BTB improvement after dexamethasone treatment. 11 This has led to the theory that dexamethasone acts

Transcription

1 J Neurosurg 90: , 1999 Early changes measured by magnetic resonance imaging in cerebral blood flow, blood volume, and blood-brain barrier permeability following dexamethasone treatment in patients with brain tumors LEIF ØSTERGAARD, M.D., FRED H. HOCHBERG, M.D., JAMES D. RABINOV, M.D., A. GREGORY SORENSEN, M.D., MICHAEL LEV, M.D., LYNDON KIM, M.D., ROBERT M. WEISSKOFF, PH.D., R. GILBERTO GONZALEZ, M.D., PH.D., CARSTEN GYLDENSTED, M.D., PH.D., AND BRUCE R. ROSEN, M.D., PH.D. Massachusetts General Hospital Nuclear Magnetic Resonance Center, Department of Radiology, Massachusetts General Hospital, Charlestown, Massachusetts; Harvard Medical School, Boston, Massachusetts; Harvard Massachusetts Institute of Technology Division of Health Sciences and Technology, Cambridge, Massachusetts; Department of Neurology, Massachusetts General Hospital, Boston, Massachusetts; and Department of Neuroradiology and Positron Emission Tomography Center, Århus University Hospital, Århus, Denmark Object. In this study the authors assessed the early changes in brain tumor physiology associated with glucocorticoid administration. Glucocorticoids have a dramatic effect on symptoms in patients with brain tumors over a time scale ranging from minutes to a few hours. Previous studies have indicated that glucocorticoids may act either by decreasing cerebral blood volume (CBV) or blood-tumor barrier (BTB) permeability and thereby the degree of vasogenic edema. Methods. Using magnetic resonance (MR) imaging, the authors examined the acute changes in CBV, cerebral blood flow (CBF), and BTB permeability to gadolinium-diethylenetriamine pentaacetic acid after administration of dexamethasone in six patients with brain tumors. In patients with acute decreases in BTB permeability after dexamethasone administration, changes in the degree of edema were assessed using the apparent diffusion coefficient of water. Conclusions. Dexamethasone was found to cause a dramatic decrease in BTB permeability and regional CBV but no significant changes in CBF or the degree of edema. The authors found that MR imaging provides a powerful tool for investigating the pathophysiological changes associated with the clinical effects of glucocorticoids. KEY WORDS magnetic resonance imaging blood-tumor barrier permeability cerebral blood flow dexamethasone brain tumor cerebral blood volume F 300 OR the past three decades, dexamethasone has been widely used in the clinical management of patients with brain tumors because of its rapid (ranging from minutes to a few hours), often dramatic clinical effects on symptoms associated with intracranial neoplasms. 9 However, the mode of action of dexamethasone remains poorly understood. Theories advanced to explain this effect are mainly based on the observed decrease in blood-brain barrier (BBB) permeability to small solutes, 12 as well as the degree of peritumoral edema 9,21 following the administration of dexamethasone. Tightening of the blood-tumor barrier (BTB), with a consequent decrease in the extravasation of plasma proteins and hence vasogenic edema, 13,21 is thought to lower intracranial pressure (ICP) and increase regional cerebral blood flow (rcbf) to reverse local ischemia. 12 Contrary to this theory, animal studies have shown that edema resorption may take place without significant BTB improvement after dexamethasone treatment. 11 This has led to the theory that dexamethasone acts primarily by restoring cerebral water homeostasis, thus reducing edema formation by increasing parenchymal resistance to fluid transport rather than tightening the BBB to passage of small solutes. Others have hypothesized that a decrease in regional cerebral blood volume (rcbv) rather than reduced edema may be the source of acute clinical improvement. 14 Dexamethasone, by inhibiting release of prostacyclin, a potent vasodilator possibly involved in the normal regulation of cerebrovascular tone, 2 is thought to cause a generalized decrease in CBV, thereby decreasing ICP and relieving symptoms. In support of this theory, positron emission tomography studies in humans have demonstrated significant decreases in rcbf and CBV in tissue contralateral to the tumor 12 hours after dexamethasone treatment. 14 The observations of changes in BBB permeability,

2 Acute effects of dexamethasone in patients with brain tumors TABLE 1 Lesion type and radiological findings in six patients with brain tumors* Case BBB Raised Time of No. Sex Tumor Type, Grade Edema Leakage CBV Exam 1 M oligodendroglioma, II/IV M astrocytoma, III/IV F oligoastrocytoma, III/IV + (+) 1 4 F astrocytoma, II/IV M primary CNS lymphoma F astrocytoma, II III/IV * Biopsy specimens graded according to the method of Daumas-Duport, et al. Also listed are the radiological findings according to MR examination. Abbreviations: + = present; = absent. Time of posttreatment examination in hours. The tumor showed slight enhancement on postcontrast MR images. However, the leakage was too small to be quantified using our technique. Diffuse large B-cell type with some T cells. CBF, CBV, and degree of edema have not yet been demonstrated in humans over a time scale comparable to that of the observed clinical improvement (a few hours). The time scale and sequence of these changes are important in understanding the acute pathophysiological changes that occur in conjunction with dexamethasone treatment in brain tumors. In this study, we used magnetic resonance (MR) imaging to detect early changes (1 6 hours posttreatment) in brain and tumor pathophysiology after dexamethasone administration by measuring CBV, CBF, and BBB permeability to gadolinium-diethylenetriamine pentaacetic acid (Gd-DTPA). In cases in which an early decrease in BBB permeability to Gd-DTPA was observed, the change in degree of edema was assessed by measuring the apparent diffusion coefficient (ADC) of water. Clinical Material and Methods Patient Population Six patients with steroid-naive brain tumors were examined to determine BBB permeability, CBV, and CBF. The participants were consecutive patients seen in our neurooncology clinic with brain tumors or brain tumor recurrence. Four patients were examined 1 hour after intravenous administration of 20 mg dexamethasone (Decadron), and two were examined 6 hours after administration of the same dose of this drug. Table 1 shows the tumor type and radiological findings in each patient. In two patients with a BBB leak (interstitial buildup of contrast agent), the ADC was measured both before and 1 hour after steroid administration to quantify changes in the degree of edema. Baseline scans were obtained 5 to 12 hours before dexamethasone administration to allow complete tracer clearance (plasma half-life 90 minutes). As a control, one additional patient with a Grade IV/IV astrocytoma who received a steady dose of oral dexamethasone (6 mg four times daily) was examined twice (with a 3-week interval) to determine the stability of our measurements. The initial patient population was medically and neurologically stable, and thus did not reveal the evolution of quantifiable neurological deficits. All examinations were performed according to a standard perfusion protocol approved by the Massachusetts General Hospital Subcommittee on Human Studies. Prior to examination, written informed consent was obtained from each patient. Imaging Protocol Dynamic Imaging for CBV and CBF Measurements. Images were obtained using a 1.5-tesla MR imager retrofitted for echoplanar imaging capabilities (Signa; General Electric Medical Systems, Milwaukee, WI; retrofitted with Instascan; Advanced NMR Systems, Wilmington, MA). All patients received 0.2 mmol/kg of a Gd-based contrast agent (Gd-DTPA, Magnevist; Berlex Laboratories, Inc., Wayne, NJ) delivered by a prototype MR-compatible power injector (Medrad, Inc., Pittsburgh, PA) to an antecubital vein at a rate of 5 ml/second. Echoplanar imaging was performed using a TR of 1.5 seconds and a TE of 75 msec. Images were acquired in a imaging matrix with a cm field of view, resulting in mm pixels, with a slice thickness of 6 mm and an interslice gap of 1 mm. Ten slices were obtained simultaneously to cover the whole tumor region. Water Diffusion Measurements. The ADC was collected in 17 slices. For two different diffusion weightings (b = 47 seconds/mm 2, b = 1009 seconds/mm 2, where b represents a measure of diffusion weighting as defined by Stejskal and Tanner 25 ), three acquisitions were made in three directions to calculate the trace of the diffusion tensor. 24 For the follow-up images, patients were carefully repositioned in the magnet. The complete examination could typically be performed in 45 minutes. Image Analysis Tissue and Arterial Gd-DTPA Concentrations. Dynamic susceptibility contrast imaging 29 was used to determine tissue as well as arterial concentration time curves for the contrast agent. 31 The tissue and arterial concentration could thus be calculated in arbitrary units from the change in MR signal intensity using the spin-echo imaging sequence described previously. The time course of the arterial input function was obtained from four to eight pixels displaying large susceptibility effects caused by artery tissue contrast agent gradient around large vessels. This has been shown to allow accurate noninvasive determination of the arterial input function. 20 Measurement of CBV. The rcbv was measured by determining the area under the tissue concentration time curve according to the Central Volume Theorem. 22,23,26 The spin-echo imaging sequence used in our experiments has been shown to be specific for microvascular paramagnetic tracer levels, 31 allowing us to measure the microvascular blood volume. Measurement of CBF. The tissue tracer concentration time curve was deconvolved with the arterial input function on a pixel-by-pixel basis, and CBF was determined as the height of the deconvolved tissue response function We did not perform invasive measurements of blood tracer levels in these studies, and relative CBV and CBF values were consequently determined by normalizing to the white matter contralateral to the tumor in each patient. Blood-Brain Barrier Permeability. The BBB permeability surface area product to Gd-DTPA was measured by 301

3 L. Østergaard, et al. TABLE 2 Permeability of BTB to Gd-DTPA before and after dexamethasone treatment in six patients* Normalized BTB Permeability Case Change No. Pretreatment Posttreatment in % control * Values are normalized to the capillary input. In Cases 1, 3, and 4, permeability values were not significantly larger than the background level in normal tissue. Abbreviation: = no change. Values are expressed as 10 2 ml/100 ml/minute one standard deviation; however, a normalization factor is involved, so these values are not directly comparable to those obtained using other techniques. The patient was receiving constant, long-term, small-dose dexamethasone at the time of both measurements, which were obtained 3 weeks apart. fitting parameters to a model describing the combined T 1 and T 2 relaxation effects due to extra- and intravascular Gd-DTPA, respectively. 30 The resulting values for the permeability surface product were normalized to the integrated area of the capillary input function to allow comparisons among patients. We performed a control study to determine the reproducibility of the technique. Comparison of Imaging Data. In the resulting maps, the hemisphere contralateral to the tumor was carefully divided into gray and white matter based on CBV maps and anatomical images. In one patient (Case 5), the tumor crossed the midline in the posterior portion of the brain, and the frontal region of the brain was consequently used as a reference. In a second patient (Case 6) the tumor was localized to the cerebellum, and consequently only supratentorial structures were used as references. The tumor region was identified based on breakdown in BBB and consequent leakage of Gd-DTPA as seen in the BBB permeability maps and postcontrast MR imaging and focal increase in relative CBV. 1 Peritumoral edema was identified on T 2 -weighted images acquired as part of the MR examination. Results We measured CBV, CBF, and BBB permeability in six TABLE 3 Values of ADC in two patients who showed early (within 1 hour) decrease in BBB permeability ADC ( m 2 /second)* Control Region Tumor Edema Region Case No. Pretreatment Posttreatment Pretreatment Posttreatment * Values given are one standard deviation. Deep gray matter. patients and one control volunteer. Table 2 shows the resulting BTB permeabilities measured in these patients, including three patients and one control volunteer who had significant leakage of Gd-DTPA. The results of ADC measurements in the two patients who showed an early decrease (1 hour posttreatment) in BTB permeability are shown in Table 3. Notice the drop (53 10%) in BTB permeability over time periods as short as 1 hour. Figure 1 shows maps of permeabilities superimposed on anatomical images for the patient in Case 2. The diffusion measurements showed no changes in ADC measured within the edema or control regions. None of the patients showed changes in the degree of edema on T 2 -weighted images. Table 4 shows the measured CBV ratios in peritumoral gray matter (directly adjacent to edema) and white matter tissue (white matter within the edema) as well as gray matter in the contralateral hemisphere. There was a pronounced decrease in relative CBV of 15% on average in the peritumoral tissue. Regionally, these changes were much larger on occasion. Figure 2 shows an example (Case 4) in which relative CBV in gray matter in the periphery of the edematous zone decreased locally by 50%, to values within the range of normal gray matter. On average, the six patients showed a small (5%) but statistically significant decrease in contralateral gray matter CBV relative to normal white matter. Average CBF values in peritumoral and contralateral gray matter were all greater than normal white matter flow rates before dexamethasone treatment. Except in one patient (Case 5), the average peritumoral and contralateral flow rate ratios did not differ and showed no systematic changes after dexamethasone treatment. The patient in Case 5 showed low initial peritumoral gray matter blood flow relative to contralateral white matter (1.2 compared with 1.8 in normal gray matter). After dexamethasone treatment, the flow increased to near-normal values (1.6). Discussion Our measurements demonstrated profound changes in BTB permeability to small solutes over a time period of just 1 hour after administration of dexamethasone. Our results thus support the hypothesis that BTB closure may be important in the series of events leading to the observed clinical improvement. These findings agree with but somewhat exceed the results of the positron emission tomography studies reported by Jarden, et al., 12 who found a 25% reduction of the BTB permeability for the potassium analog rubidium-82 6 hours after the administration of steroids. We found a 15% decrease in peritumoral gray matter CBV relative to normal white matter in our study (Fig. 2). Contralateral normal gray matter CBV also showed this tendency but to a lesser extent. Leenders, et al., 14 observed a general decrease in gray as well as white matter CBV contralateral to the tumor 12 hours or more after dexamethasone treatment. Our measurements are in agreement with this finding, although the normalization of CBV to contralateral white matter did not allow us to estimate absolute changes in peritumoral and contralateral CBV. Our methodology did not allow us to address the biochemical basis of the observed changes in BTB perme- 302

4 Acute effects of dexamethasone in patients with brain tumors FIG. 1. Case 2. Maps of BBB permeability overlaid on anatomical images before (left column) and 1 hour after (right column) dexamethasone administration in two adjacent slices. Yellow indicates highest permeability and blue means low BTB permeability (color bar on right). Areas with no permeability to Gd-DTPA are transparent, showing only the anatomical image. PS = permeability surface area product. ability and peritumoral CBV. However, it is interesting to note that dexamethasone is a potent inhibitor of arachidonic acid formation following cell injury. Arachidonic acid profoundly destabilizes the BBB, mediating vasogenic edema, 6 but also triggers a cascade of reactions, releasing leukotrienes and prostaglandins, among these prostacyclin. Because these metabolites have biological half-lives of just a few minutes, arachidonic acid is a ratelimiting factor in the production of prostacyclin. 19 Therefore, dexamethasone may play a key role in restoring the vasomotor state as well as the barrier function of endothelium by controlling the release of arachidonic acid and thus prostacyclin. 14 Apart from the possible role of arachidonic acid in changing BTB permeability to small solutes, brain tumors have been shown to release proteinaceous vascular permeability factors into the interstitium. 5 These factors increase vascular permeability and induce angiogenesis, possibly accounting for the edema and increased neovascularization found in conjunction with brain tumors. 1,27 It has been hypothesized that dexamethasone may cause BTB closure by inhibiting the tumor expression of vascular permeability factors 5 or by directly modulating their action on endothelial cells. 7 Although our approach did not allow us to address fully the cellular mechanisms of BTB closure, the rapid changes in BTB permeability detected in this study seem to implicate a direct pharmacological action on the vascular endothelium of dexamethasone. As mentioned previously, symptoms associated with brain tumors have been ascribed to local increases in ICP caused by edema 9,12 or increased CBV. 14 In the patient in Case 5, CBV was increased locally by 100% relative to normal gray matter and subsequently normalized 1 hour after administration of dexamethasone. Given normal values of rcbv, this corresponds to a regional decrease in tissue volume from approximately 10% to approximately 5% (volume/volume). The decrease in CBV thus may be associated with substantial decompression of the peritumoral tissue. The ADC has been shown to be sensitive to shifts of water from intra- to extracellular space in stroke 303

5 L. Østergaard, et al. TABLE 4 Values for rcbv in contralateral gray matter and peritumoral gray and white matter relative to contralateral white matter before and after dexamethasone treatment* rcbv Ratios Contralat Peritumoral Peritumoral Gray Matter Gray Matter White Matter Case No. Pre Post Pre Post Pre Post ND ND ND ND avg change 5% ( 0.05) 15% ( 0.05) +6.5% (NS) (p value) * The edema zone was divided into gray and white matter except for the patient in Case 6, who had a lesion in the cerebellum for which segmentation could not be performed. Significant changes were seen only in gray matter, both contralaterally and in the edema zone. Abbreviations: avg = average; ND = not done; NS = not significant; post = after dexamethasone treatment; pre = before dexamethasone treatment. According to paired two-tailed Student s t-test. and experimental conditions with known cell swelling. 4 Given the signal-to-noise ratio of the measurements in our experiment and the known sensitivity of ADC changes to changes in extracellular water content, 28 we estimate that our ADC measurements were sensitive to changes in extracellular volume fraction of as little as 3% of the total volume. At least in the two patients examined, BTB closure thus did not seem to be associated with any immediate decrease in edema or source of peritumoral tissue decompression. Neither did edema resorption seem to precede BTB closure in these patients. These results indicate that acute changes in peritumoral CBV may be the prime source of the impressive clinical effect of dexamethasone. Although acute BTB closure was dramatic, it may cause edema resorption and tissue decompression over somewhat longer time periods. This is supported by measurements indicating that edema resorption takes place over periods of days rather than hours. 3 Our measurements of CBF showed no signs of pretreatment ischemia in peritumoral or contralateral gray matter. One patient showed an increase in peritumoral gray matter flow rate, in agreement with the observations of Reulen, et al. 21 It is interesting to note that the only patient who showed CBF changes after dexamethasone administration also showed a substantial drop in BTB permeability and peritumoral gray matter CBV. A diagnosis of primary central nervous system lymphoma, a tumor type known to show a strong response to dexamethasone treatment, had been made in this patient. 10 The data needed for CBF, CBV, and BBB permeability measurements can be acquired on most standard clinical MR imaging systems by performing dynamic echoplanar imaging during routine contrast injection. Even combined with ADC measurements, important pathophysiological information can therefore be gained while adding only a few minutes to conventional imaging time. Although we were only able to include one control volunteer in our study, we believe the techniques are sufficiently robust to allow studies of pathophysiological changes in individual patients. The techniques presented here would therefore be useful in studying, for example, the effects of corticosteroid analogs, or assessing the efficacy of drugs designed to modify BBB permeability for optimum delivery of chemotherapeutic agents. FIG. 2. Case 4. Maps of CBV before (A) and 1 hour after (B) dexamethasone treatment. Regional peritumoral gray matter CBV values were focally increased by more than 100% relative to normal gray matter before treatment. One hour after dexamethasone administration, CBV values had normalized. Arrowheads in row A indicate areas of abnormally high grey matter CBV. 304

6 Acute effects of dexamethasone in patients with brain tumors Conclusions This study demonstrates that dexamethasone reverses increased BTB permeability and reduces CBV in patients with brain tumors over very short time periods, whereas CBF and the degree of edema were seemingly unaltered. Measurement of these pathophysiological parameters can be performed as part of routine MR imaging, and may become useful in the management of individuals with brain tumors. Acknowledgments The authors thank Professor Jens Astrup, Dr. Claus Andersen, and Professor Albert Gjedde for helpful discussions. References 1. Aronen HJ, Gazit IE, Louis DN, et al: Cerebral blood volume maps of gliomas: comparison with tumor grade and histologic findings. Radiology 191:41 51, Axelrod L: Inhibition of prostacyclin production mediates permissive effect on glucocorticoids on vascular tone. Lancet 1: , Bell BA, Smith MA, Kean DM, et al: Brain water measured by magnetic resonance imaging. Correlation with direct estimation and changes after mannitol and dexamethasone. Lancet 1: 66 69, Benveniste H, Hedlund LW, Johnson GA: Mechanism of detection of acute cerebral ischemia in rats by diffusion-weighted magnetic resonance microscopy. Stroke 23: , Bruce JN, Criscuolo GR, Merrill MJ, et al: Vascular permeability induced by protein product of malignant brain tumors: inhibition by dexamethasone. J Neurosurg 67: , Chan PH, Fishman RA, Caronna J, et al: Induction of brain edema following intracerebral injection of arachidonic acid. Ann Neurol 13: , Criscuolo GR, Lelkes PI, Rotrosen D, et al: Cytosolic calcium changes in endothelial cells induced by a protein product of human gliomas containing vascular permeability factor activity. J Neurosurg 71: , Daumas-Duport C, Scheithauer B, O Fallon J, et al: Grading of astrocytomas. A simple and reproducible method. Cancer 62: , Galicich JH, French LA, Melby JC: Use of dexamethasone in the treatment of cerebral edema associated with brain tumors. Lancet 81:46 53, Hochberg FH, Miller DC: Primary central nervous system lymphoma. J Neurosurg 68: , Hossmann KA, Hürter T, Oschlies U: The effect of dexamethasone on serum protein extravasation and edema development in experimental brain tumors of cat. Acta Neuropathol 60: , Jarden JO, Dhawan V, Poltorak A, et al: Positron emission tomographic measurement of blood-to-brain and blood-to-tumor transport of 82 Rb: the effect of dexamethasone and whole-brain radiation therapy. Ann Neurol 18: , Klatzo I: Disturbances of the blood-brain barrier in cerebrovascular disorders. Acta Neuropathol Suppl 8:81 88, Leenders KL, Beaney RP, Brooks DJ, et al: Dexamethasone treatment of brain tumor patients: effects on regional cerebral blood flow, blood volume, and oxygen utilization. Neurology 35: , Østergaard L, Johannsen P, Høst-Poulsen P, et al: Cerebral blood flow measurements by magnetic resonance imaging bolus tracking: comparison with [ 15 O]H 2 O positron emission tomography in humans. J Cereb Blood Flow Metab 18: , Østergaard L, Smith DF, Vestergaard-Poulsen P, et al: Absolute cerebral blood flow and blood volume measured by magnetic resonance imaging bolus tracking: comparison with positron emission tomography values. J Cereb Blood Flow Metab 18: , Østergaard L, Sorensen AG, Kwong KK, et al: High resolution measurements of cerebral blood flow using intravascular tracer bolus passages. Part II. Experimental comparison and preliminary results. Magn Reson Med 36: , Østergaard L, Weisskoff RM, Chesler DA, et al: High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. Part I. Mathematical approach and statistical analysis. Magn Reson Med 36: , Pickard JD: Role of prostaglandins and arachidonic acid derivatives in the coupling of cerebral blood flow to cerebral metabolism. J Cereb Blood Flow Metab 4: , Porkka L, Neuder M, Hunter G, et al: Arterial-input function measurement with MRI. Proc Soc Magn Reson Med 1:120, 1991 (Abstract) 21. Reulen HJ, Hadjidimos A, Schürmann K: The effect of dexamethasone on water and electrolyte content and on CBF in perifocal brain edema in man, in Reulen HJ, Schurman K (eds): Steroids and Brain Edema. New York: Springer Verlag, 1972, pp Rosen BR, Belliveau JW, Buchbinder BR, et al: Contrast agent and cerebral hemodynamics. Magn Reson Med 19: , Rosen BR, Belliveau JW, Vevea JM, et al: Perfusion imaging with NMR contrast agents. Magn Reson Med 14: , Sorensen AG, Weisskoff RM, Reese TR, et al: Optimization of diffusion-weighted MR imaging for evaluation of acute stroke. Proc Soc Magn Res 3:1383, 1995 (Abstract) 25. Stejskal EO, Tanner JE: Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. J Chem Physiol 42: , Stewart GN: Researches on the circulation time in organs and on the influences which affect it. J Physiol (Lond) 15:1 41, Strugar J, Rothbart D, Harrington W, et al: Vascular permeability factor in brain metastases: correlation with vasogenic brain edema and tumor angiogenesis. J Neurosurg 81: , Szafer A, Zhong J, Gore JC: Theoretical model for water diffusion in tissues. Magn Reson Med 33: , Villringer A, Rosen BR, Belliveau JW, et al: Dynamic imaging with lanthanide chelates in normal brain: contrast due to magnetic susceptibility effects. Magn Reson Med 6: , Weisskoff RM, Boxerman JL, Sorensen AG, et al: Simultaneous blood volume and permeability mapping using a single Gd based contrast injection. Proc Soc Magn Reson 1:279, 1994 (Abstract) 31. Weisskoff RM, Zuo CS, Boxerman JL, et al: Microscopic susceptibility variation and transverse relaxation: theory and experiment. Magn Reson Med 31: , 1994 Manuscript received March 3, Accepted in final form September 10, This work was supported by The Søster and Verner Lippert Foundation, The Beckett Foundation, The Hjortenberg Foundation, The Einar Willumsen Foundation, The King Christian X Foundation, The Danish Research Academy (J. no. S940197), The Danish Medical Research Council ( , 1993) and by Public Health Service Grant Nos. RO1-CA40303 and RO1-HL39810, and Whitaker Foundation Grant Nos. RO1-CA66072 and PO1- CA Address reprint requests to: Leif Østergaard, M.D., Department of Neuroradiology, Århus Kommunehospital, Nørrebrogade 44, DK-8000 Århus, Denmark. leif@pet.auh.dk. 305