Cerebral metabolism after fluid-percussion injury and hypoxia in a feline model

Transcription

1 J Neurosurg 97: , 2002 Cerebral metabolism after fluid-percussion injury and hypoxia in a feline model ALOIS ZAUNER, M.D., TOBIAS CLAUSEN, M.D., OSCAR L. ALVES, M.D., ANN RICE, PH.D., JOSEPH LEVASSEUR, M.S., HAROLD F. YOUNG, M.D., AND ROSS BULLOCK, M.D., PH.D. Division of Neurosurgery, Medical College of Virginia, Virginia Commonwealth University, Richmond, Virginia Object. Currently, there are no good clinical tools to identify the onset of secondary brain injury and/or hypoxia after traumatic brain injury (TBI). The aim of this study was to evaluate simultaneously early changes of cerebral metabolism, acid base homeostasis, and oxygenation, as well as their interrelationship after TBI and arterial hypoxia. Methods. Cerebral biochemistry and O 2 supply were measured simultaneously in a feline model of fluid-percussion injury (FPI) and secondary hypoxic injury. After FPI, brain tissue PO 2 decreased from 33 5 mm Hg to 10 4 mm Hg and brain tissue PCO 2 increased from 55 2 mm Hg to 81 9 mm Hg, whereas cerebral ph fell from to (p 0.05 for all three measures). After 40 minutes of hypoxia, brain tissue PO 2 and ph decreased further to 0 mm Hg and , respectively (p 0.05), whereas brain tissue PCO 2 remained high at mm Hg. Secondary hypoxic injury caused a drastic increase in cerebral lactate from M/L to M/L (p 0.05). The lactate/glucose ratio increased from to after hypoxia was introduced. The O 2 consumption decreased significantly from l/mg/hr to l/mg/hr after hypoxia was induced. Conclusions. Cerebral metabolism, O 2 supply, and acid base balance were severely compromised ultra-early after TBI, and they declined further if arterial hypoxia was present. The complexity of pathophysiological changes and their interactions after TBI might explain why specific therapeutic attempts that are aimed at the normalization of only one component have failed to improve outcome in severely head injured patients. KEY WORDS traumatic brain injury hypoxia cerebral metabolism brain chemistry cat C EREBRAL metabolism, O 2 supply, and acid base balance are severely compromised immediately after TBI and cerebral ischemia. In addition, arterial hypoxia, which is often present after injury and before resuscitation, commonly worsens the clinical picture. Even though neuroscientists have achieved a better understanding of pathophysiological events that occur after severe head injury, as measured in animal models as well as in clinical studies, the early changes remain widely unknown. The neuronal elements of brain tissue have a high metabolic rate, even in areas with a relatively low capillary density. In addition, the rate of O 2 consumption in the normal brain is usually not limited by the O 2 supply. Relatively moderate reductions in the rate of O 2 delivery cause symptoms of hypoxia, however. Hypoxia is defined as a reduction in tissue PO 2 to levels insufficient to maintain cellular function, metabolism, and Abbreviations used in this paper: CBF = cerebral blood flow; FPI = fluid-percussion injury; ICP = intracranial pressure; MABP = mean arterial blood pressure; PBS = phosphate-buffered saline; TBI = traumatic brain injury; TTC = 2,3,5-triphenyl tetrazolium chloride. structure. If CBF is also reduced, this exaggerates an existing cerebral hypoxia. At a PaO 2 of 35 mm Hg, there is a significant increase in cerebral glucose consumption and in lactate production, indicating a stimulation of anaerobic glycolysis. 10 At this stage, blood ph has a pronounced influence on the oxyhemoglobin dissociation curve, causing O 2 availability to decrease. Prolonged hypoxia leads to a gradual fall in blood pressure and to a reduced cerebral perfusion pressure. An initial increase in CBF is usually seen this is a physiological feedback salvage mechanism for preventing cerebral energy failure. During this period a fall in extracellular ph and a rise in the lactate/pyruvate ratio also occur The distinction between hypoxia and ischemia may be somewhat difficult because both lead to cerebral O 2 deprivation. Hypoxia alone, usually hypoxemia, is typically associated with inadequate ventilation, such as that due to apnea, in the acute stages of head injury. Ischemia differs from hypoxia in that it is not only an interruption of the O 2 supply, but it also involves cessation of blood flow, and thus, the failure to eliminate metabolic products such as CO 2, lactic acid, and ammonia. In the absence of an O 2 supply, the tissue can only obtain energy by spending energy-rich phosphate reserves (phosphocreatine, aden- 643

2 A. Zauner, et al. osine triphosphate, and adenosine diphosphate) and by metabolizing limited stores of glucose and glycogen to lactic acid. The aim of this study was to evaluate the early cerebral metabolic events after TBI and arterial hypoxia, and to relate these findings to brain-injured patients. Materials and Methods These nonsurvival experiments were performed after approval was received from the Virginia Commonwealth University Institutional Animal Care And Use Committee. A total of 10 male cats weighing between 4.5 and 5 kg were included in this study. After surgical preparation and sensor placement, baseline values were obtained and all animals underwent FPI followed by induction of arterial hypoxia. All animals were killed at the end of the experiment. Animal Anesthesia and Surgery Animals received no food for 12 hours. Initial sedation was provided by an intramuscular injection of 22 mg/kg ketamine administered in the animal holding facility. After sedation, 10 mg/kg intravenous methohexital was used to induce anesthesia. Anesthesia was then maintained by a mixture of 0.8 to 2.7% isoflurane, 70% N 2 O, and 30% O 2, delivered through an endotracheal tube. The animals received mechanical ventilation throughout the experiments, and the depth of anesthesia was closely monitored by continuously observing arterial blood pressure, heart rate, end-tidal PCO 2, and by intermittent arterial blood gas analysis. The PaCO 2 was kept at 34 2 mm Hg throughout the studies. The rectal temperature was monitored as well and kept at 35.5 to 37 C by using two heating blankets. One femoral artery and one femoral vein were cannulated for monitoring and for venous access. After sedation and intubation, the cats were placed in a stereotactic frame and prepared and draped in accordance with sterile surgical methods. A 4-cm midsagittal incision was made along the vertex of the scalp. A total of four burr holes were made and special care was taken to avoid injury to the underlying dura (Fig. 1). The first burr hole was made in the right supraorbital and frontal areas for placement of a two-lumen skull bolt (Codman, Raynham, MA). This bolt allowed introduction and placement of the sensor (Neurotrend; Codman) and the microdialysis probe (CMA/Microdialysis, Acton, MA). The sensor as well as the microdialysis probe was inserted into brain tissue with a special introduction guide to ensure minimal tissue damage. A horizontal rather than a vertical placement of both probes was performed to accommodate the proper location and insertion depth within the volume of the brain (Fig. 1). The second skull opening, which was in the right frontoparietal area, was used for attachment of the fluid-percussion coupling device. The remaining two small burr holes served for placement of an ICP microsensor (Codman) and for obtaining cortical biopsy samples during the study. All burr holes were sealed with dental acrylic cement or with bone wax. Brain Tissue Monitoring We chose a minimally invasive sensor that uses fiberoptic technology to provide continuous measurements of tissue PO 2 and PCO 2, ph, and brain temperature. This sensor is supplied as a sterile, single-use device, to be inserted with an introducer into brain tissue. All four sensing elements are combined into a single device 0.5 mm in diameter that has a sensing length of less than 20 mm. The sensor is packed with a tonometer containing a buffer solution, which maintains the sensor s hydration and serves as a calibration medium as well. The system underwent an automated calibration cycle for 30 minutes before insertion. All monitored parameters were recorded on a spreadsheet for further analysis. After a stabilization period of approximately 1 hour after surgical manipulation, the recording of baseline values was started. Baseline readings were not initiated before all online readings were stable. Baseline values were collected for a period of 1 hour before FPI. A microdialysis probe with a membrane length of 10 mm and a FIG. 1. Schematic drawing of the experimental setup showing positions of the Neurotrend sensor and microdialysis probe, which were inserted through a bolt. The bolt was inserted supraorbitally, through the right frontal sinus. The region of the FPI is indicated, as well as the area from which the biopsy samples were obtained. cutoff rate of 20 kd was used. The probe was perfused with sterile 0.9% NaCl solution by using a microinjection pump (CMA/Microdialysis) at a flow rate of 2 l/minute. The time intervals for microdialysis sampling were 20 minutes. Microdialysis samples were immediately frozen and stored at 22 C. Cerebral concentrations of glucose, lactate, pyruvate, and glutamate were analyzed by enzymatic fluorometric assays on a microdialysis analyzer (CMA 600; CMA/Microdialysis). The presented data are the concentrations measured in the microdialysate, without corrections for probe recovery rates. The ICP was continuously monitored using an intraparenchymal ICP microsensor. The data were transferred to a spreadsheet for further analysis. Fluid-Percussion Injury and Hypoxia A frontoparietal FPI was performed, after all sensors and catheters had been placed, and after obtaining baseline measurements. The animals were removed from the stereotactic frame before the insult was delivered. All sensors had to be temporarily withdrawn from the brain tissue to avoid damage to the apparatus during the FPI. The 10-mm-diameter, right frontoparietal burr hole was used as the injury site. A straight hollow cylinder was carefully inserted into the burr hole without injuring the underlying dura; the device was secured using acrylic cement. This tube was then filled with normal saline and connected to a fluid-percussion device. The FPI was accomplished by allowing a 4.8-kg steel pendulum to fall through a gravity-dependent arc of 22 and impact on a rubber cushion. On impact, a pressure was coursed through the saline-filled cylinder of the FPI device and hit the dura, causing TBI. High injury levels at atms were selected to produce TBI. The level of injury, expressed in atmospheres for each animal, was mseasured using a pressure pulse recorder on an oscilloscope. Immediately postinjury, the FPI device was removed, the craniotomy defect sealed with Gelfoam, and the animal was placed again in the stereotactic frame. Ten cats underwent FPI and were monitored for at least 2 hours, after which the animals were exposed to arterial hypoxia for 40 minutes, by receiving ventilation with 10% O 2 only. In one animal this resulted in hemodynamic instability, and this cat was excluded from the study. At the end of the experiments, all animals were killed by 644

3 Cerebral metabolism after traumatic brain injury and hypoxia in cats intravenous injection of 5 ml pentobarbital (65 mg/ml). All probes were immediately removed, the skull was opened, and the brain was carefully removed. The harvested brain was chilled on dry ice for 5 minutes, then cut into 5-mm coronal slices. Selected slices were then immersion-fixed in TTC solution in 2% saline. After 30 minutes of TTC immersion, the slices were transferred into a 10% formaldehyde solution, in which they were immersion-fixed overnight. Oxygen Consumption Brain tissue samples (1 mm 3 ) were removed from representative areas adjacent to the cortical injury (fourth burr hole), at the end of the baseline period, at 1 and 2 hours after FPI, and after 1 hour of hypoxia. The harvested tissue was placed on icecold Ringer glucose PBS for further analysis in the cartesian diver microspirometer to measure O 2 consumption. Brain tissue O 2 consumption, as measured with the microspirometer, is expressed in microliters per millgram per hour. The procedure was described fully in an earlier work by Levasseur, et al. 25 Electron Microscopy Brain tissue samples were obtained in areas close to the injury site at the end of the baseline period, 60 minutes after FPI, and 40 minutes after induction of hypoxia (Fig. 1). The biopsy samples were fixed in 2.5% glutaraldehyde (in 0.1 M PBS, ph 7.2) for 48 hours. After being washed in PBS three times for 10 minutes, the samples were placed in 1% osmium (in 0.1 M PBS, ph 7.2) for 1 hour at 4 C. Thereafter, the tissue was dehydrated in ethanol at ascending concentrations, then rinsed in propylene oxide, and placed in propylene oxide plastic resin overnight. Ultrathin sections (6 8 nm) were cut and stained with 1% uranyl acetate in 70% ethanol, then visualized and photographed at various magnifications on a transmission electron microscope. Statistical Analysis The analysis, including descriptive statistics, was performed using statistical software (StatView; Abacus Concepts, Berkeley, CA). The Wilcoxon rank-sum test was used to compare baseline values with those after FPI and after hypoxia. Differences were considered statistically significant at a probability value of less than Results Baseline Measurements Readings from nine male cats weighing between 4.5 and 5 kg were included in the final analysis. The MABP was mm Hg, the baseline mean ICP was 6 1 mm Hg. Baseline arterial blood gas values were kept constant during baseline measurements and were mm Hg for PaO 2, 36 3 mm Hg for PaCO 2, and for arterial ph (Table 1). Brain tissue PO 2 was 33 5 mm Hg, brain tissue PCO 2 was 55 2 mm Hg, and brain ph was Cerebral dialysate glucose during the same period was M/L, lactate was M/L, pyruvate was M/L, and glutamate was M/L. Oxygen consumption during normoxia and baseline conditions was l/mg/hr. Fluid-Percussion Injury All nine animals underwent FPI at atm and were observed for at least 2 hours, before induction of hypoxia, with all monitoring devices in place. After FPI, the MABP (78 17 mm Hg) and blood gas values for PaO 2, PaCO 2, and arterial ph ( mm Hg, 37 6 mm Hg, and , respectively) remained stable during this 2-hour period. The ICP increased significantly, to TABLE 1 Summary of physiological parameters measured in nine cats during the experiment Parameter Baseline Post-FPI Posthypoxia MABP (mm Hg) 81 ± ± ± 22 ICP (mm Hg) 6 ± 1 28 ± 3* 25 ± 3* PaO 2 (mm Hg) 147 ± ± ± 10* PaCO 2 (mm Hg) 36 ± 3 37 ± 6 34 ± 6 arterial ph 7.33 ± ± ±.09 *p 0.05 compared with baseline value mm Hg immediately after injury, and declined to approximately 20 mm Hg toward the end of the FPI period. Brain tissue PO 2 fell by more than 70%, to 10 4 mm Hg after FPI (Fig. 2). Oxygen consumption fell slightly, to l/mg/hr (not significant). Cerebral acid base homeostasis was also significantly altered postinjury. Brain tissue PCO 2 increased by approximately 32%, to 81 9 mm Hg (Fig. 2), whereas brain ph decreased to during the same period (Fig. 3). A significant increase in dialysate lactate ( M/L), pyruvate ( M/L), and glutamate levels ( M/L) occurred after FPI (p 0.05 in all cases). Dialysate glucose levels changed insignificantly after FPI (Table 2). The lactate/glucose ratio increased significantly after FPI, from to (Fig. 4); however, no significant change in the lactate/pyruvate ratio occurred. Arterial Hypoxia During the episode of arterial hypoxia, PaO 2 was reduced to mm Hg, whereas PaCO 2 and arterial ph remained unchanged from baseline, at 34 6 mm Hg and , respectively. During this period the MABP was mm Hg and the mean ICP was 25 3 mm Hg (Table 1). Cerebral tissue oxygenation fell further, from 10 mm Hg after FPI, to 0 0 mm Hg after hypoxia was induced in the animals (Fig. 2). Cerebral tissue O 2 consumption fell by almost 30%, to l/mg/hr. Arterial hypoxia worsened the already existing cerebral tissue acidosis after FPI, by further accumulating brain tissue PCO 2 (83 13 mm Hg) and by lowering brain ph to (Figs. 2 and 3). Dialysate lactate ( M/L) and pyruvate levels ( M/L) increased significantly after 40 minutes of hypoxia (p 0.05 in both instances), whereas the glucose level fell to M/L from baseline (Table 2). The dialysate glutamate level decreased compared with the FPI period, but was significantly elevated compared with baseline. In addition, the lactate/glucose and the lactate/pyruvate ratios were calculated for assessment of cerebral tissue hypoxia. The lactate/glucose ratio increased 12-fold, to (Fig. 4), whereas the increase in the lactate/pyruvate ratio was nonsignificant ( ). Mitochondrial Function and Ultrastructure The TTC staining of brain slices obtained after killing the animals indicated impaired mitochondrial metabolism in the right hemisphere beyond the FPI area. In the remaining, normal brain tissue, TTC was converted to its 645

4 A. Zauner, et al. TABLE 2 Summary of brain parameters measured during the experiment Parameter Baseline Post-FPI Posthypoxia brain tissue PO 2 (mm Hg) * 0 0* brain tissue PCO 2 (mm Hg) * 83 13* brain ph * * glucose ( M/L) * lactate ( M/L) * * pyruvate ( M/L) * * glutamate ( M/L) * * O 2 consumption ( l/mg/hr) * lactate/pyruvate ratio lactate/glucose ratio * 9.1 2* *p 0.05 compared with baseline value. FIG. 2. Graph showing the effect of FPI and arterial hypoxia on brain tissue PO 2 (ptio2), and PCO 2 (ptico2). The mean brain tissue PO 2 decreased significantly after FPI and fell further after hypoxia was induced. During the same period, brain tissue PCO 2 increased substantially, but changed only slightly during the period of arterial hypoxia. reduced red form, indicating normally functioning mitochondria. The extent of the injury and the proper placement of the Neurotrend and microdialysis probes are demonstrated in a representative brain slice (Fig. 5). After FPI, mitochondrial swelling and changes in the mitochondrial ultrastructure were seen, along with normalappearing mitochondria. Nevertheless, hypoxic injury in addition to FPI significantly worsened the ultrastructure of mitochondria. All mitochondria demonstrated massive swelling and cell disruptions after hypoxia (Fig. 6). Discussion Over the past several decades, our appreciation of the mechanisms of brain damage has increased enormously. It has become clear that pathophysiological cascades initiated by TBI, stroke, and subarachnoid hemorrhage have many similar features, with brain ischemia being the common denominator. This phenomenon cannot be diagnosed in the living brain, however, without CBF studies or a new monitoring method. Neurological monitoring interpreted according to coma scales thus remains the mainstay in efforts to detect worsened brain function and to guide therapeutic decisions. This is not possible, however, when patients are comatose, or experiencing pharmacological paralysis. Monitoring of the brain-injured patient has thus recently been extended to include measurement of ICP, cerebral perfusion pressure, electrophysiological function, and other parameters. These measures are often used to detect global changes, however, and do not provide early information on worsening of cerebral oxygenation, CBF, and brain chemistry levels before the onset of permanent damage. Manipulations directed against early changes in cerebral biochemistry, cerebral acid base homeostasis, and substrate delivery for treatment of the injured brain are now becoming feasible, and may improve outcome, but only if directed by measurements of these parameters. This study was therefore designed to evaluate early, regional changes in brain energy metabolism after FPI that was complicated by a secondary hypoxic insult. To detect these changes, we used the same technologies now available for clinical use in severely head injured patients. FIG. 3. Graph showing the effect of FPI and arterial hypoxia on brain tissue ph. Brain ph decreased significantly after FPI and fell even more after arterial hypoxia started. FIG. 4. Bar graph showing the effect of FPI and arterial hypoxia on the extracellular lactate/glucose (LG) ratio. The lactate/ glucose ratio increased significantly 2 hours after FPI and had an even more dramatic increase after arterial hypoxia was induced. *p

5 Cerebral metabolism after traumatic brain injury and hypoxia in cats FIG. 5. Photograph of a representative cat brain slice after staining with 2% TTC (tetrazolium red), showing the influence of FPI and arterial hypoxia on TTC staining as an indicator of mitochondrial function. Large, pale areas are seen in the right hemisphere, indicating impaired mitochondrial metabolism, whereas the unaffected tissue shows normal staining of the reduced TTC form. The arrow indicates the locations of the microdialysis probe and Neurotrend sensor within the right hemisphere. Cerebral ischemia, accompanied by hypoxia resulting in secondary brain damage, is one of the major factors determining the outcome in patients after severe TBI. 3,4 On postmortem examination, 80 to 90% of patients who die with severe TBI have histological features of hypoxia and/or ischemia. 16 Several well-identified intracranial factors, which include hematomas, cerebral edema, and raised ICP, all contribute to posttraumatic ischemic damage by reducing CBF. 28 Arterial hypotension and arterial hypoxia are common systemic causes of cerebral ischemia in these patients. 9 The neurochemical response to a cerebral metabolic mismatch is an increase in dialysate lactate, which may indicate a shift from aerobic to anaerobic metabolism. Under baseline conditions, brain lactate output is suppressed by the presence of sufficient brain tissue O 2. 1,21 Lactate accumulation represents a slowing of the citric acid cycle metabolic pathway within the mitochondria in favor of anaerobic utilization of glucose. Pyruvate, a product of the glycolytic breakdown of glucose, undergoes anaerobic conversion to lactate by lactate dehydrogenase. 22 During this process, the reduced form of nicotinamide adenine dinucleotide is reoxidized to create the oxidized form of nicotinamide adenine dinucleotide, which is necessary for the continuation of glucose breakdown to pyruvate. Thus, an O 2 delivery/demand mismatch, as it occurs in the first hours after severe TBI, originates glycolytic accumulation of lactate. 11 Increased levels of lactate in the dialysate or even in cerebrospinal fluid in severely head injured patients, have been shown to correlate with poor outcome or clinical deterioration. 12,15,26,27,36 As expected from other studies, a significant rise in lactate levels was seen after FPI. An immediate dramatic increase in lactate was seen, however, once arterial hypoxia was introduced. This may mean that normoxia and a normal O 2 supply is essential for good outcome in patients after TBI. 8 Hypoxia may further lead to astrocyte swelling that lengthens the diffusion distances between capillaries and mitochondria, thus reducing substrate delivery still more. The decrease in cerebral oxidative metabolism after TBI depends on the severity of the injury. 17,19 In noncomplicated FPI, the increased glucose metabolism rapidly resolves, and by 6 hours postinjury, glucose metabolism actually becomes reduced compared with baseline values. 1,20,23 Similarly, positron emission tomography studies have demonstrated an increased regional and global cerebral FIG. 6. Electron photomicrographs showing the mitochondrial ultrastructure of brain tissue samples obtained in representative areas at various times. Original magnification Left: At the end of the baseline period, normal cell structures are seen, and the mitochondria appear normal. Center: Two hours after FPI, swollen mitochondria and a few relatively normal mitochondria are seen. Right: After arterial hypoxia, massive swelling of mitochondria is evident. 647

6 A. Zauner, et al. glucose metabolism after severe head injury in humans. 2 The transient increase in the metabolic requirement after TBI is the result of marked ionic fluxes caused by mechanical injury and by the release of excitatory amino acids. Seizure activity may also play a role. In our study, the injury was probably not severe enough to result in a marked decrease in oxidative metabolism, and thus the glucose levels did not fall as expected. Oxidative metabolism was substantially impaired, however, when O 2 depletion was induced by arterial hypoxia. At that moment of the experiment, glucose levels decreased abruptly, indicating that the injured brain is particularly at risk during hypoxia. Oxygen consumption, a direct indicator of oxidative metabolism, depends on CBF and on the metabolic requirements of brain tissue. The Cartesian diver method, which is used to measure O 2 consumption of normal or injured brain tissue, 25 was used to evaluate the time course of O 2 consumption after TBI and arterial hypoxia. In contrast with other animal models, in our study we failed to demonstrate an increase in O 2 consumption following TBI, and this correlates with the lack of change in dialysate glucose levels after FPI. In other studies, however, an increase in O 2 consumption was reported within 1 hour after TBI 25 or during a transient hypermetabolic state in animal 33 as well as in human studies. 2 As expected after the induction of arterial hypoxia in this study, however, O 2 consumption decreased significantly because of mitochondrial damage, as seen in the sections prepared for ultrastructural studies (Fig. 6). There was significantly more mitochondrial swelling and damage after FPI and a 40-minute period of hypoxia. To measure O 2 delivery, we used a multiparameter sensor, a technology extensively validated in previous studies. 13,18,34,35 Measurements of brain tissue PO 2 have shown low PO 2 levels in approximately 25 to 30% of patients with severe TBI in the first 12 hours postinjury. 14,32,33 Furthermore, under normoxic conditions, tissue oxygenation was directly related to CBF and inversely related to dialysate lactate and glutamate levels in these studies. Thus, this technology provides a reliable indirect measure of local substrate delivery. In this study, extracellular glutamate levels were markedly increased after TBI. This accords closely with the results in our human studies. 6,7,24 The highest glutamate levels in human studies were seen when intracranial hematomas or secondary ischemic insults were present, as in this animal study. 5,7 Besides inducing ionic shifts and thus increasing cerebral metabolism, it is hypothesized that glutamate release after TBI is in part responsible for driving glycolysis. 1 An excessive glutamate release leads to an increase in lactate, although the exact mechanism for this increase is unknown and may be attributed either directly to an effect of glutamate on glycolytic enzymes or indirectly to mitochondrial blockade caused by Ca flux, which may in turn open the membrane transition pore. Our results confirm the involvement of increased glutamate release in the pathophysiological changes accompanying TBI, and support the conclusion that glutamate may emerge as the best marker of severity in both primary and secondary injury (that is, after TBI and after secondary ischemic damage). Conclusions This animal study simulated neurochemical changes in severely head injured patients, who suffer hypoxia before stabilization in the field or in the emergency room. The delivery of TBI complicated by arterial hypoxia resulted in a significant decrease in cerebral PO 2 and O 2 consumption, a decrease in dialysate levels of glucose, an increase in dialysate glutamate and lactate levels, and severe cerebral acidosis. Furthermore, we demonstrated an increase in the lactate/glucose ratio, whereas the lactate/pyruvate ratio remained unchanged. In TBI, the lactate/glucose ratio seems to be a better indicator of metabolic impairment associated with the injury. In pure ischemic injury, such as middle cerebral artery occlusion, the opposite is seen. The complexity of pathophysiological changes and their interactions after TBI might explain why specific therapeutic attempts aimed at the normalization of only one component have failed to improve outcome in severely head injured patients. This experiment provides evidence that after TBI complicated by hypoxia, the most useful and sensitive analytes for FPI alone were glutamate and brain tissue PO 2, whereas the lactate/glucose ratio was the most specific indicator for secondary hypoxic injury. Overall, hypoxia leads to further deterioration of the measured metabolic parameters, thus explaining in part the poor outcome seen in patients with TBI when it occurs. Acknowledgment Special thanks to Jiepei Zhu, M.D., Ph.D., Medical College of Virginia, Virginia Commonwealth University, for excellent work with the electron microscope. Disclaimer None of the authors has a financial interest in the equipment or procedures discussed in this article. References 1. Andersen BJ, Marmarou A: Post-traumatic selective stimulation of glycolysis. Brain Res 585: , Bergsneider M, Hovda DA, Shalmon E, et al: Cerebral hyperglycolysis following severe traumatic brain injury in humans: a positron emission tomography study. J Neurosurg 86: , Bouma GJ, Muizelaar JP, Choi SC, et al: Cerebral circulation and metabolism after severe traumatic brain injury: the elusive role of ischemia. J Neurosurg 75: , Bouma GJ, Muizelaar JP, Stringer WA, et al: Ultra-early evaluation of regional cerebral blood flow in severely head-injured patients using xenon-enhanced computerized tomography. J Neurosurg 77: , Brown JI, Baker AJ, Konasiewicz SJ, et al: Clinical significance of CSF glutamate concentrations following severe traumatic brain injury in humans. J Neurotrauma 15: , Bullock R, Zauner A, Myseros JS, et al: Evidence for prolonged release of excitatory amino acids in severe human head trauma. Relationship to clinical events. Ann NY Acad Sci 765: , Bullock R, Zauner A, Woodward JJ, et al: Factors affecting excitatory amino acid release following severe human head injury. J Neurosurg 89: , Cherian L, Robertson CS, Goodman JC: Secondary insults increase injury after controlled cortical impact in rats. J Neurotrauma 13: ,

7 Cerebral metabolism after traumatic brain injury and hypoxia in cats 9. Chestnut RM, Marshall LF, Klauber MR, et al: The role of secondary brain injury in determining outcome from severe head injury. J Trauma 34: , Cohen PJ, Alexander SC, Smith TC, et al: Effects of hypoxia and normocarbia on cerebral blood flow and metabolism in conscious man. J Appl Physiol 23: , Combs DJ, Dempsey RJ, Maley M, et al: Relationship between plasma glucose, brain lactate, and intracellular ph during cerebral ischemia in gerbils. Stroke 21: , DeSalles AAF, Kontos HA, Becker DP, et al: Prognostic significance of ventricular CSF lactic acidosis in severe head injury. J Neurosurg 65: , Doppenberg EMR, Zauner A, Bullock R, et al: Correlation between brain tissue oxygen tension, carbon dioxide tension, ph and cerebral blood flow a better way of monitoring the severely injured brain? Surg Neurol 49: , Doppenberg EMR, Zauner A, Watson JC, et al: Determination of the ischemic threshold for brain oxygen tension. Acta Neurochir Suppl 71: , Goodman JC, Valadka AB, Gopinath SP, et al: Extracellular lactate and glucose alterations in the brain after head injury measured by microdialysis. Crit Care Med 27: , Graham DI, Ford I, Adams JH, et al: Ischaemic brain damage is still common in fatal non-missile head injury. J Neurol Neurosurg Psychiatry 52: , Hovda DA, Yoshino A, Kawamata T, et al: Diffuse prolonged depression of cerebral oxidative metabolism following concussive brain injury in the rat: a cytochrome oxidase histochemistry study. Brain Res 567:1 10, Hutchinson PJ, Al-Rawi PG, O Connell MT, et al: Monitoring of brain metabolism during aneurysm surgery using microdialysis and brain multiparameter sensors. Neurol Res 21: , Inao S, Marmarou A, Clark GD, et al: Production and clearance of lactate from brain tissue, cerebrospinal fluid, and serum following experimental brain injury. J Neurosurg 69: , Kawamata T, Katayama Y, Hovda DA, et al: Lactate accumulation following concussive brain injury: the role of ionic fluxes induced by excitatory amino acids. Brain Res 674: , Kraig RP, Petito CK, Plum F, et al: Hydrogen ions kill brain at concentrations reached in ischemia. J Cereb Blood Flow Metab 7: , Krebs EG: Protein kinases. Curr Top Cell Regul 5:99 133, Krishnappa IK, Contant CF, Robertson CS: Regional changes in cerebral extracellular glucose and lactate concentrations following severe cortical impact injury and secondary ischemia in rats. J Neurotrauma 16: , Kuroda Y, Inglis FM, Miller JD, et al: Transient glucose hypermetabolism after acute subdural hematoma in the rat. J Neurosurg 76: , Levasseur JE, Alessandri B, Reinert M, et al: Fluid percussion injury transiently increases then decreases brain oxygen consumption in the rat. J Neurotrauma 17: , Murr R, Stummer W, Schurer L, et al: Cerebral lactate production in relation to intracranial pressure, cranial computed tomography findings, and outcome in patients with severe head injury. Acta Neurochir 138: , Persson L, Hillered L: Chemical monitoring of neurosurgical intensive care patients using intracerebral microdialysis. J Neurosurg 76:72 80, Schroder ML, Muizelaar JP, Bullock MR, et al: Focal ischemia due to traumatic contusions documented by stable xenon-ct and ultrastructural studies. J Neurosurg 82: , Siesjo BK: Cell damage in the brain: a speculative synthesis. J Cereb Blood Flow Metab 1: , Siesjo BK: Editorial: Metabolic control of intracellular ph. Scand J Clin Lab Invest 32:97 104, Siesjo BK, Ljunggren B: Cerebral energy reserves after prolonged hypoxia and ischemia. Arch Neurol 29: , van Santbrink H, Maas AIR, Avezaat CJJ: Continuous monitoring of partial pressure of brain tissue oxygen in patients with severe head injury. Neurosurgery 38:21 31, Yoshino A, Hovda DA, Kawamata AT, et al: Dynamic changes in local cerebral glucose utilization following cerebral contusions in rats: evidence of a hyper- and a subsequent hypometabolic state. Brain Res 561: , Zauner A, Bullock R, Di X, et al: Brain oxygen, CO2, ph, and temperature monitoring: evaluation in the feline brain. Neurosurgery 37: , Zauner A, Doppenberg E, Woodward JJ, et al: Multiparametric continuous monitoring of brain metabolism and substrate delivery in neurosurgical patients. Neurol Res 19: , Zauner A, Doppenberg EM, Woodward JJ, et al: Continuous monitoring of cerebral substrate delivery and clearance: initial experience in 24 patients with severe acute brain injuries. Neurosurgery 41: , 1997 Manuscript received December 3, Accepted in final form May 2, Funding for this study was provided by the National Institutes of Health Grant No. NS Address reprint requests to: Alois Zauner, M.D., Interventional Neuroradiology, University of California at Los Angeles, Box , Los Angeles, California Azauner@ mednet.ucla.edu. 649