Stefanie Dedeurwaerdere

Neuronuclear Assessment of Patients With Epilepsy

Epilepsy is a common chronic neurological disorder that is controlled with medication in approximately 70% of cases. When partial seizures are recurrent despite the use of antiepileptic drugs, resection of the epileptogenic cortex may be considered. Nuclear medicine plays an important role in the presurgical assessment of patients with refractory epilepsy. Single-photon emission computed tomography (SPECT) and positron emission tomography (PET) techniques are used to determine the seizure onset zone, which needs to be resected to render a patient seizure free. Correct localization of the ictal onset zone with the use of SPECT or PET is associated with a better surgical outcome. Ictal perfusion SPECT imaging with (99m)Tc-ethyl cysteinate dimer (ECD) or (99m)Tc-hexamethylpropyleneamine oxime (HMPAO) enables one to detect the seizure onset zone in a majority of cases, especially in patients with temporal lobe epilepsy. Interictal SPECT imaging, which is more widely available, is unreliable to determine the ictal onset zone and is usually only used for comparison with ictal SPECT images. Assessment of the ictal onset zone using subtracted ictal and interictal studies, overlayed on structural imaging has proven to be more sensitive and more specific compared with visual assessment. Video-electroencephalography monitoring in combination with ictal SPECT imaging, however, is only available in specialized centers. It is important to inject the perfusion tracer as early as possible after the beginning of a seizure and to be aware of patterns of seizure propagation. Interictal (18)F-fluorodeoxyglucose (FDG)-PET is routinely used to detect brain areas of hypometabolism, which usually encompass, but tend to be larger than, the seizure onset zone. Also, for assessment of FDG-PET, it is advisable to use an automated technique comparing the patients images to a normal database in addition to visual interpretation of the images, since automated techniques have proven to be more accurate. In view of the thickness of the cortical ribbon, which may be below the resolution of the PET camera, posthoc partial volume correction or PET reconstruction incorporating the anatomical information of magnetic resonance imaging (MRI), may be useful for optimal assessment of glucose metabolism. Perfusion SPECT and interictal FDG-PET are able to demonstrate areas of abnormal perfusion and metabolism at a distance from the ictal onset zone, which may be associated with cognitive and psychiatric comorbidities, and may represent the functional deficit zone in epilepsy. Part of the functional deficit zone is a dynamic seizure-related process, which may resolve with cessation of seizures. In recent years, novel PET tracers have been developed to visualize not only glucose metabolism but also a wide variety of specific receptor systems. In patients with epilepsy, changes in the gamma-amino-butyric acid(A) receptor, opioid receptor, 5-HT(1A) serotonin receptor, nicotinic acetylcholine receptor systems, and others have been described. Because these tracers are not widely available and the superiority of studying these receptor systems over glucose metabolism in the presurgical evaluation of patients with refractory epilepsy remains to be proven, their use in clinical practice is limited at the moment. Finally, advances in small animal PET scanning allow the in vivo study of the process of epileptogenesis, starting from an initial brain insult to the development of seizures, in animal models of epilepsy. Potential new therapeutic targets may be discovered using this translational approach.

Thalamic and limbic involvement in the mechanism of action of vagus nerve stimulation, a SPECT study

PURPOSEnTo unravel the mechanism of action of neurostimulation as a treatment for seizures, functional neuroimaging tools allow minimally invasive research in humans. We performed single-photon emission computed tomography (SPECT) in patients with epilepsy, treated with vagus nerve stimulation (VNS). Changes in regional cerebral blood flow (rCBF) at the time of initial stimulation as well as after chronic treatment were correlated with long-term clinical efficacy.nnnMETHODSnIn this pilot study, 27 patients (14 female and 13 male) who were treated with VNS at Ghent University Hospital for refractory epilepsy underwent a (99m)Tc-ECD (ethyl cystein dimer) SPECT activation study at the time the first stimulation train was administered. 12 patients underwent an additional (99m)Tc-ECD SPECT activation study 6 months later. Image acquisition was performed on a high-resolution triple-headed gamma camera. Significant rCBF changes were correlated with prospectively assessed clinical efficacy data.nnnRESULTSnSignificant rCBF changes were found in the thalamus, the hippocampus and the parahippocampal gyrus. Acute limbic hyper-perfusion and chronic thalamic hypo-perfusion correlate with positive clinical efficacy.nnnCONCLUSIONSnAcute and chronic electrical stimulation of the vagus nerve induces rCBF changes that can be measured by SPECT on a group-basis. The thalamus and the limbic system are thought to play a key role in the mechanism of action of VNS.

Positron Emission Tomography in Basic Epilepsy Research: A View of the Epileptic Brain

Summary:u2002 The neurobiological processes that result in epilepsy, known as epileptogenesis, are incompletely understood. Moreover, there is currently no therapy that effectively halts or impedes the development or progression of the condition. Positron Emission Tomography (PET) provides valuable information about the function of the brain in vivo, and is playing a central role in both clinical practice and research. This technique reliably reveals functional abnormalities in many epilepsy syndromes, particularly temporal lobe epilepsy. Unfortunately, epileptogenesis is extremely difficult to study in human patients who usually present with established epilepsy, rather than at the early stages of the process. Animal models offer the advantage of permitting the assessment of the pre‐, developing, and chronic epileptic states. However, traditional techniques (e.g., histology) are only able to examine the brain at one time point during epileptogenesis in any one individual. Recent advances in dedicated small animal PET (saPET) allow researchers for the first time to study in vivo biomolecular changes in the brain during epileptogenesis by means of serial acquisitions in the same animal. Repeated application of in vivo imaging modalities in the same animal also decreases the effect of biological inter‐individual variability and the number of animals to be used. The availability of novel PET tracers permits the investigation of a broad range of biochemical and physiological processes in the brain. Besides research on epileptogenesis, saPET can also be applied to investigate in vivo the biological effect of novel treatment strategies. saPET is widely used in many fields of pathophysiological investigation and is likely to significantly enhance epilepsy research.

Vagus nerve stimulation does not affect spatial memory in fast rats, but has both anti-convulsive and pro-convulsive effects on amygdala-kindled seizures

Vagus nerve stimulation (VNS) is an adjunctive treatment for refractory epilepsy. Using a seizure-prone Fast-kindling rat strain with known comorbid behavioral features, we investigated the effects of VNS on spatial memory, epileptogenesis, kindled seizures and body weight. Electrodes were implanted in both amygdalae and around the left vagus nerve of 17 rats. Following recovery, rats were tested in the Morris water-maze utilizing a fixed platform paradigm. The VNS group received 2 h of stimulation prior to entering the Morris water-maze. Rats were then tested in the kindling paradigm wherein the VNS group received 2 h of stimulation prior to daily kindling stimulation. Finally, the abortive effects of acute VNS against kindling-induced seizures were determined in fully kindled rats by applying VNS immediately after the kindling pulse. Body weight, water consumption and food intake were measured throughout. Memory performance in the Morris water-maze was not different between control and vagus nerve stimulation rats. Similarly, kindling rate was unaffected by antecedent VNS. However, pro-convulsive effects (P<0.05) were noted, when VNS was administered prior to the kindling pulse in fully kindled rats. Yet, paradoxically, VNS showed anti-convulsant effects (P<0.01) in those rats when applied immediately after the kindling stimulus. Body weight was significantly lower throughout kindling (P<0.01) in VNS-treated rats compared with controls, which was associated with reduced food intake (P<0.05), but without difference in water consumption. VNS appears to be devoid of significant cognitive side effects in the Morris water-maze in Fast rats. Although VNS exhibited no prophylactic effect on epileptogenesis or seizure severity when applied prior to the kindling stimulus, it showed significant anti-convulsant effects in fully kindled rats when applied after seizure initiation. Lastly, VNS prevented the weight gain associated with kindling through reduced food intake.

In-vivo imaging characteristics of two fluorinated flumazenil radiotracers in the rat

Purpose[11C]Flumazenil shows promise as a clinical and research PET radiotracer to image changes in GABAA central benzodiazepine receptor (cBZR), but its widespread use has been limited by practical limitations of [11C]. This study evaluated the imaging characteristics of two fluorinated PET radiotracers in rats in vivo: [18F]fluoroflumazenil ([18F]FFMZ) and [18F]flumazenil ([18F]FMZ).MethodsPET acquisitions were performed on a small-animal scanner following injection of [18F]FFMZ in nine rats and [18F]FMZ in eight rats. The following treatments were investigated: (1) injection of the tracer dose, (2) presaturation then injection of the tracer dose, and (3) injection of the tracer dose followed by a displacement injection. Unchanged tracer was measured in plasma and brain structures in four animals 10 and 30xa0min after injection, and ex-vivo autoradiography was also performed.ResultsFor both [18F]FFMZ and [18F]FMZ maximal brain activity peaked rapidly, and was highest in the hippocampus (1.12±0.06xa0SUV, 1.24±0.10xa0SUV, respectively), and lowest in the pons (1.00±0.07xa0SUV, 1.03±0.09xa0SUV, respectively). By 50xa0min after injection, maximal uptake for [18F]FFMZ and [18F]FMZ had decreased in the hippocampus to 18±3% and 80±1% (p<0.01), respectively. The presaturation and displacement studies showed a higher nonspecific component for [18F]FFMZ than for [18F]FMZ. Metabolite studies showed that at 30xa0min only 10% of the signal was from [18F]FFMZ in the brain. This nonspecific binding was apparent on autoradiography. In contrast, [18F]FMZ accounted for >70% of the signal in the brain, which resulted in well-defined regional binding on autoradiography.ConclusionThese results demonstrate that [18F]FMZ is a superior radiotracer to [18F]FFMZ for in-vivo PET imaging of the GABAA/cBZR, having slower metabolism and leading to lower concentrations of metabolites in the brain that results in a substantially better signal-to-noise ratio.

Manganese-enhanced MRI reflects seizure outcome in a model for mesial temporal lobe epilepsy

The neurobiological processes resulting in epilepsy, known as epileptogenesis, are incompletely understood. Manganese-enhanced MRI (MEMRI) can potentially aide in this quest as it provides superior tissue contrast, particularly of the hippocampal subregions. This longitudinal study aims to characterise the changes in the hippocampus of the post kainic acid-induced status epilepticus (KASE) rat model of mesial temporal lobe epilepsy using MEMRI in vivo. Serial acquisition of T(1)-weighted MEMRI images were taken before, 2 days and 6 weeks after KASE (10-30 mg/kg, i.p.) in 14 rats and in 11 control rats, while a second cohort of control (N=6) and epileptic animals (N=10) was imaged at 2 months post KASE only. MnCl(2) (50 mM, 10 μl) was administered in the right lateral ventricle 1 day before scanning. Regions of interest were drawn around the hippocampus and several subregions of the hippocampus (CA1, CA3 and dentate gyrus). Markers of epilepsy such as spontaneous recurrent seizures, hippocampal neuronal loss and mossy fiber sprouting were quantified. A persistent increase in MEMRI signal intensity was found in the hippocampus, CA1 and dentate gyrus in the KASE group compared to the control group (ANOVA P<0.05). The intensity signal in the hippocampus and subregions correlated inversely with the frequency of spontaneous recurrent seizures in the chronic epileptic phase, however there was no relationship observed between histopathological changes such as cell loss and mossy fiber sprouting with seizures. This study demonstrates that MEMRI is able to detect imaging changes in the hippocampus during the course of epileptogenesis relevant for seizure expression. These data strongly indicate a relationship between manganese enhancement and spontaneous seizure outcome, suggesting that MEMRI could provide a preclinical biomarker for the severity of epileptogenesis in vivo in animal models.

Changes in hippocampal GABAA/cBZR density during limbic epileptogenesis: relationship to cell loss and mossy fibre sprouting.

Reduced GABA(A)/central benzodiazepine receptor (GABA(A)/cBZR) density, mossy fibre sprouting (MFS) and hippocampal cell loss are well described pathological features of human temporal lobe epilepsy (TLE), and animal models thereof. However, the temporal relationship of their development, and their roles in the emergence of the epilepsy, are uncertain. This was investigated in the kainic acid (KA)-induced post-status epilepticus (SE) model of TLE. Male Wistar rats (7 weeks, n=53) were randomised into control and KA groups. At 24h, 2, 4 or 6 weeks sham and KA post-SE animals were euthanised, brains extracted and GABA(A)/cBZR density, neuronal loss and MFS measured in hippocampal sub-regions. GABA(A)/cBZR density (B(max)) was measured by saturation-binding analysis using [(3)H]-flumazenil. At 24h post-SE GABA(A)/cBZR density was increased in almost all hippocampal subregions, but was decreased at the later time points with the exception of the dentate gyrus. There was significant neuronal loss in the CA3 SPc region (-24 ± 9.3%, p<0.05) at 24h, which remained stable at the later time points associated with an elevated GABA(A)/cBZR density per surviving neuron at 24h post-SE (+56.4%; p<0.05) which returned to control levels by 6 weeks post-SE. MFS in the dentate gyrus progressively increased over the 6 weeks following SE (+70.6% at 6 weeks), at which time there was a significant inverse relationship with GABA(A)/cBZR binding (r(2)=0.87; p=0.02). The temporal evolution of GABA(A)/cBZR density changes post-KA-induced SE, and the relationship with decreases in hippocampal pyramidal cell numbers and MFS, may point to a key role for these changes in the pathogenesis of acquired limbic epileptogenesis.

In vivo measurement of hippocampal GABAA/cBZR density with [18F]-flumazenil PET for the study of disease progression in an animal model of temporal lobe epilepsy.

Purpose Imbalance of inhibitory GABAergic neurotransmission has been proposed to play a role in the pathogenesis of temporal lobe epilepsy (TLE). This study aimed to investigate whether [18F]-flumazenil ([18F]-FMZ) PET could be used to non-invasively characterise GABAA/central benzodiazepine receptor (GABAA/cBZR) density and affinity in vivo in the post-kainic acid status epilepticus (SE) model of TLE. Methods Dynamic [18F]-FMZ -PET scans using a multi-injection protocol were acquired in four male wistar rats for validation of the partial saturation model (PSM). SE was induced in eight male Wistar rats (10 weeks of age) by i.p. injection of kainic acid (7.5–25 mg/kg), while control rats (nu200a=u200a7) received saline injections. Five weeks post-SE, an anatomic MRI scan was acquired and the following week an [18F]-FMZ PET scan (3.6–4.6 nmol). The PET data was co-registered to the MRI and regions of interest drawn on the MRI for selected structures. A PSM was used to derive receptor density and apparent affinity from the [18F]-FMZ PET data. Key Findings The PSM was found to adequately model [18F]-FMZ binding in vivo. There was a significant decrease in hippocampal receptor density in the SE group (p<0.01), accompanied by an increase in apparent affinity (p<0.05) compared to controls. No change in cortical receptor binding was observed. Hippocampal volume reduction and cell loss was only seen in a subset of animals. Histological assessment of hippocampal cell loss was significantly correlated with hippocampal volume measured by MRI (p<0.05), but did not correlate with [18F]-FMZ binding. Significance Alterations to hippocampal GABAA/cBZR density and affinity in the post-kainic acid SE model of TLE are detectable in vivo with [18F]-FMZ PET and a PSM. These changes are independent from hippocampal cell and volume loss. [18F]-FMZ PET is useful for investigating the role that changes GABAA/cBZR density and binding affinity play in the pathogenesis of TLE.

Fluctuating and constant valproate administration gives equivalent seizure control in rats with genetic and acquired epilepsy

PURPOSEnControlled-release formulations of Valproate (VPA) reduce side effects by minimizing peak plasma VPA concentrations in patients with epilepsy. However, the impact of this on anti-seizure efficacy has not been thoroughly explored. Here the pharmacokinetics and pharmacodynamics of chronic intermittent (consequently, peak VPA concentrations) and continuous VPA administration were directly compared in two rat models of epilepsy.nnnMETHODSnGenetic Absence Epilepsy Rats from Strasbourg (GAERS) received a single acute bolus of VPA (100 mg/kg intravenously) combined with electroencephalography (EEG) and/or blood sampling for 180 min post-injection. GAERS and epileptic rats post-kainic acid-induced status epilepticus were chronically infused intravenously (3-5 days, respectively) with (i) saline followed by in random order (ii) intermittent and (iii) continuous VPA (42 mg/kg/h), separated by two days of wash-out. Seizures were quantified using video-EEG monitoring and VPA levels measured in brain, cerebrospinal fluid and plasma.nnnRESULTSnFollowing acute VPA administration seizure suppression in GAERS persisted after plasma VPA levels became very low. Chronic intermittent and continuous VPA significantly suppressed seizures in both models (p<0.01) with no difference between administration regimens. In GAERS, the pattern of seizure suppression during intermittent treatment was constant, in contrast to the fluctuating VPA plasma and brain levels. There was discordance between the temporal pattern of plasma, brain VPA levels and seizure suppression efficacy in GAERS.nnnCONCLUSIONnAdministration regimes that result in fluctuating VPA blood levels achieve equivalent sustained seizure suppression as those that maintain steady mid-range concentrations.

A novel system allowing long-term simultaneous video-electroencephalography recording, drug infusion and blood sampling in rats.

INTRODUCTIONnElectroencephalography (EEG) recording and drug administration is commonly used for neurological experiments in rats, but is typically cumbersome due the use of multiple lines. We have developed a unique system, which allows long-term simultaneous video-electroencephalography recording, drug infusion and blood sampling in rats.nnnMETHODSnThe vEEG/drug infusion system was designed and tested on two contrasting animal models of epilepsy. Animals were implanted with EEG-electrodes and a jugular vein cannula fixed in a head cap, avoiding an additional cable for tethering. In an acute infusion study (n=16), repeated blood samples were taken after i.v. bolus injection of valproate. In a subset of these rats (n=10), paired blood samples were removed from the jugular vein and the heart after valproate administration. In a chronic infusion study (n=38), heparinised (4IU/h) saline or valproate (42mg/kg/h) was infused continuously for up to 17 days.nnnRESULTSnIn the acute study, repeated blood samples showed a decrease in plasma valproate levels over time following bolus injection. In the chronic study, high quality continuous EEG was achieved and 79% of animals were successfully infused throughout the planned infusion period (13-17 days), with 66% of projected blood samples able to be taken during the infusion. There was a high correlation between the jugular vein and cardiac plasma levels of valproate (Spearman test, r=0.69; p<0.05).nnnCONCLUSIONnThis system is ideal for pharmacokinetic/dynamic studies and long-term drug infusion where simultaneous EEG and/or frequent blood sampling are desired.