JOURNAL OF NEUROTRAUMAVolume 18, Number 12, 2001Mary Ann Liebert, Inc.

Elevation of Extracellular Glutamate in the Final, IschemicStage of Progressive Epidural Mass Lesion in Cats

SHINGO TOYOTA, RUDOLF GRAF, CHRISTIAN DOHMEN, MARIO VALENTINO,MARTIN GROND, KLAUS WIENHARD, and WOLF-DIETER HEISS

ABSTRACT

Epidural mass lesions may cause ischemia due to progressive intracranial hypertension. In order toinvestigate the impact of intracranial pressure on accumulation of neuroactive substances, we grad-ually raised intracranial pressure in five halothane anesthetized cats by inflation of an epidural bal-loon. We evaluated in the parietal cortex contralateral to the site of balloon inflation, alterations ofextracellular glutamate and purine catabolites and of the lactate/pyruvate ratio in relation to changesof intracranial, cerebral perfusion and mean arterial blood pressure. In a complementary experi-ment, regional cerebral blood flow was assessed by sequential positron emission tomography. In thissimplified mass lesion model, extracellular glutamate increased in all cats at a late, critical stage af-ter tentorial herniation, when intracranial pressure had increased to more than 90 mm Hg, cere-bral perfusion pressure had decreased below 40–50 mm Hg. Positron emission tomography assess-ments revealed that the ischemic threshold for glutamate accumulation was in the range of 15–20mL/100 g/min. Purine catabolites and the lactate/pyruvate ratio increased somewhat earlier thanglutamate, but also after reaching the critical, terminal stage. We conclude that in this model ofprogressive epidural compression, glutamate-mediated excitotoxic processes at sites remote from theinitial focal lesion depend on processes such as delayed ischemia in combination with tentorial her-niation and systemic hypotension. These processes seem to be initiated by a decrease of cerebralperfusion pressure below a threshold of 40–50 mm Hg.

Key words: cats; cerebral perfusion pressure; intracranial pressure; microdialysis; PET; progressive masslesion

1349

INTRODUCTION

DURING THE EVOLUTION of progressive mass lesions,intracranial hypertension is an essential, life threat-

ening process that is commonly seen in stroke and neu-rotrauma causing a decrease in cerebral perfusion pres-sure (CPP) and thereby cerebral blood flow (CBF). A lowCPP may jeopardize regions of the brain primarily notinvolved due to compression of cerebral vessels, and en-hancement of CPP may help to avoid both regional andglobal ischemia (Nagao et al., 2000). Many reports rec-

ommend that CPP should be maintained above a certainvalue; however, the management of CPP and intracranialpressure (ICP) is still controversial (Rosner et al., 1995;Juul et al., 2000; Robertson et al., 1999).

Neurochemical monitoring is an excellent method toinvestigate brain diseases. With intracerebral microdial-ysis, excessive extracellular accumulation of excitatoryamino acids has been shown in animal models of bothexperimental stroke and neurotrauma in the acute phase(Shimada et al., 1989; Hillered et al., 1989; Butcher etal., 1990; Nilsson et al., 1990; Bullock et al., 1991; Tak-

MPI für neurologische Forschung, Köln, Germany.

agi et al., 1993), but it may also be involved in delayedimpairment (Matsumoto et al., 1993, 1996; Taguchi etal., 1996). Recently, microdialysis has been introducedin intensive care units as a tool for neurochemical mon-itoring in neurotrauma, subarachnoid hemorrhage, andstroke patients (Persson and Hillered, 1992; Bullock etal., 1995, 1998; Persson et al., 1996; Berger et al., 1999),especially in order to predict secondary deterioration. Therelevance of microdialysis as a tool for neuromonitoringhas also been documented by relating the results to re-sults obtained with regional imaging techniques such aspositron emission tomography (Enblad et al., 1996).Epidural hematoma represents one more important ex-ample of progressive mass lesion (McNealy and Plum,1962). For this pathology, the relevance of ICP elevationon alteration of extracellular substrate concentrations isunclear, and experimental models have not been tested.We therefore examined in a simplified brain compressionmodel produced by epidural balloon inflation in cats al-terations of neurochemical substances as related to ICP,CPP, and CBF alterations.

MATERIALS AND METHODS

Five cats weighing 4.3–6.0 kg were used. The studywas approved by the local Animal Care Committee andthe Regierungspräsident of Cologne and is in compliancewith the German Laws for Animal Protection. Cats wereinitially anesthetized with ketamine hydrochloride (25mg/kg i.m.). After cannulating the left femoral artery andvein, the animals were tracheotomized and immobilizedwith pancuronium bromide (0.2 mg/kg i.v.). Thereafter,artificial ventilation was initiated, and anesthesia waschanged to halothane (0.6 –1.2% in a 70% nitrous ox-ide/30% oxygen gas mixture).

Intravenous infusion of Ringer solution (147.1 mM NaCl/4.0 mM KCl/2.25 mM Ca21) containing 5mg/kg/h gallamine triethiodide for muscle relaxation wasmaintained throughout the experiment. Mean arterialblood pressure (MABP) was measured continuously, ar-terial blood gases were measured intermittently. Deepbody temperature was controlled at 37.0°C using a heat-ing blanket feedback controlled by a rectal temperatureprobe.

A burr hole of 6-mm-diameter was drilled into the skullabove the right parietal region to insert a latex miniatureballoon into the epidural space. The balloon was inflatedwith water at a rate of 8ml/min using a microinfusionpump (CMA/100; Carnegie Medicine) to elevate in-tracranial pressure. A second burr hole of 6-mm-diame-ter was drilled into the skull above the marginal gyrus ofthe left cerebral cortex. Stereotactic coordinates of this

recording site were 8-mm anterior/3-mm lateral (Reinoso-Suarez, 1961). At this site, the dura was incised undermicroscopic control, and a microdialysis probe (mem-brane length 1 mm, membrane diameter 0.25 mm, cutoff6,000 Dalton) inserted into gray matter of the marginalgyrus using a micromanipulator. The depth of insertionwas adjusted to 1.5 mm. On the cortical surface of thesame site, at a distance of 3–4 mm to the microdialysisprobe, a strain-gauge MicroSensor (Codman/Johnson &Johnson Professional, Inc., Randolph, MA) measuredICP, and a thermocouple measured regional brain tem-perature. After completion of the preparation, both burrholes were completely sealed with dental cement to pre-vent CSF leakage.

Microdialysis probes were manufactured as concentrictubes with an inner and outer Silica tube, and a capillarydialysis membrane at the tip glued to the outer tube(Cuprophan, Akzo Nobel, Wuppertal, Germany; cutoff6,000 Dalton; diameter 250 mm; length of the activemembrane, 1 mm). The probes were continuously per-fused with artificial CSF (1.2 mM CaCl2, 145 mM NaCl,2.7 mM KCl,1.0 mM MgCl2, adjusted to pH 7.4 withphosphate buffer) at a flow rate of 1 mL/min. At this flowrate, in vitro recovery of glutamate for the probes usedin the study amounted to 17.9% 6 5.1%. For purinecatabolites and lactate/pyruvate, in vitro recovery was notspecifically determined. Recovery rates tested at the sameperfusion rate for the type of microdialysis probes usedin the present experiments were in the same range (de-viation , 15%) as those for glutamate. Amino acids wereanalyzed in 20-min samples with a RF-535 fluorescencedetector (Shimadzu, Kyoto, Japan; excitation wavelength330 nm, emission wavelength 480 nm) after separationon a 5-mm Adsorbosphere OPA HS column (Alltech,München, Germany; 100 3 4.6 mm) using a gradient elu-tion profile (buffer A: 50 mM sodium acetate, pH 5.7;buffer B: 100% methanol) following precolumn deriva-tization with o-phthaldialdehyde (Shimada et al., 1993;with modifications). Purine catabolites including adeno-sine and hypoxanthine were analyzed with an ultravioletdetector (Knauer, Berlin, Germany) at a wavelength of254 nm after separation on a 3-mm C18-Nucleosil 100column (60 3 4 mm; Knauer) using an elution system of10 mmol/L ammonium dihydrogen phosphate/6%methanol (pH 4.2), with a flow rate of 1.0 mL/min (Mat-sumoto et al., 1992, with modifications). Metabolites ofglycolysis (lactate and pyruvate) were assessed with anultraviolet detector (Knauer) at a wavelength of 214 nmafter separation on a 5-mm Ultrapac ODS-120T column(250 3 4.6 mm; LKB,Bromma, Sweden) using an elu-tion solvent of 100 mmol/L sodium dihydrogen phos-phate (pH 2.3), with a flow rate of 1.0 mL/min (Staub etal., 2000).

TOYOTA ET AL.

1350

One out of the five described experiments was performed in a clinical high-resolution PET camera(Siemens/CTI ECAT EXACT HR) to additionally obtainsequential regional determinations of cerebral bloodflow. The method has been described before (Heiss et al.,1997). In brief, the scanner has a field of view of 15 cm,an in-plane spatial resolution of 3.6 mm full width at halfmaximum (FWHM), and an axial resolution of 4.0 mmFWHM (Wienhard et al., 1994). The animal was posi-tioned and kept in the scanner gantry throughout the en-tire experiment to guarantee positional stability. CBF wasdetermined after i.v. bolus injection of 20 mCi 15O-la-beled water. Experimental background and limitations ofthis method for measurement of cerebral hemodynamicshave been previously discussed (Baron et al., 1989). Thecat used in the PET experiment was otherwise treated inthe same way as the rest of the animals included in thestudy.

After a 4-h stabilization period, control microdialysissamples were collected. Thereafter, inflation of theepidural balloon was started and continued until CPP de-creased below 0 mm Hg. The various parameters (ICP,blood pressure, brain temperature, EtCO2) were contin-uously recorded and analyzed using a PC-based data acquisition system (DASY LAB, DATALOG, Mön-chengladbach, Germany). CPP was calculated online inthe acquisition system using the formula: CPP 5

MABP 2 ICP. For statistical analysis, data were nor-malized in time categories and calculated as means 6

SD. Statistical evaluation was performed using one wayanalysis of variance (ANOVA) with posthoc comparison(ScheffÈ) to test significance (p , 0.05) within groups(Statistica; StatSoft, USA).

RESULTS

Physiological parameters including PaO2, PaCO2, andpH were kept within physiological ranges during controland 60 min after balloon inflation (Table 1). Deep bodytemperature was kept at 37.0°C.

As shown in an actual recording in Figure 1, inflationof the balloon resulted in an almost immediate slow andthen progressively increasing rise of ICP measured on theside contralateral to the balloon. In this particular exper-iment, this phase lasted almost 150 min. Thereafter, theICP rise suddenly accelerated peaking at about 120 mmHg, remained at this high level for 10–20 min, anddropped thereafter to about 70 mm Hg. Right pupillarydilatation was observed 126 min after onset of ballooninflation. MABP, in contrast to ICP, remained relativelystable over the first 150 min, increasing slightly duringthe last 20–30 min of this phase. Corresponding to thesteep rise of ICP, MABP also increased sharply after 150min to almost 180 mm Hg and subsequently decreasedbelow control levels. As a consequence of a progressivelydecreasing ICP and an almost unchanged MABP duringthe early phase, CPP gradually decreased to about 80 mmHg during the primary phase of balloon inflation until, atthe time point of the steep ICP rise, it began to drop morerapidly finally reaching almost 0 mm Hg. On the con-trary, extracellular substrates measured by microdialy-sis/HPLC were altered only in the very late phase of theexperiment after the latest steep increase of ICP. Adeno-sine, presumably as an intermediate product of ATP de-pletion, increased more in the beginning than did the ex-citatory amino acid glutamate. The adenosine elevationleveled off at about 2 mM, while glutamate continued toincrease throughout the observation period reaching toabout 25 mM.

In all animals, similar courses of alterations of the var-ious parameters were obtained. The mean value of theICP peak following the late steep rise was 111.4 6 13.3mm Hg, and the mean time point of this peak was172.2 6 45 min after balloon inflation. At this time, bal-loons had been inflated to 1.38 6 0.36 mL. For summa-rizing time course alterations, data were standardized intime categories by dividing the individual times betweenstart of balloon inflation and ICP peak into 10 aliquotsand thereby using these aliquots (14 in total) as catego-rized time units (mean: 17.2 6 4.3 min; Fig. 2). Basi-cally, the summarized data confirmed the observations

GLUTAMATE RISE IN FINAL STAGE OF MASS LESION

1351

TABLE 1. PHYSIOLOGICAL DATA

pH pCO2 (mm Hg) pO2 (mm Hg)

Control before balloon inflationMean 7.368 28.5 145.5SD 0.014 2.8 16.4

60 min after inflationMean 7.349 29.7 138.5SD 0.026 2.5 18.6

described above for the exemplified experiment. Statis-tical treatment of the data was in so far interesting as itallowed to compare the time of significant alterationagainst control (time category 0) among the multiple pa-rameters. Following such consideration, changes of ICPand CPP (time category 6) were earliest. MABP (timecategory 10), adenosine, hypoxanthine (time category11), and finally glutamate changes (time category 12)were characteristic for the late, critical phase of the ex-periments. Increases in lactate/pyruvate ratio (not shown)followed a time course pattern similar to that of thepurines and increases of other neuroactive amino acidssuch as aspartate and GABA (not shown) followed a pat-tern similar to that of glutamate.

To better understand the significance of CPP reduction

on extracellular glutamate elevation during balloon in-flation, these parameters were plotted against each other(Fig. 3). A threshold-like correlation was obtained for thetwo parameters. The CPP threshold for glutamate eleva-tion amounted to 40–50 mmHg.

To show the impact of gradual ICP rise on regionalcerebral blood flow, sequential PET determinations wereperformed in a complementary experiment. As shown inconsecutive images of one coronal plane of an individ-ual cat (Fig. 4, top panel), CBF gradually decreased inthe right parietal region documenting the effect of pro-gressive balloon inflation at this site. The effect of bal-loon inflation on CBF in the other brain regions wasslower, and only in the last three images, almost generalcessation of CBF was obtained. Quantification of this

TOYOTA ET AL.

1352

FIG. 1. Plots of actual recordings obtained in a single animal in marginal gyrus contralateral to the inflated balloon. Note rightpupillary dilation (marked by an eye symbol) 120–130 min after balloon inflation, the abrupt increase of ICP around 150 min af-ter inflation, the subsequent elevation of MABP presumably representing a Cushing reflex, and the concurrent sharp decrease ofCPP. Only after this event, adenosine and glutamate accumulated in the extracellular space. ICP, intracranial pressure; MABP,mean arterial pressure; CPP, cerebral perfusion pressure.

process was performed by analyzing in five consecutivecoronal planes circular regions of interest in the parietalregion of the hemisphere contralateral to the balloon (Fig.4, bottom panel). Extracellular glutamate elevation wasfinally observed when PET-CBF fell below 20 mL/100g/min.

DISCUSSION

Studies on intracranial hypertension induced by con-trolled expansion of an epidural balloon have been startedalready in the 1970s, and multiple parameters have mean-while been investigated in such experimental models(Hekmatpanah, 1970a,b; Lewis and McLaurin, 1972;Sullivan et al., 1977; Jakobsson et al., 1990; Wolfla etal., 1996; Nagai et al., 1997; Burger et al., 1999; Nagaoet al., 2000). The results of the present study regardingICP, CPP, and CBF alterations basically confirm formerstudies. In particular, the decline of ICP after progressiveincrease during balloon inflation, and the concomitant de-

crease of MABP can be interpreted as a disturbance ofmedullar function that is triggered by herniation of supra-tentorial brain tissue into infratentorial space. It is there-fore believed that the drop of ICP and MABP marks theterminal stage of herniation.

The main purpose of our study was to compare mon-itoring techniques such as ICP and MABP recording withneurochemical monitoring by microdialysis regardingtheir ability to indicate this terminal, devastating stage ofherniation at an early time point during progressive in-tracranial hypertension.

In traumatic cerebral ischemia, prolonged duration ofmass lesions on the brain and thereby increased ICP areimportant factors. A low CPP may jeopardize regions ofthe brain due to compression of cerebral vessels, and en-hancement of CPP may help to avoid both global and re-gional ischemia (Nagao et al., 2000). MABP has also amajor impact on CPP. After induction of cerebral is-chemia, blood flow to periischemic tissue is linearly re-lated to perfusion pressure, and even small reductions inarterial blood pressure increase the severity of the is-

GLUTAMATE RISE IN FINAL STAGE OF MASS LESION

1353

FIG. 2. Changes in multiple parameters in marginal gyri contralateral to the inflated balloon (n 5 5) as a function of time afterballoon inflation. For summarizing time course alterations, data were standardized in time categories by dividing the individual timesbetween start of balloon inflation and ICP peak into 10 aliquots and by using these aliquots (14 in total) as categorized time units(mean: 17.2 6 4.3 min). Means 6 SD are plotted. *p , 0.05: significantly different from baseline values at time point 0.

chemic insult (Hossmann, 1982). In the present study,both ICP and CPP monitoring revealed significant alter-ations in a rather early stage (Fig. 2, time point 6). It hasbeen reported that in clinical neurotrauma, the prognosisof long lasting low CPP is detrimental, and many reportsrecommend that CPP should be maintained above a cer-tain value, notwithstanding that the management of CPPand ICP is still controversial (Rosner et al., 1995; Juul etal., 2000; Robertson et al., 1999). In our noncontusiontrauma model, a management of CPP seems desirable toprevent secondary ischemia. The rise of adenosine andhypoxanthine or glutamate at a CPP of around 50 mmHg would demark the start of the critical stage of the dis-order. CPP of less than 50mm Hg is thought to fulfill cri-teria of an ischemic event if it is held for more than 30min. In such a condition, glutamate has been shown torise in noncontusion neurotrauma patients (Bullock et al.,1998). However, management of ICP and CPP shouldobviously be started earlier to prevent onset of such stage.Since in our study, levels of significant elevation of glu-tamate and adenosine were in the range of 20 mm Hg forICP and 100 mm Hg for CPP, we would suggest to con-sider interventional steps if the respective parameters ex-ceed or fall below these levels.

Our study demonstrates that in this model of epiduralmass lesion, glutamate does not rise before herniation.Other relevant amino acids such as aspartate or GABA(analyzed in the same study but not shown here) are byno means faster than those of glutamate. Both the purineand the lactate/pyruvate ratios seemed to rise moresteeply and perhaps somewhat earlier than glutamate, but

also after herniation. Since increases of purines and ofthe lactate/pyruvate ratio most probably reflect metabolicdeterioration, these elevations indicate the onset of a crit-ical ischemia in the investigated tissue compartment. Thelactate/pyruvate ratio has been shown to particularly pos-sess high sensitivity for ischemia and for metabolic dis-turbances in brain tissue (Enblad et al., 1996; Persson etal., 1996). The results are also in agreement with studiesin focal cerebral ischemia in which higher cerebral bloodflow thresholds for extracellular accumulation of adeno-sine than for glutamate have been shown (Matsumoto etal., 1992).

Interestingly, glutamate increase in the PET experi-ment occurred at a time when CBF values determined byPET had decreased to 15–20 mL/100 g/min. Taking intoaccount that partial volume effects characteristic for PETstudies result in slightly underestimated CBF values be-cause white matter compartments with lower blood floware included in the volume of interest (Heiss et al., 1995),our results come close to the ischemic flow threshold of20 mL/100 g/min for glutamate elevation determined ina former study on global ischemia (Shimada et al., 1989).

The microdialysis technique has recently been appliedin stroke patients in the subacute stage as a tool for mon-itoring and possibly predicting secondary tissue deterio-ration and malignant edema formation remote from theprimary lesion (Berger et al., 1999; Bullock et al., 1995).In one of these case reports (Bullock et al., 1995), sus-tained high levels of glutamate were obtained in thedialysate. In this example, the dialysis membrane was lo-cated in the ischemic core. Presumably, with such posi-

TOYOTA ET AL.

1354

FIG. 3. Scatterplots of the relation between cerebral perfusion pressure (CPP) and glutamate elevation. Data were obtainedfrom start of epidural balloon inflation till the final peak of intracranial pressure (ICP; see Fig. 1 and 2). 20-min data were plot-ted. For this purpose, 1-sec samples of CPP obtained by the data acquisition system were averaged over the respective 20-minepisodes of microdialysis sampling.

tioning, a predictor function for secondary damage is notto be expected, since information on the size of the pri-mary lesion is not available from local recordings. In theother case report (Berger et al., 1999), however, neuro-chemical monitoring was performed in the hemispherecontralateral to the primary ischemic lesion. In this in-stance, elevation of glutamate was interpreted to havepredictor function for malignant infarction. It is obvious,therefore, that the location of microdialysis membranesneeds to be carefully defined.

In our study, glutamate elevations were not obtainedin remote, contralateral regions until the terminal stageof ICP elevation and CBF reduction. Thus, increased ICPalone is not sufficient to cause substantial elevations ofextracellular glutamate and of substrates of energy me-tabolism to make them efficient early indicators of sec-ondary damage. As to monitoring neurotrauma, our re-sults are in good agreement with clinical reports showing

the better prognosis of acute epidural as compared toacute subdural hematoma. In an acute subdural hematomamodel in rats, hippocampal glutamate was found to risesignificantly, even though regional CBF was preserved(Bullock et al., 1991). This suggests that, particular insubdural derangements, glutamate-mediated excitotoxicprocesses may be initiated in the absence of notable sec-ondary ischemia. It needs to be tested as to whether inmodels of experimental focal ischemia and in stroke pa-tients, factors other than intracranial hypertension, butpossibly also associated with progressive edema forma-tion, induce such early alterations of extracellular sub-strate concentrations then to be detected by neurochem-ical monitoring. Microdialysis studies in a larger numberof patients, and perhaps in different locations are neededto shed more light as to the processes leading finally tosecondary damage.

In conclusion, in this model of epidural mass lesion,

GLUTAMATE RISE IN FINAL STAGE OF MASS LESION

1355

FIG. 4. Top panel: Sequential PET images of cerebral blood flow obtained in one coronal plane of an individual cat (the sameexperiment as shown in Fig. 1). Note the progressive decrease of CBF in the right parietal region. At this site, the epidural bal-loon was positioned and inflated. Note that only in the last three images of the sequence, CBF in other brain regions also re-markably dropped. Bottom panel: Time course of quantified regional CBF obtained by PET (PET-CBF). Data for five consec-utive coronal planes in circular regions of interest in the left parietal cortex (diameter: 6 mm) were analyzed. Means 6 SD areplotted. For comparison, the cerebral perfusion pressure (CPP) and the extracellular glutamate elevation shown in Fig. 1 are againpresented.

glutamate-mediated excitotoxic processes at sites remotefrom the initial lesion may depend on factors associatedwith other processes such as progressive edema forma-tion, but up to a terminal stage, they probably do not de-pend on intracranial hypertension. As indicators for anearly management of epidural mass lesions, more classi-cal measures such as ICP or CPP may serve supplemen-tary function better, as do imaging techniques such asPET and MRI, since they are capable of assessing notonly the intensity of functional disturbance, but also itsregional distribution.

ACKNOWLEDGMENTS

We would like to thank P. Gabel and D. Lattacz forvaluable technical assistance. The work was supportedby a grant of the Bundesministerium für Bildung undForschung (Kompetenznetzwerk Schlaganfall) to R.Graf, M. Grond, and W.-D. Heiss.

REFERENCES

BARON, J.C , FRACKOWIAK R. S. J., HERHOLZ K., et al.(1989). Use of PET methods for measurement of cerebral en-ergy metabolism and hemodynamics in cerebrovascular dis-ease. J. Cereb. Blood Flow Metab. 9, 723–742.

BERGER, C., ANNECKE, A., ASCHOFF, A., et al. (1999).Neurochemical monitoring of fatal middle cerebral artery in-farction. Stroke 30, 460–463.

BULLOCK, R., BUTCHER, S.P., CHEN, M.H., et al. (1991).Correlation of the extracellular glutamate concentration withextent of blood flow reduction after subdural hematoma inthe rat. J. Neurosurg. 74, 794–802.

BULLOCK, R., ZAUNER, A., WOODWARD, J., et al. (1995).Massive persistent release of excitatory amino acids follow-ing human occlusive stroke. Stroke 26, 2187–2189.

BULLOCK, R., ZAUNER, A., WOODWARD, J.J., et al. (1998).Factors affecting excitatory amino acid release following se-vere human head injury. J. Neurosurg. 89, 507–518.

BURGER, R., VINCE, G.H., MEIXENSBERGER, J., et al.(1999). Interrelations of laser Doppler flowmetry and braintissue oxygen pressure during ischemia and reperfusion in-duced by an experimental mass lesion. J. Neurotrauma 16,1149–1164.

BUTCHER, S.P., BULLOCK, R., GRAHAM, D.I., et al.(1990). Correlation between amino acid release and neu-ropathologic outcome in rat brain following middle cerebralartery occlusion. Stroke 21, 1727–1733.

ENBLAD, P., VALTYSSON, J., ANDERSSON, J., et al. (1996).Simultaneous intracerebral microdialysis and positron emis-sion tomography in the detection of ischemia in patients with

subarachnoid hemorrhage. J. Cereb. Blood Flow Metab. 16,637–644.

HEKMATPANAH, J. (1970a). Cerebral circulation and perfu-sion in experimental increased intracranial pressure. J. Neu-rosurg. 32, 21–29.

HEKMATPANAH, J. (1970b). The sequence of alterations inthe vital signs during acute experimental increased intracra-nial pressure. J. Neurosurg. 32, 16–20.

HEISS, W.D., WIENHARD, K., GRAF, R., et al. (1995). High-resolution PET in cats: application of a clinical camera to ex-perimental studies. J. Nucl. Med. 36, 493–498.

HEISS, W.-D., GRAF, R., LÖTTGEN, J., et al. (1997). RepeatPET Studies in transient MCA occlusion in cats: residual per-fusion and efficacy of postischemic reperfusion. J. Cereb.Blood Flow Metab. 17, 388–400.

HILLERED, L., HALLSTROM, A., SEGERSVARD, S., et al.(1989). Dynamics of extracellular metabolites in the striatumafter middle cerebral artery occlusion in the rat monitored byintracerebral microdialysis. J. Cereb. Blood Flow Metab. 9,607–616.

HOSSMANN, K.A. (1982). Treatment of experimental cere-bral ischemia. J. Cereb. Blood Flow Metab. 2, 275–297.

JAKOBSSON, K.E., LOFGREN, J., ZWETNOW, N.N., et al.(1990). Cerebral blood flow in experimental intracranial masslesions. Part I: The compression phase. Neurol. Res. 12,147–152.

JUUL, N., MORRIS, G.F., MARSHALL, S.B., et al. (2000).Intracranial hypertension and cerebral perfusion pressure: in-fluence on neurological deterioration and outcome in severehead injury. The Executive Committee of the InternationalSelfotel Trial. J. Neurosurg. 92, 1–6.

LEWIS, H.P., and MCLAURIN., R.L. (1972). Regional cere-bral blood flow in increased intracranial pressure producedby increased cerebrospinal fluid volume, intracranial massand cerebral edema, in: Intracranial Pressure. M. Brock andH. Dietz (eds), Springer-Verlag: Berlin, pps. 160–164.

MATSUMOTO, K., GRAF, R., ROSNER, G., et al. (1992).Flow thresholds for extracellular purine catabolite elevationin cat focal ischemia. Brain Res. 579, 309–314.

MATSUMOTO, K., GRAF, R., ROSNER, G., et al. (1993). El-evation of neuroactive substances in the cortex of cats dur-ing prolonged focal ischemia. J. Cereb. Blood Flow Metab.13, 586–594.

MATSUMOTO, K., LO, E.H., PIERCE, A.R., et al. (1996).Secondary elevation of extracellular neurotransmitter aminoacids in the reperfusion phase following focal cerebral is-chemia. J. Cereb. Blood Flow Metab. 16, 114–124.

MCNEALY, D., and PLUM, F. (1962). Brainstem dysfunctionwith supratentorial mass lesions. Arch. Neurol. 7, 26–48.

NILSSON, P., HILLERED, L., PONTEN, U., et al. (1990)Changes in cortical extracellular levels of energy-related

TOYOTA ET AL.

1356

metabolites and amino acids following concussive brain in-jury in rats. J. Cereb. Blood Flow Metab. 10, 631–637.

NAGAI, H., MORITAKE, K., and TAKAYA, M. (1997). Cor-relation between transcranial Doppler ultrasonography andregional cerebral blood flow in experimental intracranial hy-pertension. Stroke 28, 603–607.

NAGAO, S., BEMANA, I., KURATANI, H., et al. (2000). Ni-ravoline, a selective kappa-opioid receptor agonist effectivelyreduces elevated intracranial pressure. Exp. Brain Res. 130,338–344.

PERSSON, L., and HILLERED, L. (1992). Chemical monitor-ing of neurosurgical intensive care patients using intracere-bral microdialysis. J. Neurosurg. 76, 72–80.

PERSSON, L., VALTYSSON, J., ENBLAD, P., et al. (1996).Neurochemical monitoring using intracerebral microdialysisin patients with subarachnoid hemorrhage. J. Neurosurg. 84,606–616.

REINOSO-SUÁREZ, F. (1961) Topographischer Hirnatlas derKatze. Darmstadt, E. Merck AG: Germany.

ROBERTSON, C.S., VALADKA, A.B., HANNAY, H.J., et al.(1999). Prevention of secondary ischemic insults after severehead injury. Crit. Care Med. 27, 2086–2095.

ROSNER, M.J., ROSNER, S.D., and JOHNSON, A.H. (1995).Cerebral perfusion pressure: management protocol and clin-ical results. J. Neurosurg. 83, 949–962.

SHIMADA, N., GRAF, R., ROSNER, G., et al. (1989). Is-chemic flow threshold for extracellular glutamate increase incat cortex. J. Cereb. Blood Flow Metab. 9, 603–606.

SHIMADA, N., GRAF, R., ROSNER, G., et al. (1993) Is-chemia-induced accumulation of extracellular amino acids incerebral cortex, white matter, and cerebrospinal fluid. J. Neu-rochem. 60, 66–71.

STAUB , F., GRAF, R., GABEL, P., et al. (2000). Multiple in-terstitial substances measured by microdialysis in patientswith subarachnoid hemorrhage. Neurosurgery 47, 1106–1116.

SULLIVAN, H.G., MILLER, J.D., BECKER, D.P., et al.(1977). The physiological basis of intracranial pressurechange with progressive epidural brain compression. An ex-perimental evaluation in cats. J. Neurosurg. 47, 532–550.

TAGUCHI, J., GRAF, R., ROSNER, G., et al. (1996). Pro-longed transient ischemia results in impaired CBF recoveryand secondary glutamate accumulation in cats. J. Cereb.Blood Flow Metab. 16, 271–279.

TAKAGI, K., GINSBERG, M.D., GLOBUS, M.Y., et al.(1993). Changes in amino acid neurotransmitters and cere-bral blood flow in the ischemic penumbral region followingmiddle cerebral artery occlusion in the rat: correlation withhistopathology. J. Cereb. Blood Flow Metab. 13, 575–585.

WIENHARD, K., DAHLBOM, M., ERIKSSON, L., et al.(1994). The ECAT EXACT HR: performance of a new highresolution positron scanner. J. Comput. Assist. Tomogr. 18,110–118.

WOLFLA, C.E., LUERSSEN, T.G., BOWMAN, R.M., et al.(1996). Brain tissue pressure gradients created by expandingfrontal epidural mass lesion. J. Neurosurg. 84, 642–647.

Address reprint requests to:Rudolf Graf, Ph.D.

MPI für Neurologische ForschungGleueler Str. 50

50931 Köln, Germany

E-mail: rudolf.graf@pet.mpin-koeln.mpg.de

GLUTAMATE RISE IN FINAL STAGE OF MASS LESION

1357