Value of preoperative EEG for carotid endarterectomy.

PURPOSE: This study was designed to determine whether the preoperative, baseline electroencephalogram (EEG) can be used for intraoperative decision making during carotid endarterectomy, and to identify circumstances where the EEG can be eliminated. METHODS: The charts of all patients undergoing carotid endartectomy at the authors' institution from June 1991 to May 1995 were reviewed to identify those patients that had adequate pre- and intraoperative EEG monitoring. EEGs during 331 carotid endartectomies in 303 patients were coded without knowledge of outcome; primary and secondary endpoints were EEG changes with clamping and clinical outcome, respectively. RESULTS: The incidence of mortality and major neurological morbidity was 1.8%. Baseline-EEGs were abnormal in 105 patients (32%). Whereas baseline-EEG changes were highly predictive of EEG changes after anesthetic induction (P < .0001), they were not predictive of EEG changes with clamping or of clinical outcome. Prior stroke (CVA) predicted abnormal baseline-EEGs (P < .0001) and abnormal post-anesthetic EEGs (P < .0001) but did not predict changes with clamping or perioperative CVA. EEG changes with clamping occurred during 18% of operations; such changes were predicted only by contralateral occlusion (P < .0016) and EEG changes during a prior contralateral carotid endartectomy (P < .0001). The only variable that predicted an adverse neurological outcome was the presence of contralateral occlusion, which increased the likelihood of a perioperative neurological event seven-fold (P = .0038). Clinical outcomes in the 57 of 105 patients with abnormal baseline-EEGs and the 49 of 83 with prior CVA who were shunted were not different from those who were not. CONCLUSIONS: baseline-EEG is not of value for the prediction of adverse events during carotid endartectomy and can be eliminated. Because contralateral occlusion is highly predictive of changes with clamping, and patients undergoing a second carotid endartectomy will usually manifest EEG changes identical to those at the first, operative EEG monitoring can also be eliminated from both these circumstances. Finally, prior stroke does not lead to a higher incidence of clamp-induced EEG changes, and thus is not an indication for shunting in and of itself.

The effects of neonatal hypoxia on kindled seizure development and electroconvulsive shock profiles.

PURPOSE: Our previous research indicated that the exposure of rat pups to an hypoxic environment during a discrete developmental period (postnatal days 10-15) produces short-term seizures and confers an enduring increase in susceptibility to pentylenetetrazol- and flurothyl-induced seizures. In this study, we evaluated the effects of hypoxic insult in this neonatal period of susceptibility to electrical kindling and corneal electroconvulsive shock. METHODS: Ten-day-old rat pups were exposed to a 3% O2 environment, as previously described, and were either kindled or exposed to corneal electroshock at adulthood (70 days old). RESULTS: Neither kindled seizure development from the septal nucleus or amygdala nor electroconvulsive shock profiles were significantly altered by hypoxic pretreatment. CONCLUSIONS: Results indicate that hypoxia produces increases in seizure susceptibility that are observable in only some experimental seizure models but not in others. This outcome serves to target some anatomic systems more than others in the mechanisms involved in hypoxia-induced neural reorganization.

Evidence for the interaction of brainstem systems mediating seizure expression in kindling and electroconvulsive shock seizure models.

Amygdala kindling was observed to increase significantly the proportion of rats that exhibited tonic hindlimb extension in response to corneal electroshock stimulation. Mechanical brainstem lesions which abolished electroshock-induced tonic hindlimb extension failed to alter either the expression of fully generalized kindled seizures or the development of amygdala kindled seizures. Results suggest that while kindling can alter the sensitivity of brainstem systems involved in the expression of tonic hindlimb extension, these same systems are not necessary for either the development or expression of amygdala kindled seizures.

Epileptogenic effect of hypoxia in the immature rodent brain.

The response to cerebral hypoxia/ischemia may be different in the neonate compared to other age groups. An in vivo model was developed in the rat to determine whether there are age-dependent differences in the effects of hypoxia on electroencephalographic (EEG) activity. EEG recordings were obtained from Long Evans hooded rats deprived of oxygen at five ages: postnatal days 5 to 7, 10 to 12, 15 to 17, 25 to 27, and 50 to 60. Oxygen concentration was varied from 0, 2, 3, and 4% between animals. EEGs were recorded in all animals before, during, and at 1 hour after exposure to the hypoxic condition and at 1 to 7 days afterward in a subset of animals. All animals were deprived of oxygen until the onset of apnea and bradycardia to 20 to 40% of baseline heart rate values. Hypoxia resulted in isoelectric EEG significantly more frequently in the animals deprived of oxygen at postnatal days 25 to 27 and 50 to 60 than in the younger age groups. A highly significant effect was that the animals deprived at postnatal days 5 to 17 revealed a high incidence of epileptiform EEG activity during hypoxia. In contrast, the older animals exhibited only rare isolated EEG spikes before reaching an isoelectric EEG. The severity of hypoxia-induced epileptiform EEG changes was highest in the animals subjected to moderately hypoxic conditions (3% and 4% oxygen) at postnatal days 10 to 12. Furthermore, epileptiform changes persisted for hours to days following prolonged episodes of hypoxia in the younger animals. This study demonstrates a unique response of the immature brain to exhibit epileptiform activity during hypoxia.

Perinatal exposure to anoxia alone does not alter the susceptibility to amygdaloid-kindled seizures in the adult rabbit.

It is suggested that asphyxia on newborns increases the susceptibility to epileptic syndromes. The effect of perinatal and postnatal anoxia on subsequent seizure susceptibility was assessed by amygdaloid kindling in adult rabbits. Rabbits from 1 day pre-term to 53 days were exposed to 100% N2 for an average of 7 min or until the heart rate was reduced by 70%. Non-anoxic littermates served as controls. At 2 months of age, animals were implanted with bilateral amygdalae electrodes. After a postsurgical recovery period, afterdischarge (AD) thresholds were determined for the electrode sites and a kindling paradigm was performed. There were no significant differences in the rate of kindling in all groups studied (control, anoxic at 1 day pre-term or at term, anoxic at 44 and 53 days). These results demonstrate that perinatal anoxia did not alter the seizure susceptibility in the adult rabbit kindling model.

Evidence for a norepinephrine-dependent brain-stem substrate in the development of kindling antagonism.

Kindling antagonism' is a modification of the standard kindling paradigm in which two limbic system structures are stimulated on alternate trials. The consistent outcome of this procedure is that one site ('dominant site') develops typical generalized seizures, while kindled seizure development from the other site ('suppressed site') is arrested at an intermediate stage. Our recent research suggests that kindling antagonism is dependent on hindbrain norepinephrine (NE) systems. The present study explores this relationship further. Neonatal rats were treated with intracerebral injections of 6-hydroxydopamine (6-OHDA). At adulthood, these animals received either (1) a low dose of 6-OHDA (75 micrograms), (2) a high dose of 6-OHDA (200 micrograms), or (3) vehicle. Regional NE concentrations were determined by high pressure liquid chromatography. Neonatal 6-OHDA followed by vehicle resulted in decreases in forebrain and increases in hindbrain NE concentrations. Low dose 6-OHDA at adulthood depleted cerebellar, but not pontine-medullary NE. High dose 6-OHDA resulted in depletions of both cerebellar and pontine-medullary NE. Only high dose 6-OHDA significantly interfered with the development of antagonism. Neonatal 6-OHDA facilitated the rate of dominant site kindling independently of subsequent adult treatment regimens. Results suggest that the influence of NE on kindling antagonism is mediated through a brain-stem substrate and that the influence of NE on kindling rate is mediated through a forebrain substrate.

Stepwise progression of kindling: perspectives from the kindling antagonism model.

What is the nature of the kindling process? We hypothesize that kindling is a discontinuous process involving discrete, stepwise transitions from one state of neural organization to another. Our data from the kindling antagonism paradigm argue that there are two critical transitions in the kindling process. These transitions constitute major steps in kindled seizure development. They act as "gates" controlling the ability of afterdischarge (AD) activity to effect the necessary reorganization of neural function which drives the kindling process. We identify two critical gates: 1) a forebrain gate which is dependent on norepinephrine (NE) and effects a discrete transition from nonconvulsive, stage 1 and 2 behaviors to stage 3 seizures, and 2) a brainstem gate which is also NE-dependent and effects a transition from stage 3 seizures to stage 4 and 5 seizures. These gates separate the kindling process into 3 "phases" which are different from, but overlap, the traditional behavioral stages of kindling defined by Racine. Current data suggest that these phases involve independent neural circuitry. They may also involve different physiological mechanisms, but this remains to be determined. This hypothesis is designed to provide a framework for the kindling process within which to search for kindling mechanisms.

Microinjections of GABA agonists into the amygdala complex attenuates kindled seizure expression in the rat.

The effects of intraamygdala injections of either gamma-vinyl GABA or muscimol on the behavioral and electrographic expression of stable, fully generalized, kindled seizures were assessed. Results suggest that intraamygdala administration of GABA agonists preferentially attenuates the behavioral, but not the focal, electrographic expression of kindled seizures elicited from either the insular or entorhinal cortex. These results, in conjunction with those of others, suggest that the amygdala becomes an integral and necessary structure for the expression of seizures kindled from a variety of forebrain areas.

Kindling does not cause persistent changes in firing rates or transmitter sensitivities of substantia nigra pars reticulata neurons.

These studies were undertaken to determine whether the kindling process induces persistent alterations in the functional status of neurons of the substantia nigra pars reticulata, a brain area identified previously as a site important in regulating the expression of generalized motor seizures. Extracellular, single-unit recordings of pars reticulata neurons were made in chloral hydrate-anesthetized, fully kindled rats (2-3 weeks after the last seizure), or unkindled control rats of the same age and weight. Kindling caused no alterations in several electrophysiological parameters examined. For instance, neither the number of active pars reticulata cells encountered, nor their firing rates, were significantly different between kindled and control groups. In addition, kindling failed to alter the sensitivities of pars reticulata neurons to iontophoretic application of two inhibitory transmitters, gamma-aminobutyric acid and glycine, and two transmitters that excite these cells, glutamate and acetylcholine. These results suggest that while kindling produces enduring increases in seizure susceptibility, it causes no persistent interictal changes in either basal activity or several measures of transmitter sensitivity of substantia nigra pars reticulata neurons.

Kindling antagonism: a role for hindbrain norepinephrine in the development of site suppression following concurrent, alternate stimulation.

The concurrent, alternate electrical stimulation of the septal nucleus and the entorhinal cortex results in the development of fully generalized seizures at one site (dominant site) and the lack of development of kindled seizures at the other (suppressed or antagonized site). We have labeled this phenomenon kindling antagonism. Previous work from our laboratory has demonstrated that the whole brain depletion of norepinephrine (NE) eliminates the development of kindling antagonism. In the present study animals were treated with the neurotoxin 6-hydroxydopamine (6-OHDA) as neonates. The neonatal administration of 6-OHDA produced robust increases in brainstem and cerebellar NE levels and depletions of forebrain NE levels when assayed at maturity. Striatal dopamine levels were spared by this treatment. Neonatal 6-OHDA did not alter the development of the kindling antagonism phenomenon which is typically observed following concurrent, alternate stimulation of the septal nucleus and entorhinal cortex. Neonatal 6-OHDA treatment significantly facilitated the rate of kindled seizure development at dominant sites but failed to alter thresholds for the elicitation of afterdischarges (AD) or patterns of development of AD durations. Other characteristics of kindling antagonism were similarly unaffected by 6-OHDA treatment. These data suggest that brainstem and/or cerebellar NE are sufficient to mediate the development of kindling antagonism in the relative absence of forebrain NE.

Etiologic and preventive aspects of epilepsy in the child--bridging the gap between laboratory and clinic.

Four broad categories of basic phenomena are pertinent to developing ways to prevent epilepsy. These include mechanisms of epileptogenesis, ictal initiation and temporary entrainment by the seizure discharge of normally functioning brain, seizure propagation, and control mechanisms that function both to restrain the cascade of epileptic events culminating in a seizure and to arrest the epileptic event and restore the interictal state. In newborns and children, hypoxia-ischemia is a major factor leading to epileptogenesis, and several schemes are proposed to classify, quantify, and prevent hypoxic-ischemic encephalopathy. Control mechanisms must be better understood in order to develop prophylactic recommendations for epilepsy, and an experimental model of "kindling antagonism" may increase our understanding of these. Programs of prevention of seizures in children will evolve only if basic researchers and clinicians work productively together to develop an adequate understanding of factors important in epileptogenesis and antiepileptogenic control mechanisms.

Kindling antagonism: effects of norepinephrine depletion on kindled seizure suppression after concurrent, alternate stimulation in rats.

The concurrent, alternate electrical stimulation of the entorhinal cortex and septal nucleus results in the development of fully generalized seizures at one electrode site and the suppression of seizure development at the other. We have labeled this phenomenon kindling antagonism. Selective, whole-brain depletion of norepinephrine (NE) virtually eliminates the development of kindling antagonism such that fully generalized seizures develop at both sites in a majority of animals. This effect occurs in the absence of appreciable changes in kindling characteristics of these animals compared with either untreated or vehicle-treated controls. These results suggest that the suppression of seizure development observed in the kindling antagonism model is normally maintained by a NE-dependent mechanism. Our results support those of earlier studies using single-site kindling paradigms in which NE depletion facilitates the rate of kindled seizure development. We suggest that the NE-dependent mechanism responsible for the seizure suppression observed to follow concurrent, alternate stimulation and the suppression of seizure development using single-site kindling paradigms may be the same. The nature of this NE-dependent seizure suppression mechanism and the anatomic locus or loci critical for this effect remain questions for future research.

The effects of chronic desmethylimipramine on entorhinal cortical kindling in rats.

Chronic pretreatment of rats with desmethylimipramine (DMI) significantly slowed the rate of seizure generalization elicited by repeated electrical stimulation of the entorhinal cortex (kindling). An identical drug regimen administered to either fully kindled rats or rats partially kindled to early motor seizure stages failed to significantly alter kindling profiles in these animals. Under these latter conditions, in fact, there was a tendency for chronic DMI to exacerbate seizure activity. The effect of chronic DMI pretreatment to slow the development of kindled seizure generalization did not occur if a two-week delay was interposed between the end of drug treatment and the beginning of kindling trials. Results suggest that the retardation of entorhinal cortical kindling rate is dependent on DMI-induced CNS adaptations which recover within two weeks following treatment, and this effect is dependent on the presence of DMI-induced adaptations in a naive (unkindled) nervous system. Alterations of either the kindled state or the adaptational state produced by chronic drug eliminate the slowing of seizure generalization observed when both conditions are present.

Comparative pharmacological effects on visual cortical neurons in monocularly deprived cats.

Monocularly deprived (MD) cats show a loss of responsiveness to visual stimulation of the deprived eye among visual cortical neurons. Several lines of evidence suggest that this effect involves, at least in part, a suppression of deprived eye input, possibly mediated by GABA inhibition. In order to better understand the nature of this suppression we have evaluated the effectiveness of different types of disinhibitory and excitatory agents to reverse the effects of MD. We investigated bicuculline (a GABA antagonist); picrotoxin (a GABA antagonist with a different mechanism of action from bicuculline); strychnine (a glycine antagonist); ammonium ion (a blocker of membrane chloride channels); physostigmine (a cholinesterase inhibitor); and naloxone (an opiate antagonist and also a GABA antagonist). All drugs were given intravenously. Bicuculline restored binocularity to 50% of the visual cortical neurons tested and naloxone to 36%. With both drugs, receptive fields of the normal eye tended to lose specificity. The emergent deprived eye receptive fields were usually similar to those of the normal eye after drug administration. Ammonium ion produced binocular responses in 27% of neurons tested, but receptive fields were grossly abnormal; moreover, ammonium infusion tended to depress neuronal responsiveness. All other drugs tested failed to restore binocularity. These experiments lend further credence to the hypothesis that GABA inhibition contributes to the cortical effects of MD, since only drugs with GABA antagonistic action were effective in restoring neuronal responsiveness to the deprived eye.

Microiontophoretic bicuculline restores binocular responses to visual cortical neurons in strabismic cats.

Rearing cats with surgically induced strabismus resulted in an abnormally high percentage (80%) of monocularly driven neurons in visual cortex. Microiontophoretic application of the GABA antagonist bicuculline restored binocular responses to over 50% of these monocular cells. Elevation of spontaneous rate by glutamate failed to produce binocularity. These results indicate that intracortical inhibition plays a role in the abnormal ocular dominance distribution of strabismic cats.

Extraction of clinical information from electroencephalographic background activity: the combined use of brain electrical activity mapping and intravenous sodium thiopental.

Traditional visual inspection of electroencephalographic (EEG) tracings and computer-assisted topographic mapping were compared in their abilities to detect and locate supratentorial lesions following intravenous sodium thiopental administration. Of the 13 subjects, 8 had atrophic and 5 had mass lesions, all defined by computed tomographic scan and 11 confirmed at operation. EEGs made before and after thiopental administration were evaluated separately from topographic maps of statistical difference between EEGs at multiple frequency ranges made before and after thiopental injection. Topographic mapping of statistical difference accurately detected all thirteen lesions, whereas EEG detected eight. In addition to demonstrating reduced beta production overlying structural abnormalities, topographic mapping revealed regionally augmented beta, especially over irritative lesions. Moreover, localization was possible with the topographic method after the first thiopental injection, whereas a second injection was required for EEG localization by visual inspection. Changes in slow (delta) activity were also useful in delineation of atrophic lesions, in which delta was usually augmented but occasionally regionally reduced. Multielectrode studies with topographic mapping appear essential in delineating cerebral abnormalities, because both slow and fast activities may be increased or reduced over such areas. The relative response of EEG background activity to thiopental at different frequencies may assist lesion characterization as well as localization.

Animal models of strabismic amblyopia: physiological studies of visual cortex and the lateral geniculate nucleus.

Receptive field properties of visual cortical and lateral geniculate cells were studied in 4 models of amblyopia in the cat: monocular deprivation (MD cats), surgical esotropia (esotropic cats), optically induced concomitant strabismus (stationary prism cats) and optically induced incomitant strabismus (rotating prism cats). Comparison observations were made in normal cats. Recordings in visual cortex indicated a reduction in responsiveness to the treated eye in MD and rotating prism cats. Esotropic and stationary prism cats showed mainly a loss of binocular cells. Recordings in the lateral geniculate nucleus indicated a reduction in the spatial resolving capacity of X-cells driven by the treated eye in MD, esotropic and rotating prism cats. The magnitude of this effect was comparable in all of these preparations. Stationary prism cats showed comparable spatial resolving capacities in X-cells driven by either eye. Y-cells were unaffected in any preparation except MD where there were reduced frequencies of Y-cells driven by the treated eye. These results indicate that: (1) interocular differences in spatial patterns without form deprivation are sufficient to produce a loss of responsiveness to one eye in visual cortex; (2) incomitant disparities are necessary to produce the physiological correlates of amblyopia in cats; and (3) deficits in spatial resolution in geniculate neurons are comparable in magnitude in various amblyopic preparations.