British Journal of Anaesthesia 89 (2): 287±93 (2002) Regional response of cerebral blood volume to graded hypoxic hypoxia in rat brain C. Julien-Dolbec1, I. Tropres1, O. Montigon1, H. Reutenauer1, A. Ziegler1, M. Decorps1 and J.-F. Payen2* 1 INSERM 438 Unit and 2Department of Anaesthesia, The University of Grenoble School of Medicine, Grenoble, France *Corresponding author: INSERM U438, Pavillon B, HoÃpital Albert Michallon, BP 217, F-38043 Grenoble, France Background. The response of cerebral blood ¯ow to hypoxic hypoxia is usually effected by dilation of cerebral arterioles. However, the resulting changes in cerebral blood volume (CBV) have received little attention. We have determined, using susceptibility contrast magnetic resonance imaging (MRI), changes in regional CBV induced by graded hypoxic hypoxia. Methods. Six anaesthetized rats were subjected to incremental reduction in the fraction of inspired oxygen: 0.35, 0.25, 0.15, and 0.12. At each episode, CBV was determined in ®ve regions of each hemisphere after injection of a contrast agent: super®cial and deep neocortex, striatum, corpus callosum and cerebellum. A control group (n=6 rats) was studied with the same protocol without contrast agent, to determine blood oxygenation level dependent (BOLD) contribution to the MRI changes. Results. Each brain region exhibited a signi®cant graded increase in CBV during the two hypoxic episodes: 10±27% of control values at 70% SaO2, and 26±38% at 55% SaO2. There was no difference between regions in their response to hypoxia. The mean CBV of all regions increased from 3.6 (SD 0.6) to 4.1 (0.6) ml (100 g)±1 and to 4.7 (0.7) ml (100 g)±1 during the two hypoxic episodes, respectively (Scheffe F-test; P<0.01). Over this range, CBV was inversely proportional to SaO2 (r2=0.80). In the absence of the contrast agent, changes due to the BOLD effect were negligible. Conclusions. These ®ndings imply that hypoxic hypoxia signi®cantly raises CBV in different brain areas, in proportion to the severity of the insult. These results support the notion that the vasodilatory effect of hypoxia is deleterious in patients with reduced intracranial compliance. Br J Anaesth 2002; 89: 287±93 Keywords: blood, volume, cerebral; brain, magnetic resonance imaging; complications, hypoxia Accepted for publication: April 9, 2002 It is established that hypoxic hypoxia is accompanied by an increase in cerebral blood ¯ow (CBF) so that delivery of oxygen tends to be maintained.1 2 This increase is due to the dilation of cerebral arterioles.3 As a result, cerebral blood volume (CBV) rises. However, the change in CBV cannot be inferred from that in CBF because there is no clear relationship between the two.4 5 There have been few studies of changes in CBV induced by hypoxia: reports describe either only small effects on CBV6 7 or marked increases.8 9 By measuring regional CBV response to hypoxic hypoxia in anaesthetized, mechanically ventilated rats, our goal was to determine: (i) the relationship between the degree of the hypoxic insult and CBV changes; and (ii) possible differences in responsiveness to hypoxia between brain areas. We used susceptibility contrast magnetic resonance imaging (MRI) to measure CBV in dorsoparietal neocortex, striatum, corpus callosum and cerebellum during incremental reduction in the fraction of inspired oxygen (FIO2): 0.35, 0.25, 0.15 and 0.12. Materials and methods Two groups of fed Wistar female rats (200±220 g) were studied sequentially. Group 1 (n=6 rats) was used for the Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2002 Julien-Dolbec et al. determination of regional CBV using susceptibility contrast MRI. Susceptibility contrast MRI exploits the increase in the magnetic susceptibility difference (Dc) between the vascular and the extravascular compartments induced by the presence of a long-lived contrast agent con®ned in the vascular bed. This increase in Dc results in an increase DR2* of the decay rate (R2*=1/T2*) of the NMR signal from water protons, which is proportional to regional CBV, as previously shown:10 11 rCBV ˆ 3 DR2CBV 4 DB0 1† where g is the gyromagnetic ratio, and B0 is the magnetic ®eld in the absence of sample. It has been shown that deoxygenated haemoglobin acts as a natural intravascular contrast agent, which is the basis for the BOLD image contrast.12 This may interfere with the accuracy of CBV measurements during hypoxia.13 Therefore, a second group of rats (Group 2, n=6 rats) was studied during the same protocol without the contrast agent, to delineate the in¯uence of the BOLD effect on the images obtained with contrast agent. Animal preparation Preparation of animals was similar in the two groups and conformed to the guidelines of the French Government (decree No 87-848 of October 19, 1987, licenses 006683 and A38071). Anaesthesia was induced with 4% halothane and then maintained with an intraperitoneal injection of thiopental (40 mg kg±1). One percent lidocaine was injected subcutaneously for local anaesthesia at all surgical sites. After tracheostomy, rat lungs were mechanically ventilated with 65% nitrous oxide, 35% oxygen using a rodent ventilator (Model 683, Harvard Apparatus Inc., South Natick, MA, USA). Ventilation was adjusted to maintain PaCO2 at »35 mm Hg. FIO2 was continuously monitored (MiniOX I analyzer, Catalyst Research Corporation, Owings Mills, MD, USA). A 0.7 mm indwelling catheter was inserted into the left femoral artery to monitor mean arterial blood pressure (MABP) via a chart recorder (8000S, Gould Electronic, Ballainvilliers, France). Blood gases (PaO2 and PaCO2), arterial saturation of haemoglobin in oxygen (SaO2), arterial pH and haemoglobin content (Hb) were analysed in arterial blood samples of less than 0.1 ml (ABL 510, Radiometer, Copenhagen, Denmark). Another 0.7 mm indwelling catheter was inserted into the left femoral vein to continuously infuse normal saline containing epinephrine (1.5 mg ml±1) and sodium bicarbonate (0.025 mmol ml±1) at a rate of 2 ml h±1 throughout the study. Epinephrine was required to prevent the adverse effects of combined anaesthesia and hypoxic hypoxia on the cardiovascular system. Sodium bicarbonate was used to prevent arterial acidosis. Cannulation of the femoral vein was also required for the injection of the contrast agent (Group 1). Rectal temperature was maintained at 37.5 (0.5)°C by using a heating pad placed under the abdomen. Blood gases and arterial pH were corrected for rectal temperature. Experimental protocol Animals were subjected to a stepwise lowering of FIO2: 0.35, 0.25, 0.15 and 0.12. The basic cycle was started after more than 30 min of equilibration at FIO2 of 0.35 (control). The initial criteria for exclusion from the study were: MABP <100 mm Hg, arterial pH <7.30, PaO2 <100 mm Hg, arterial haemoglobin content <10 g dl±1. Subsequent episodes were then ®rst induced by lowering the inhaled oxygen for FIO2=0.25, then by replacing the oxygen by air (FIO2 of 0.15 and 0.12). During these four episodes, fractions of inspired nitrous oxide were 0.65, 0.75, 0.25 and 0.40, respectively. Each episode lasted 10 min: a 5 min equilibrium period followed by NMR acquisition and determination of MABP and arterial blood sampling. Preliminary studies showed that PaO2 reached a steady value within 5 min. When the cycle of measurements ended the rat was killed by administration of an overdose of thiopental (50 mg kg±1). MRI measurement MRI was performed with a SMIS console (SMIS Ltd, Guildford, UK) equipped with a 2.35 T, 40 cm diameter horizontal bore magnet (Bruker Spectrospin, Wissembourg, France) and a 20 cm diameter actively shielded gradient coil (Magnex Scienti®c Ltd., Yarnton, Oxford, UK). The rat was prone, its head secured via ear bars, and a 30 mm diameter surface coil was located directly above the brain. After radiofrequency coil matching and tuning, the magnetic ®eld homogeneity was adjusted to obtain a linewidth for water less than 0.5 parts per million (ppm) in the brain. Six adjacent horizontal slices (from 2 mm below bregma) were chosen from a T1 transverse scout image. A series of images for each slice at different echo times was acquired using a multi gradient-echo sequence with an interecho interval of 4.2 ms (repetition time TR=2 s; ®rst echo time TE=7.6 ms; number of slices=6; ®eld of view=35335 mm; slice thickness=1 mm; 64332 image matrix; number of averages=2). Acquisition of all images of the six slices took about 3 min. In Group 1 (measurement of CBV), superparamagnetic iron oxide particles (200 mmol iron kg±1 body mass of AMI 227, Sineremâ; Guerbet, Aulnay-sous-Bois, France) were injected intravenously 30 min after the start of the experiment (FIO2 of 0.35). Images were acquired before (n=24 echoes, pre-contrast image) and 3 min after injection (n=12 echoes, post-contrast image). Acquisition of postcontrast images was then repeated at the end of each subsequent PaO2 episode (FIO2 of 0.25, 0.15 and 0.12). 288 Cerebral blood volume response to hypoxia Data analysis Image processing and determination of regional CBV were performed using an Ultrasparc workstation (Sun Microsystems, Pasadena, CA, USA). In Group 1, for each PaO2 episode, T2* images were calculated by a least squares monoexponential ®t of the signal intensity vs the echo time on a pixel by pixel basis. Differences in relaxation rates in each pixel were then calculated according to the formula: DR2 ˆ 1 T2post ÿ 1 T2pre 2† with T2* pre and T2* post being the decay time constants before and after administration of the contrast agent (Group 1), respectively. The DR2* values were obtained from the T2* post-contrast values during the four successive episodes. Five regions of interest (ROI) were de®ned in the two hemispheres: super®cial and deep neocortex, corpus callosum, striatum and cerebellum. Selection of regions was made on slice 1 for super®cial neocortex, on slice 2 for deep neocortex, on slice 3 for corpus callosum, and on slice 5 for striatum and cerebellum, by comparing the images to an anatomical atlas.14 Large DR2* values (>200 s±1) assigned to large vessels were discarded. A correction for clearance of the contrast agent from the plasma (elimination half-time »4.5 h) was applied since the post-contrast experiments lasted »60 min. This correction is described elsewhere.15 In Group 2, DR2* BOLD was calculated using equation (2), where T2* pre and T2* post are the decay time constants during control and subsequent episodes, respectively. Assuming a similar BOLD effect within the selected brain regions for each episode, a mean value of T2* post was then determined from T2* values obtained in the ®ve brain regions. DR2* BOLD has been used to correct the DR2* values measured in Group 1 for the changes in deoxygenated haemoglobin concentration during hypoxia:16 DR2COR ˆ DR2 ÿ DR2BOLD 3† Regional CBV, expressed as the percentage of blood volume in each voxel, or ml (100 g)±1 tissue, was then determined according to equation (1). For an injection of AMI-227 of 200 mmol of iron kg±1 of body mass, Dc=0.571 ppm at 2.35T in large vessels.17 We assumed that the average haematocrit in the brain microcirculation was 0.83 of that in large vessels,4 5 resulting in a Dc value of 0.688. Finally, we assumed that the brain haematocrit remains constant during hypoxic hypoxia, as previously shown in most brain areas.6 Statistical analysis Data were expressed as mean (SD). Analysis for statistical signi®cance of changes during the successive episodes was performed using one-way analysis of variance (ANOVA) for repeated measurements (StatView SE program, SAS Institute Inc., Cary, NC, USA). Each value at a given episode was compared to that obtained at another episode using the Scheffe F-test post-hoc test. To look for a regional difference in the responsiveness to hypoxia, interaction between brain regions and episodes was assessed using a two-way ANOVA (brain region3episode) for repeated measurements. Differences between the two hemispheres was tested using a non-parametric Wilcoxon signed rank test. If no signi®cant difference was found between the two hemispheres, pooled data were subjected to the analysis. A stepwise regression analysis was used to estimate the respective in¯uence of SaO2 and of other factors (MABP, PaCO2) on the CBV changes for each episode. Statistical signi®cance was set at P<0.05. Results Physiological data are shown in Table 1. Hypoxic hypoxia caused a signi®cant decrease in MABP at FIO2 of 0.15 and 0.12. There was also signi®cant hypocapnia at FIO2 of 0.12. Typical T2* images of different coronal sections before and after the injection of the contrast agent and during the successive episodes of hypoxic hypoxia are shown in Figure 1. Administration of the contrast agent results in a decrease in T2* values, allowing determination of CBV in cerebral regions during the control period (FIO2 of 0.35). As FIO2 decreased, the brightness of T2* images was reduced, re¯ecting an increase in CBV. Table 2 shows CBV values in the selected brain regions. No signi®cant difference in CBV was found between the two hemispheres. Each brain region exhibited a signi®cant graded increase in CBV during the two hypoxic episodes, with the greatest degree of hypoxia (FIO2 of 0.12) yielding the largest regional CBV changes (Table 2). Percentage change in regional CBV was 10±27% of control values at FIO2 of 0.15, and 26±38% at FIO2 of 0.12. There was no signi®cant interaction between brain regions and episodes (F-test=1.3; P=0.22), meaning that no evidence of regional difference was found in the responsiveness to hypoxia. Mean CBV of the ®ve brain regions for each animal (mCBV) was also calculated. Hypoxic hypoxia signi®cantly raised mean CBV from 3.6 (SD 0.6) (FIO2 of 0.35) to 4.1 (0.6) ml (100 g)±1 and to 4.7 (0.7) ml (100 g)±1 at SaO2 70% (FIO2 of 0.15) and 55% (FIO2 of 0.12), respectively (P<0.01). Since SaO2 and PaO2 are two parameters independently measured by the blood gas analyser, we plotted experimental SaO2 and PaO2 values against mCBV (Fig. 2A and B). No relationship between mCBV changes and MABP or PaCO2was found. In contrast, there was a negative linear relationship between SaO2 and mCBV changes, with an r2 value of 0.80 (P<0.01): mCBV (% of control)=165.5±0.65 SaO2 (4) To quantify the error due to the BOLD effect in these changes, DR2* BOLD values were measured in Group 2 (no 289 Julien-Dolbec et al. contrast agent). In this group, DR2* BOLD values were +0.48 (2.04), +1.02 (2.37), and +2.36 (2.66) s±1, at FIO2 of 0.25, 0.15 and 0.12, respectively. These values corresponded respectively to 0.7%, 1.3% and 2.8% of those observed in Group 1. The increase in R2* BOLD was signi®cant at FIO2 of 0.12 only (P<0.05). Discussion The present study indicates that a direct relationship exists between CBV and hypoxic hypoxia. To our knowledge, this is the ®rst in vivo study measuring regional CBV in a model of graded hypoxia. CBV increased in proportion to the severity of the hypoxic insult, suggesting that vasodilatory capacity is not limited to this range of hypoxia. In addition, the changes were not signi®cantly different between the various brain regions investigated. Methodological critique Fig 1 Typical T2* images of horizontal slices (from slice A on the top to slice D at the bottom) before (FIO2 of 0.35) and after injection of the contrast agent at each episode (FIO2 of 0.35, 0.25, 0.15, and 0.12). Selection of regions of interest was made on slice A for super®cial neocortex, on slice B for deep neocortex, on slice C for corpus callosum, and on slice D for striatum and cerebellum. The use of exogenous contrast agent allowed us to monitor regional CBV changes during graded hypoxic hypoxia. Such techniques have been successfully used to investigate cerebrovascular changes in the brain during challenges such as hypercapnia, ischaemia or functional stimulation.15 18 19 Equation (1), used to determine absolute CBV, is based on a highly simpli®ed model of the brain vessel architecture.10 Despite this approximation, control values of regional CBV (2.8±4.3 ml (100 g)±1) obtained are in reasonable agreement with other studies in rats using different techniques5 9 or bolus tracking MRI.20 Deoxygenated haemoglobin, acting as an endogenous paramagnetic contrast agent, contributes to the difference in magnetic susceptibility between blood vessels and surrounding tissue.12 During hypoxic hypoxia, the increase in deoxygenated haemoglobin affects in a linear manner the changes in R2* with respect to the control state.13 21 We found a 2±3 s±1 increase in DR2* when SaO2 fell to 55%, in close agreement with those studies. Because of the large doses of contrast agent used in the present study, DR2* changes due to those in deoxyhaemoglobin concentration accounted for less than 5% of the CBV changes. Despite epinephrine, the greatest level of hypoxic hypoxia used in this study (FIO2 of 0.12) was associated with hypocapnia and a decrease in MABP, which might have interfered with CBV changes. For example, using a similar MRI procedure, we found that marked hypocapnia (PaCO2 25 mm Hg) resulted in a decrease in regional CBV of 12±17% in normoxic rats.22 However, the present change in PaCO2 was of smaller magnitude. In a recent study using Table 1 Physiological and biochemical data in Group 1 (with contrast agent) during successive episodes of reduced FIO2. Values are mean (SD). *P<0.05 vs FIO2 0.35 (Scheffe F-test) FIO2 MABP (mm Hg) PaO2 (mm Hg) SaO2 (%) PaCO2 (mm Hg) Arterial pH Haemoglobin (g dl±1) Temperature (°C) 0.35 0.25 0.15 0.12 131 (10) 142.3 (16.9) 100 37.1 (2.7) 7.32 (0.01) 13.0 (1.1) 37.5 (0.3) 134 (10) 97.3 (9.3)* 95.6 (1.8) 36.0 (4.8) 7.32 (0.01) 12.5 (0.9) 37.6 (0.3) 111 (15)* 55.0 (9.6)* 70.7 (7.3)* 31.7 (2.8) 7.36 (0.03) 12.0 (0.7) 37.4 (0.1) 92 (16)* 41.5 (3.5)* 53.2 (5.3)* 30.0 (2.7)* 7.36 (0.04) 11.9 (1.1) 37.5 (0.2) 290 Cerebral blood volume response to hypoxia MRI contrast imaging, an increase of only 10% in regional CBV was reported during progressive haemorrhagic hypotension in rats (MABP between 40 and 10 mm Hg).23 The lack of a signi®cant in¯uence of MABP and PaCO2 on the CBV (see stepwise regression analysis) indicates that both parameters probably have only minor effects on the present CBV changes. Another potential confounding factor in the CBV changes was the associated change in inspired concentration of nitrous oxide (FIN2O) during the successive episodes. Since nitrous oxide is a potent cerebrovasodilator,24 any change in its concentration might have interfered with our results. However, signi®cant increase in CBV was found during the two hypoxic episodes in which the fraction of inspired nitrous oxide was lowered (FIN2O of 0.25 and 0.40). Consequently, it is possible that the CBV changes would have been larger if the nitrous oxide fraction had been maintained constant. CBV response to hypoxic hypoxia The present study shows that CBV is signi®cantly increased by »15% at SaO2 70% (PaO2 55 mm Hg) and this rise reaches »30% at SaO2 55% (PaO2 40 mm Hg). These ®ndings are in accordance with other studies which reported a gradual change in CBF during graded hypoxic hypoxia.1 2 A gradual change in cerebral haemodynamics is seen as the oxygen content is lowered; this tends to maintain a constant oxygen supply to brain.25 We found that a stepwise reduction of the FIO2 raised regional CBV by 26±38% in all brain areas at SaO2 of 55%. It is recognized that the CBV values in normoxia differ among brain regions.4 In the present study, similar regional CBV responses to hypoxia were found in all regions, in agreement with studies measuring the CBF response to hypoxia.26 This suggests that the hypoxic stimulus may affect the various brain regions in a comparable manner regardless of their baseline blood ¯ow or blood volume. Recently, D'Arceuil and co-workers8 reported a 40±50% increase in cortical CBV at SaO2 of 40% in neonatal rabbits. In addition, under moderate hypoxia and hypercapnia, a 50% increase in CBV was found in the cortex of newborn piglets.27 In moderately hypoxic rats, a 30% increase in cortical CBV was reported by Shockley and LaManna.9 Although these results were obtained with various techniques (MRI, autoradiography, or optical methods), they are in agreement with our measurements of the changes in CBV found in both super®cial and deep neocortical regions at comparable levels of hypoxia. However, other studies have reported smaller CBV changes during hypoxia. Bereczki and co-workers6 found that moderate hypoxia (PaO2 40 mm Hg) increased microvascular volume by <20% in most areas of rat brain. However, in that study results were obtained in parenchy- Fig 2 Relationship between mCBV changes (% value in control) and SaO2 (A) and PaO2 (B) at differing FIO2 (0.35, 0.25, 0.15, and 0.12). *P<0.05 vs FIO2 0.35. Table 2 CBV values (ml (100 g)±1) in different brain regions at differing FIO2: control (FIO2 0.35). Values are mean (SD). *P<0.05 vs FIO2 0.35; ²P<0.05 vs FIO2 0.25; ³P<0.05 vs FIO2 0.15 (Scheffe F-test) Region of interest FIO2 0.35 Super®cial cortex Deep cortex Corpus callosum Striatum Cerebellum 4.28 3.26 2.84 3.13 4.28 0.25 (0.97) (1.02) (0.47) (0.72) (1.38) 4.41 3.36 2.91 3.19 4.71 0.15 (0.90) (1.00) (0.40) (0.71) (1.77) 291 4.68 3.52 3.30 3.72 5.42 0.12 ( 0.95) (0.94) (0.44)*² (0.86)*² (2.16)* 5.58 4.32 3.82 4.22 5.43 (1.46)*²³ (1.35)*²³ (0.69)*²³ (0.79)*²³ (1.82)* Julien-Dolbec et al. mal microvessels with diameter <50 mm, still having small baseline regional CBV values (0.4±2.2 ml (100 g)±1). CBV measured with the gradient-echo pulse sequence, used here, is less sensitive to vessel size, and re¯ects total CBV after the injection of a large dose of contrast agent.11 19 Considering that large vessels, with estimated diameter >200 mm, were excluded (see Materials and methods), the present changes in CBV should include pial arterioles, parenchymal arterioles, and venules. Therefore, if most of the CBV changes during hypoxia do indeed occur in these vessels, it is not surprising that methods detecting microvascular changes are associated with smaller changes. In contrast, it has been reported that hypoxic hypoxia (SaO2 70±75%) increased CBV by only 5±8% in human studies,7 28 instead of the 15% we found in rats. These differences may result from differences in the status (awake vs anaesthetized), species, or methods used for measuring CBV. There is no evidence that cerebrovascular responses to hypoxia should have been increased by the use of thiopental in rats. The well-known depressive effect of thiopental on brain metabolism would reduce baseline CBV, but the per cent response to hypoxia is not different from awake animals.29 Similarly, there is no reason to suspect a greater sensitivity to hypoxia in rats in comparison with humans. Therefore, differences between methods in their ability to detect CBV changes might be possible. For example, Fortune and co-workers7 used single-photon emission computed tomography (SPECT) with 99m-Tclabelled erythrocytes. In addition to its limited spatial resolution (4 cm in that study), the unavoidable problem with SPECT is extracranial contamination by labelled cells, possibly resulting in a large attenuation of CBV changes during respiratory challenges. The study by Hampson and colleagues28 used near-infrared spectroscopy (NIRS), whose reliability for determining CBV changes during hypoxia has not been established. It is thus possible that the changes in human CBV in response to hypoxia may have been underestimated. deleterious consequences of transient episodes of hypoxic hypoxia on cerebral haemodynamics in such patients. In summary, the present study shows that graded hypoxic hypoxia results in signi®cant increase in CBV in proportion to the severity of the insult. CBV changes were not signi®cantly different between the various brain regions investigated. We conclude that hypoxic hypoxia can signi®cantly contribute to increased intracranial pressure in subjects with reduced intracranial compliance. Acknowledgements This study was supported by a grant from the the DAAD/HSP II program, RhoÃnes-Alpes Region and the Ligue Nationale contre le Cancer. AMI-227 was kindly provided by Guerbet Laboratories. The authors thank Jonathan Coles and Raymond Koehler for helpful comments on the manuscript. References Clinical implications Identifying a relationship between the severity of a hypoxic insult and the range of CBV changes is of particular clinical relevance in patients with compromised intracranial compliance (i.e. head-injured patients). Any vasodilatory stimulus, for example hypoxic hypoxia, can aggravate intracranial hypertension and reduce cerebral perfusion pressure in such patients.30 In addition, the most severe degree of hypoxia led to the largest increase in CBV, showing no limit in cerebrovascular capacity for vasodilation within this range of hypoxia. If these experimental data are applicable in humans, our results indicate that a fall of SaO2 to 70% progressively increases the CBV by 15%. This is considerably more than the 7% blood volume change required to raise ICP to the threshold of 20 mm Hg in headinjured patients.31 Our results therefore underline the 292 1 Koehler RC, Traystman RJ, Zeger S, Rogers MC, Jones MD, Jr. Comparison of cerebrovascular response to hypoxic and carbon monoxide hypoxia in newborn and adult sheep. J Cereb Blood Flow Metab 1984; 4: 115±22 2 Ulatowski JA, Bucci E, Razynska A, Traystman RJ, Koehler RC. Cerebral blood ¯ow during hypoxic hypoxia with plasma-based hemoglobin at reduced hematocrit. Am J Physiol 1998; 274: H1933±42 3 Kontos HA, Wei EP, Raper AJ, Rosenblum WI, Navari RM, Patterson JL. Role of tissue hypoxia in local regulation of cerebral microcirculation. Am J Physiol 1978; 234: H582±91 4 Cremer JE, Seville MP. Regional brain blood ¯ow, blood volume, and haematocrit values in the adult rat. J Cereb Blood Flow Metab 1983; 3: 254±6 5 Todd MM, Weeks JB, Warner DS. The in¯uence of intravascular volume expansion on cerebral blood ¯ow and blood volume in normal rats. Anesthesiology 1993; 78: 945±53 6 Bereczki D, Wei L, Otsuka T, et al. Hypoxia increases velocity of blood ¯ow through parenchymal microvascular systems in rat brain. J Cereb Blood Flow Metab 1993; 13: 475±86 7 Fortune JB, Feustel PJ, deLuna C, Graca L, Hasselbarth J, Kupinski AM. Cerebral blood ¯ow and blood volume in response to O2 and CO2 changes in normal humans. J Trauma 1995; 39: 463±71; discussion 471±2 8 D'Arceuil HE, Crespigny AJ, Rother J, Moseley M, Rhine W. Serial magnetic resonance diffusion and hemodynamic imaging in a neonatal rabbit model of hypoxic-ischemic encephalopathy. NMR Biomed 1999; 12: 505±14 9 Shockley RP, LaManna JC. Determination of rat cerebral cortical blood volume changes by capillary mean transit time analysis during hypoxia, hypercapnia and hyperventilation. Brain Res 1988; 454: 170±8 10 Yablonskiy DA, Haacke EM. Theory of NMR signal behavior in magnetically inhomogeneous tissues: the static dephasing regime. Magn Reson Med 1994; 32: 749±63 11 Boxerman JL, Hamberg LM, Rosen BR, Weisskoff RM. MR contrast due to intravascular magnetic susceptibility perturbations. Magn Reson Med 1995; 34: 555±66 12 Ogawa S, Lee TM, Barrere B. The sensitivity of magnetic resonance image signals of a rat brain to changes in the cerebral venous blood oxygenation. Magn Reson Med 1993; 29: 205±10 13 Lin W, Paczynski RP, Celik A, Kuppusamy K, Hsu CY, Powers WJ. Experimental hypoxemic hypoxia: changes in R2* of brain Cerebral blood volume response to hypoxia 14 15 16 17 18 19 20 21 22 parenchyma accurately re¯ect the combined effects of changes in arterial and cerebral venous oxygen saturation. Magn Reson Med 1998; 39: 474±81 Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordinates. North Ryde: Academic Press, 1982 Payen JF, Vath A, Koenigsberg B, Bourlier V, Decorps M. Regional cerebral plasma volume response to carbon dioxide using magnetic resonance imaging. Anesthesiology 1998; 88: 984±92 Kennan RP, Scanley BE, Gore JC. Physiologic basis for BOLD MR signal changes due to hypoxia/hyperoxia: separation of blood volume and magnetic susceptibility effects. Magn Reson Med 1997; 37: 953±6 Tropres I, Grimault S, Vaeth A, et al. Vessel size imaging. Magn Reson Med 2001; 45: 397±408 Hamberg LM, Boccalini P, Stranjalis G, et al. Continuous assessment of relative cerebral blood volume in transient ischemia using steady state susceptibility-contrast MRI. Magn Reson Med 1996; 35: 168±73 Mandeville JB, Marota JJ, Kosofsky BE, et al. Dynamic functional imaging of relative cerebral blood volume during rat forepaw stimulation. Magn Reson Med 1998; 39: 615±24 Lin W, Celik A, Paczynski RP, Hsu CY, Powers WJ. Quantitative magnetic resonance imaging in experimental hypercapnia: improvement in the relation between changes in brain R2 and the oxygen saturation of venous blood after correction for changes in cerebral blood volume. J Cereb Blood Flow Metab 1999; 19: 853±62 Jezzard P, Heineman F, Taylor J, et al. Comparison of EPI gradient-echo contrast changes in cat brain caused by respiratory challenges with direct simultaneous evaluation of cerebral oxygenation via a cranial window. NMR Biomed 1994; 7: 35±44 Payen JF, Briot E, Tropres I, Julien-Dolbec C, Montigon O, Decorps M. Regional cerebral blood volume response to 293 23 24 25 26 27 28 29 30 31 hypocapnia using susceptibility contrast MRI. NMR Biomed 2000; 13: 384±91 Zaharchuk G, Mandeville JB, Bogdanov AA, Jr, Weissleder R, Rosen BR, Marota JJ. Cerebrovascular dynamics of autoregulation and hypoperfusion. An MRI study of CBF and changes in total and microvascular cerebral blood volume during hemorrhagic hypotension. Stroke 1999; 30: 2197±204; discussion 2204±5 Archer DP, Lebrecque P, Tyler JL, Meyer E, Trop D. Cerebral blood volume is increased in dogs during administration of nitrous oxide or iso¯urane. Anesthesiology 1987; 67: 642±8 Brown MM, Wade JP, Marshall J. Fundamental importance of arterial oxygen content in the regulation of cerebral blood ¯ow in man. Brain 1985; 108: 81±93 Pelligrino DA, Koenig HM, Albrecht RF. Nitric oxide synthesis and regional cerebral blood ¯ow responses to hypercapnia and hypoxia in the rat. J Cereb Blood Flow Metab 1993; 13: 80±7 Bauer R, Bergmann R, Walter B, Brust P, Zwiener U, Johannsen B. Regional distribution of cerebral blood volume and cerebral blood ¯ow in newborn piglets±effect of hypoxia/hypercapnia. Brain Res Dev Brain Res 1999; 112: 89±98 Hampson NB, Camporesi EM, Stolp BW, et al. Cerebral oxygen availability by NIR spectroscopy during transient hypoxia in humans. J Appl Physiol 1990; 69: 907±13 Donegan JH, Traystman RJ, Koehler RC, Jones MD, Jr, Rogers MC. Cerebrovascular hypoxic and autoregulatory responses during reduced brain metabolism. Am J Physiol 1985; 249: H421±9 Richards HK, Czosnyka M, Pickard JD. Assessment of critical closing pressure in the cerebral circulation as a measure of cerebrovascular tone. Acta Neurochir 1999; 141: 1221±7 Marmarou A. Increased intracranial pressure in head injury and in¯uence of blood volume. J Neurotrauma 1992; 9 [Suppl 1]: S327±32