-old rat brain (78). These studies suggest that over-activation of NMDA receptors may be a final common pathway for hypoxic–ischemic injury, as well as other types of developmental brain injury.

EVIDENCE THAT EAA NEUROTRANSMITTERS PLAY A ROLE IN SEIZURE-RELATED INJURY

The link between excessive EAA neurotransmitter activity and epileptic brain injury was initially strengthened by Meldrum’s pioneering studies in subhuman primates (11). These studies showed that seizures alone could damage the brain and that the EAA analog kainic acid produced sustained seizures and brain injury with a regional histologic pattern resembling epileptic brain injury. Similar studies of selective injury in the hippocampus from kainic acid injection into the brain have been performed in rodents by Nadler et al (79). Two additional models in adult animals have also provided additional, more direct experimental evidence.

In experiments conducted by Sloviter, electrical stimulation of the perforant pathway that projects into the hippocampus from the entorhinal cortex leads to neuronal degeneration in hippocampal zones innervated by perforant pathway fibers (80). The perforant pathway is the major excitatory glutamatergic afferent pathway into the hippocampus (81). The acute neuropathologic changes produced by the persistent stimulation resemble the histopathology seen when glutamate analogues or other convulsant substances are injected into the hippocampus (82).

The pathology replicates the typical “epileptic” pattern of injury with damage to dentate basket cells, hilar cells, and CA3 and CA1 pyramidal cells of the hippocampus. CA2 neurons are typically spared. Electron microscopy shows acutely swollen dendritic segments distributed in a laminar pattern corresponding to the neuronal-receptive fields of EAA pathways in the perforant projection. Somatostatin-containing neurons are especially vulnerable, whereas GABA neurons are relatively spared (83).

In a second model of epileptic brain injury developed by Collins and Olney, persistent focal seizure activity is induced by focal application of the GABA receptor-blocking drug bicuculline onto the cerebral cortex (10). The seizure activity causes acute neurodegenerative changes in specific thalamic regions innervated by the corresponding corticothalamic pathway. This model also produces acute neuropathologic changes that resemble glutamate-like neuronal degeneration. Using this model, Olney’s group demonstrated that ketamine and MK-801, two powerful blockers of the action of NMDA-type EAA receptors, completely abolish the thalamic damage caused by the focal seizure activity (84). MK-801 reduced but did not prevent the electrical seizure activity itself, suggesting that the neuroprotective action resulted from blockade of NMDA-type EAA receptor activation rather than from a reduction in neurotransmitter release. The results of these experiments, along with the electrical stimulation studies, support the hypothesis that EAA receptor stimulation plays an important role in seizure-induced neuronal injury. They also suggest that the neurotoxic effects of seizures can be dissociated from effects of excessive electrical activity.

In addition to etiology, associated seizure burdens often seem to be the defining factor for the emergence of refractoriness to first line AEDs (85). When it comes to neonatal seizures, the ability to determine the true seizure burden of clinically ambiguous behavioral manifestations remains critical in correctly gauging both seizure frequency and response to treatment. Since electroclinical dissociation is now a well-accepted concept (86,87) where cessation of behavioral manifestations of seizures after drug administration represents a false sense of having effectively managed seizures, it is imperative that to truly understand molecular and pharmacologic consequences of seizures, true seizure burdens are ascertained. It has been shown in rodents (88) that there are regional differences in the ontogeny of Cl(-) transport, wherein most thalamic neurons in perinatal rat pups were found to maintain low [Cl(-)] and were inhibited by GABA as well as phenobarbital. In contrast, most neocortical neurons maintained higher [Cl(-)], and were excited by GABA(A)R activation. This may explain why seizure activity in the cortex is not suppressed by anticonvulsants that block the transmission of seizure activity through subcortical networks.

Consequences of Prolonged Febrile Seizures

The FEBSTAT study prospectively studied the relationship between febrile status epilepticus (FSE) and acute hippocampal damage, as well as the development of mesial temporal sclerosis, epilepsy (particularly temporal lobe epilepsy), and impaired hippocampal function (89). The FEBSTAT study has concluded that the HHV-6B infection is commonly associated with FSE and HHV-7 infection is less frequently associated with FSE. Together, they account for one-third of FSE, a condition associated with an increased risk of both hippocampal injury and subsequent temporal lobe epilepsy (90). Earlier studies have reported that mutations in the HCN2 channel may play an important role in the occurrence of febrile seizures; however, these mutations were only found in 2.4% of subjects with febrile seizures and genetic epilepsy with febrile seizures (91) over the 0.2% incidence in blood bank controls. Additionally, the HCN1 and HCN2 variants that have been identified failed to show statistical differences during functional analyses for channel properties (92). Therefore, if gain of function in the HCN channel does play a role in metabolic and pharmacologic consequences of febrile seizures, that role has not been credibly established to underlie FSE.

Children with FSE are at risk for acute hippocampal injury and a substantial number also have abnormalities in hippocampal development (93). Therefore, the conclusion that the immature brain is not susceptible to cell death drawn from animal models of hyperthermia, induced by external heat, with relatively low seizure burdens (94) may not be replicating the clinical condition. In contrast, modifications in the model by exposing the pups to longer duration of heat have also revealed hippocampal cell death (95). This indicates that the severity of the acute injury in addition to the etiology defines the metabolic consequences of FSE.

Hippocampal sclerosis (HS) is the most common neuropathologic pattern observed in pharmacoresistant epilepsy and represents a critical feature in mesial temporal lobe epilepsy syndrome (96). Since febrile seizures (FS) are a prevalent antecedent of temporal lobe epilepsy (TLE), it is likely that the severity of the initial FSE modulates the severity of HS and, therefore, also pharmacoresistance.

GABAergic Consequences of Seizures and Role in Pathogenesis of Epilepsy

It is well known that prolonged seizures in the young brain can produce injury that is aggravated when the inflammatory component is added to the etiology (97,98). The resulting injury contributes to synaptic reorganization that likely plays a major role in the development of chronic epilepsy. Additionally, specific loss of susceptible subsets of interneurons following SE have been implicated to contribute in the process of epileptogenesis (99). Dominant mutations in SCN1A, which encodes the Nav1.1 VGSC α-subunit, underlie several forms of epilepsy, including Dravet syndrome (DS) and genetic epilepsy with febrile seizures plus (GEFS+). However, mouse model studies have shown that 69% of Nav1.1 immunoreactive neurons were also positive for parvalbumin (100). The inactivation of one SCN1A allele in interneurons of the neocortex and hippocampus in the model was sufficient to reduce thresholds to flurothyl- and hyperthermia-induced seizures, whereas thresholds were unaltered following inactivation in excitatory cells. These findings indicate that the GABAergic system plays a predominant role in the generation and pharmacologic consequences of certain seizure syndromes.

Non-Neuronal Consequences of Seizures: Role of Astrocytes

Investigation of specimens from patients with pharmacoresistant temporal lobe epilepsy and epilepsy models have revealed alterations in expression, localization, and function of astroglial K+ and water channels. In addition, malfunction of glutamate transporters and the astrocytic glutamate-converting enzyme, glutamine synthetase, has been observed in epileptic tissue. Dysfunctional astrocytes in gliotic tissue play a crucial role in epilepsy (101).

Decreased expression of inwardly rectifying potassium (Kir) channels in astrocytes are also known to contribute to impaired K⁺ buffering following seizures; this phenomenon is tied to the local inflammatory environment (102). Acute inflammatory reaction after seizures has long been suspected due to clinical observation of pleocytosis, without any evidence of infection, in the cerebrospinal fluid (CSF) and peripheral blood of patients with recent generalized convulsions. Steroids or adrenocorticotropic hormone (ACTH), with a suppressive effect on inflammation or immune reactions, have been used to effectively treat children with intractable epilepsy.

Seizures and Inflammation

Brain inflammation may be a common substrate contributing to seizures and following seizures in drug-resistant epilepsies of different etiologies, and recurrent seizures can per se be a major cause of long-term inflammation (103,104). Following this view, new research shows that pharmacologic blockade of specific inflammatory molecules and pathways can significantly reduce seizures in experimental models of seizures and epilepsy (105). From a clinical standpoint, a role of inflammation in the pathophysiology of human epilepsy is still hypothetical, although this possibility is supported by abundant evidence (106). Increased BBB permeability and changes in endothelial transporters function induced by seizures, inflammation, or both have clinical implications. BBB leakage allows the entry of compounds with immunogenic or inflammatory potentials, as predicted—for example, in interferon-induced seizures (107)—but it may also facilitate the entry of compounds with therapeutic potential with limited or no access to the CNS. Activation of the ACTH–GC axis in response to antecedent injury or stress, leading to hyperfunction of CRH–neuronal pathways, has been suggested to play a role in the pathogenesis of West syndrome (108).

Role of Glucose Transport

Hypoxic–ischemic brain injury is associated with transient compensatory changes targeted at protecting glucose delivery to fuel cellular energy metabolism by transient increases in the neuronal brain glucose transporter isoform (GLUT-3), which then may delay the processes of apoptosis and cell necrosis (109,110).

MECHANISMS FOR SEIZURES AND INJURY IN THE IMMATURE BRAIN

Stay updated, free articles. Join our Telegram channel

Join