| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1995;59:579-584
© 1995 The Society of Thoracic Surgeons
Division of Cardiac Surgery, the Johns Hopkins Medical Institutions and the Kennedy-Krieger Research Institute, Baltimore, Maryland
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
-amino-3-hydroxy-5-methylisoazole-4-propionic acid/kainate or AMPA/kainate subtype, may predominate in the adult brain. In this experiment, a selective AMPA antagonist, NBQX, was used in a canine survival model of hypothermic circulatory arrest. Twelve male dogs (20 to 25 kg) were placed on closed-chest cardiopulmonary bypass, subjected to 2 hours of hypothermic circulatory arrest at 18°C, and rewarmed on cardiopulmonary bypass. All were mechanically ventilated and monitored for 20 hours before extubation and survived for 3 days. Six dogs received NBQX beginning 2 hours after arrest (3 mg/kg for 3 hours then 1.5 mg/kg for 2 hours). Control dogs received vehicle only. Neurologic recovery was assessed every 12 hours using a species-specific behavior scale that yielded a neurodeficit score ranging from 0 (normal) to 500 (brain dead). After sacrifice at 72 hours, brains were examined by receptor autoradiography and histologically for patterns of selective neuronal necrosis and scored blindly from 0 (normal) to 100 (severe injury). Dogs given NBQX had better neurologic function compared with controls (neurodeficit score, 58.6 ± 15 versus 204 ± 30; p < 0.004) and had less neuronal injury (18.2 ± 3 versus 52.5 ± 6; p < 0.004). Densitometric receptor autoradiography revealed preservation of neuronal NMDA receptor expression only in dogs given NBQX. These results suggest that antagonism of the non-NMDA glutamate receptor AMPA may be neuroprotective in adults after hypothermic circulatory arrest. |
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
We have shown previously that inhibiting the action of the neurotransmitter glutamate by pharmacologic antagonism of the N-methyl-D-aspartate (NMDA) glutamate receptor reduces the neurologic injury caused by profound hypothermia and circulatory arrest [1]. The NMDA glutamate receptor subtype predominates in the immature brain. During development, however, other glutamate receptor subtypes are transiently over-expressed; the non-NMDA receptor -amino-3-hydroxy-5-methylisoazole-4-propionic acid/kainate (AMPA/kainate) may be more prominent in the mature brain [2].
To evaluate the role of the AMPA/kainate receptor in diffuse brain injury associated with hypothermic circulatory arrest (HCA), we used the selective AMPA antagonist 6-nitro-7-sulfamoyl-benzo(f)quinoxaline-2,3-dione or NBQX in a canine survival model of HCA. We hypothesized that NBQX could ameliorate the selective neuronal necrosis caused by prolonged periods of circulatory arrest, rendering this antagonist a potentially effective cerebroprotectant in older patients after HCA.
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Bilateral tympanic membrane, nasopharyngeal, and rectal temperature probes were placed. Tympanic membrane temperature (Ttm) correlates very closely with brain temperature. Swan-Ganz and arterial catheters were placed percutaneously. Electroencephalography (EEG) was recorded with a Grass M 8-16 (Grass Co, Quincy, MA). Nine needle electrodes (Grass) were inserted in the dog scalp and the head insulated with transparent adhesive material and aluminum foil to reduce artifact while on cardiopulmonary bypass (CPB). Electroencephalographic recording parameters included a time constant of 0.3 to 1 Hz, a high-frequency filter of 35 Hz, 30 mm/s paper speed, and initial sensitivity of 70 µV. Electrode impedence was measured less than 3 k
. Eight channels were used in bipolar montage and a variable number of channels in referential montage. Electroencephalographic baseline was recorded under anesthesia before the initiation of CPB. Electrocerebral silence was defined as absence of EEG activity at a gain of 2 µV.
Cardiopulmonary Bypass and Hypothermic Circulatory Arrest
The CPB circuit included a Bentley-10 Plus bubble oxygenator (Baxter Healthcare, Irvine, CA), a 40-µm in-line arterial filter, and a Sarns roller pump (Sarns Inc, Ann Arbor, MI). Heparin was administered (300 U/kg intravenously) and animals were cannulated for closed-chest CPB; venous cannulas (16F to 18F) were advanced to the level of the right atrium through the right femoral and right external jugular veins; the arterial cannula (12F to 14F) was placed in the descending thoracic aorta via the right femoral artery.
Cardiopulmonary bypass was initiated and animals were cooled to a Ttm of 18°C using surface (ice bags around head and cooling blanket) and core (CPB) cooling. During CPB, mean blood pressure was maintained at 50 to 55 mm Hg with pump flows of 80 to 100 mL/kg, reduced to 50 to 60 mL/kg when the Ttm was less than 32°C. Arterial blood gases were controlled using the alpha-stat strategy. The arterial pump was turned off and the venous blood allowed to drain by gravity, thereby exsanguinating the animal. Circulatory arrest was maintained for 2 hours at 18°C.
Then CPB with rewarming was reinstituted. At normothermia (37°C), the dogs were weaned from CPB and decannulated. After decannulation, the right femoral artery and vein were ligated. Protamine was administered and the wounds were irrigated and closed.
Animals subsequently recovered on a ventilator and underwent 20 hours of intensive care monitoring. Anesthesia was maintained using nitrous oxide (2.5 L/min) and intravenous fentanyl. Cardiac rhythm, arterial blood pressure, Swan-Ganz parameters, and urine output were monitored, along with arterial blood gases and levels of hemoglobin and glucose; appropriate adjustments were made as needed. Animals were weaned from ventilatory support at the end of 20 hours and extubated using standard protocol.
Animals were followed up for 72 hours from the cessation of CPB. Neurologic assessment was performed every 12 hours. At the conclusion of the 3-day period, after the final neurologic examination, animals were reanesthetized, intubated, and ventilated. Through a median sternotomy, the ascending aorta was cannulated and the descending aorta clamped. Dogs were sacrificed by perfusion-fixation of the brain with 3 L of paraformaldehyde (3%, pH 7.4) via the aortic cannula. After 24 hours, the brains were removed and further fixed for 14 days before being sliced into sections for subsequent microscopic examination.
NBQX Protocol
Experimental dogs (treated group, n = 6) received NBQX in an investigator-blinded manner. Two hours after completion of HCA (at the end of CPB), each dog received 3 mg/kg for 3 hours, then 1.5 mg/kg for 2 hours. This dosing regimen was chosen because we previously have shown the importance of a weaning schedule for glutamate receptor antagonists [1]. Control animals (control group, n = 6) received vehicle only.
Neurologic Injury Evaluation
Animals were neurologically assessed using a species-specific behavioral scale developed for dogs at the International Resuscitation Research Center, University of Pittsburgh [3]. Neurologic deficit scoring consisted of five components including level of consciousness, cranial nerve function, breathing pattern, motor and sensory function, and behavior. A score of 100 was assigned to each category; 0 was normal function, 500 was brain death.
Histopathology
Histopathologic analysis was performed on all dogs. After fixation, each brain was sliced into 3-mm coronal sections; 18 to 20 sections were embedded in paraffin, and 10-µm sections of each block were stained with hematoxylin and eosin, cresyl violet (for cell bodies), and luxol-fast-blue (for white matter and myelin), Sections of 25 anatomic areas were examined blindly by a neuropathologist for evidence of neuronal injury. Each population of neurons was scored as normal = 0; mild changes of neuronal injury (shrunken and angular) = 1; moderate neuronal injury (hypereosinophilic) = 2; frank neuronal necrosis with loss of some neurons = 3; and neuronal and vessel necrosis = 4. The neuronal populations then were divided into nine anatomically related regions for comparison between experimental groups of dogs: the neocortex included frontal, parietal, temporal, occipital, insular, and cingulate regions with a cumulative injury score of 24. The basal ganglia included globus pallidus, putamen, caudate nucleus, and thalamus; the worst score was 16. The brain stem included midbrain, substantia nigra, periaqueductal grey mater, pons, and medulla; the worst score was 20. The cerebellum included Purkinje cells and granular and molecular layers; the worst score was 12. Deep central white matter included the anterior commissure and corpus callosum; worst score was 12. The remaining regions included hippocampus, dentate gyrus, entorhinal cortex, and amygdala. The total histopathology score for each brain was the sum of all scores for each of the 25 anatomic areas. The worst possible score was 100.
Receptor Autoradiography
To quantify glutamate receptor subtype expression, densitometric receptor autoradiography was performed on 3 dogs in each group. After sacrifice, 20-µm sections of fresh brain were processed for autoradiographic imaging by labeling with 3H-tritiated glutamate (to image NMDA receptors) or with 3H-tritiated AMPA (non-NMDA receptors) by a method described previously [4]. Density of receptors were calculated as picomoles per microgram of protein and compared with 3 normal or negative-control dogs (ie, dogs not undergoing CPB or HCA).
Animal Care
Animals received humane care in compliance with the ``Guide for the Care and Use of Laboratory Animals'' published by the National Institutes of Health (NIH publication 85-23, revised 1985).
Statistical Analysis
All values are expressed as mean ± standard deviation. Comparisons between groups were made using analysis of variance for repeated measures or Student's t test where appropriate. Receptor autoradiographic density values were compared between groups using a nested analysis of variance.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
During the 20-hour recovery period, there were no differences in mean arterial pressure between groups; cardiac output ranged from 2.1 to 2.8 L/min for all dogs. Acidosis associated with HCA was rapidly corrected in both groups with no significant differences in arterial CO2 tension levels or bicarbonate requirements during recovery. Mild hyperglycemia and hemodilution were similar in both groups. There were no significant differences between groups in the quantities of anesthetic agents required during the procedure or the 20-hour recovery period.
Electroencephalography
In treated dogs, EEG silence was obtained within 22.1 ± 2 minutes after the initiation of CPB, at a Ttm of 20.8° ± 1°C. This was not significantly different in controls, where EEG silence occurred within 21.6 ± 3 minutes at a Ttm of 21.3° ± 1°C; continous EEG activity returned after 24.4 ± 3 minutes of rewarming on CPB in the treated group corresponding to a Ttm of 23.4° ± 1°C. This was similar to control dogs where EEG activity returned after 24.1 ± 2 mins of rewarming on CPB, corresponding to a Ttm of 24.9° ± 2°C. In both groups, the predominant frequency at 2 hours after arrest was nonrhythmic and diffuse with a frequency of 18 to 25 Hz. We previously have reported this pattern as a predictor of poor neurologic outcome after prolonged HCA [5].
Neurologic Outcome
At each postarrest neurologic evaluation, the neurologic deficit score was significantly higher in the control group compared with the treated group (p < 0.001 by analysis of variance (Fig 1
). Apart from mild abnormalities of gait in 3 dogs, animals receiving NBQX were neurologically normal before sacrifice on postarrest day 3; all could maintain posture and feed normally.
|
The final neurologic deficit score was 204 ± 30 for the control group compared with 58.6 ± 15 in the treated group (p < 0.004).
Histopathology
The neuronal injury observed on microscopy was that of selective neuronal necrosis (Fig 2, Table 1). The most severely affected regions were the pyramidal cells of the CA-1 hippocampus, the molecular layer of the dentate nucleus, and entorhinal cortex; the Purkinje cells of the cerebellum, lamina 3 and 5 of the neocortex, and the basal ganglia (particularly the globus pallidus) also sustained severe injury. No injury was observed in the white matter or brain stem in any animal.
|
|
Receptor Autoradiography
Densitometric receptor autoradiography in the control group revealed a significant decline in NMDA-sensitive 3H-glutamate binding after HCA in the hippocampus (Fig 3). Receptor autoradiography in the hippocampus showed that in the CA-1 (representing the stratum oriens and pyramidal cell layer) and CA-3 there was a significant reduction in NMDA receptor density of 75% and 64%, respectively, in the control group, compared with normal dogs (ie, negative controls or the normal group). In the corresponding regions in the treated group, however, there was no significant decline in NMDA-sensitive 3H-glutamate binding, indicating preservation of glutamate receptor expression of these neuronal populations by NBQX.
|
), the 20% and 11% reduction in the CA-1 and CA-3 regions, respectively, was not significant compared with the normal group. Binding of AMPA remained at normal levels in treated dogs.
|
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The ionotropic glutamate receptors can be divided into NMDA and non-NMDA or AMPA/kainate based on selective agonists. Although overstimulation of NMDA receptors appears to be a major mechanism of injury including focal strokes, non-NMDA receptors also have been shown to mediate neurotoxicity, particularly in prolonged or global insults [8].
Antagonists of NMDA and non-NMDA receptors ameliorate damage mediated by glutamate in both in vitro and animal models of neuronal injury [9, 10]. We have shown previously that inhibition of the NMDA receptor with the selective antagonist dizocilpine (MK-801) significantly reduces the selective neuronal necrosis and improves functional recovery after 2 hours of HCA at 18°C [1].
In the developing mammalian brain, different receptor subtypes are temporarily over-expressed by neurons for regulation of synaptogenesis, neuronal circuitry, and cytoarchitecture [2]. Thus neuronal populations differ in their vulnerability to glutamate at any given time, depending on their repertoire of receptors. Such a distinct ontogenic profile of susceptibility to excitotoxicity implies that as the balance of receptor subtype changes from NMDA to non-NMDA during brain development, a selective AMPA/kainate antagonist may be most appropriate for adult patients undergoing procedures involving prolonged HCA and global ischemia [11].
In this study we used the selective AMPA/kainate receptor antagonist NBQX in a canine model of HCA. There was significant improvement in the functional recovery of dogs subjected to 2 hours of circulatory arrest at 18°C. Microscopy confirmed the neuroprotective capacity of NBQX in reducing the selective neuronal necrosis in the hippocampus, neocortex, basal ganglia, and cerebellum, whereas receptor autoradiography revealed preservation of both NMDA and AMPA receptor expression after HCA.
These results represent further evidence of a role for glutamate excitotoxicity in the development of HCA-induced neurologic injury. They suggest that the AMPA/kainate receptor mediates neuronal injury after a global ischemic insult to a more significant extent than previously appreciated.
The fact that NBQX afforded neuronal protection when administered after the ischemic insult and 2 hours of reperfusion on CPB has important implications. It confirms that excitotoxicity is a delayed process occurring over several hours after the inciting event and that the cascade can be arrested by administration of glutamate antagonists after a neurologic injury has occurred. Previous experiments in rodents have demonstrated that AMPA antagonism can reduce brain injury when given after the onset of ischemia [10]. It is interesting that the EEG pattern and predominant frequency of both groups of dogs 2 hours after reperfusion was predictive of severe neurologic damage. NBQX, however, given after this electrical activity had developed, prevented the anticipated injury. Should this particular property prove common to other AMPA antagonists, it would afford these compounds an attractive advantage over currently used methods of neuroprotection, including selective anterograde or retrograde cerebral perfusion [12, 13] and cold blood or asanguineous cerebroplegia [14, 15], which can be employed only in the operating room when prolonged circulatory arrest is anticipated.
Although glutamate antagonists prevent excitotoxic neuronal injury, they can cause widespread central nervous system depression and pathologic changes in neurons [16], including morphologic damage to certain defined neuronal populations in the cerebral cortex of rats treated with dizocilpine [17]. Such changes were not observed in our treated dogs. Barbiturates can prevent these changes, and the use of the anesthetic thiamylal sodium in our model may have retarded this pathomorphologic injury.
The densitometric receptor autoradiography in this study illustrates the neuroprotective properties of NBQX in critical brain regions. It is intriguing that there was preservation of NMDA receptor expression throughout the hippocampus by the non-NMDA, AMPA/kainate receptor antagonist, and yet the decrease in AMPA expression was insignificant. AMPA/kainate receptors carry most of the fast excitatory activity in the brain and they control, in part, the opening of NMDA channels by depolarizing the neuronal membranes in which they reside. Therefore, it is possible that by inhibiting the AMPA/kainate receptor, NBQX retards depolarization of the neuronal membrane and blocks voltage-dependent influx of calcium through the NMDA channel, thus ameliorating the neurotoxicity in the presence of excessive concentrations of glutamate. This may explain the observed preservation of NMDA receptor expression.
Because of the genetically programmed expression of different glutamate receptor subtypes at different stages of neurodevelopment, the neuroprotection afforded by the selective glutamate receptor antagonists, including NBQX in this study and the NMDA antagonist dizocilpine in previous work, may have important ramifications for future clinical application of these compounds in cardiac surgery. The age of the patient at the time of operation ultimately may dictate which antagonist will be most suitable for administration before or after prolonged periods of circulatory arrest. In fact, the profile of neuronal vulnerability to excitotoxicity in early infancy and childhood may help determine timing of operations for certain congenital heart defects requiring prolonged HCA to facilitate repair.
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
We thank Lars Nordholm, PhD, and Anker Jon Hansen, MD, for their expert advice and generous assistance in accomplishing this project and for providing the NBQX compound.
In addition, we express our gratitude to David Goldsborough for his help in the statistical analyses and Barbara Dobbs and Lynn DiMarcantonio for their help in preparing the manuscript.
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Address reprint requests to Dr Baumgartner, Department of Surgery, The Johns Hopkins Hospital, Blalock 618, 600 N Wolfe St, Baltimore, MD 21287-4618.
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  J. A. Williams, C. J. Barreiro, L. U. Nwakanma, M. S. Lange, L. E. Kratz, M. E. Blue, J. Berrong, N. D. Patel, V. L. Gott, J. C. Troncoso, et al. Valproic Acid Prevents Brain Injury in a Canine Model of Hypothermic Circulatory Arrest: A Promising New Approach to Neuroprotection During Cardiac Surgery Ann. Thorac. Surg., June 1, 2006; 81(6): 2235 - 2242. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Shake, E. A. Peck, E. Marban, V. L. Gott, M. V. Johnston, J. C. Troncoso, J. M. Redmond, and W. A. Baumgartner Pharmacologically induced preconditioning with diazoxide: a novel approach to brain protection Ann. Thorac. Surg., December 1, 2001; 72(6): 1849 - 1854. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Baumgartner, P. L. Walinsky, J. D. Salazar, E. E. Tseng, M. V. Brock, J. R. Doty, J. M. Redmond, M. E. Blue, M. A. Goldsborough, J. C. Troncoso, et al. Assessing the impact of cerebral injury after cardiac surgery: will determining the mechanism reduce this injury? Ann. Thorac. Surg., June 1, 1999; 67(6): 1871 - 1873. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. E. Tseng, M. V. Brock, C. C. Kwon, M. Annanata, M. S. Lange, J. C. Troncoso, M. V. Johnston, and W. A. Baumgartner Increased intracerebral excitatory amino acids and nitric oxide after hypothermic circulatory arrest Ann. Thorac. Surg., February 1, 1999; 67(2): 371 - 376. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. E. Tseng, M. V. Brock, M. S. Lange, J. C. Troncoso, C. J. Lowenstein, M. E. Blue, M. V. Johnston, and W. A. Baumgartner Nitric oxide mediates neurologic injury after hypothermic circulatory arrest Ann. Thorac. Surg., January 1, 1999; 67(1): 65 - 71. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. E. Tseng, M. V. Brock, M. S. Lange, M. E. Blue, J. C. Troncoso, C. C. Kwon, C. J. Lowenstein, M. V. Johnston, and W. A. Baumgartner Neuronal Nitric Oxide Synthase Inhibition Reduces Neuronal Apoptosis After Hypothermic Circulatory Arrest Ann. Thorac. Surg., December 1, 1997; 64(6): 1639 - 1647. [Abstract] [Full Text]
|
|
|
|
|
|
M. V. Brock, J. M. Redmond, S. Ishiwa, M. V. Johnston, W. A. Baumgartner, J. C. Laschinger, and G. M. Williams Clinical Markers in CSF for Determining Neurologic Deficits After Thoracoabdominal Aortic Aneurysm Repairs Ann. Thorac. Surg., October 1, 1997; 64(4): 999 - 1003. [Abstract] [Full Text]
|
|
|
|
 |
|
 A. J. du Plessis Topical Review: Cerebral Hemodynamics and Metabolism During Infant Cardiac Surgery. Mechanisms of Injury and Strategies for Protection J Child Neurol, August 1, 1997; 12(5): 285 - 300. [Abstract] [PDF]
|
|
|
|
|
|
A. Mori, T. Ueda, T. Nakamichi, M. Yasudo, R. Aeba, H. Odaguchi, A. Mitsumaru, T. Ito, R. Yozu, A. Koto, et al. Detrimental Effects of Exogenous Glutamate on Spinal Cord Neurons During Brief Ischemia In Vivo Ann. Thorac. Surg., April 1, 1997; 63(4): 1057 - 1062. [Abstract] [Full Text]
|
|
|
|
|
|
M. V. Brock, M. E. Blue, C. J. Lowenstein, F. A. Northington, M. S. Lange, M. V. Johnston, and W. A. Baumgartner Induction of Neuronal Nitric Oxide After Hypothermic Circulatory Arrest Ann. Thorac. Surg., November 1, 1996; 62(5): 1313 - 1320. [Abstract] [Full Text]
|
|
|
|
 |
|
 M. H. LeBlanc, X. Q. Li, M. Huang, D. M. Patel, and E. E. Smith AMPA Antagonist LY293558 Does Not Affect the Severity of Hypoxic-Ischemic Injury in Newborn Pigs Stroke, October 1, 1995; 26(10): 1908 - 1915. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
