A multimodality investigation of cerebral

Magnetic Resonance Imaging 25 (2007) 826 – 833

A multimodality investigation of cerebral hemodynamics and autoregulation in pharmacological MRI Alessandro Gozzia,4, Laura Ceolina, Adam Schwarza, Torsten Reesea, Simone Bertanib, Valerio Crestanb, Angelo Bifonea a

Department of Neuroimaging, Centre of Excellence for Drug Discovery, Psychiatry, GlaxoSmithKline Medicines Research Centre, 37135 Verona, Italy b Laboratory Animal Science, Centre of Excellence for Drug Discovery, Psychiatry, GlaxoSmithKline Medicines Research Centre, 37135 Verona, Italy Accepted 11 January 2007

Abstract Pharmacological MRI (phMRI) methods have been widely applied to assess the central hemodynamic response to pharmacological intervention as a surrogate for changes in the underlying neuronal activity. However, many psychoactive drugs can also affect cardiovascular parameters, including arterial blood pressure (BP). Abrupt changes in BP or the anesthetic agents used in preclinical phMRI may impair cerebral blood flow (CBF) autoregulation mechanisms, potentially introducing confounds in the phMRI response. Moreover, relative cerebral blood volume (rCBV), often measured in small-animal phMRI studies, may be sensitive to BP changes even in the presence of intact autoregulation. We applied laser Doppler flowmetry and MRI to measure changes in CBF and microvascular CBV induced by increasing doses of intravenous norepinephrine (NE) challenge in the halothane-anesthetized rat. NE is a potent vasopressor that does not cross the blood–brain barrier and mimics the rapid BP changes typically observed with acute drug challenges. We found that CBF autoregulation was maintained over a BP range of 60–120 mmHg. Under these conditions, no significant central rCBV responses were observed, suggesting that microvascular rCBV changes in response to abrupt changes in perfusion pressure are negligible within the autoregulatory range. Larger BP responses were accompanied by significant changes in both CBV and CBF that might confound the interpretation of phMRI results. D 2007 Elsevier Inc. All rights reserved. Keywords: phMRI; CBF; CBV; Autoregulation; Rat; Blood pressure

1. Introduction Pharmacological MRI (phMRI) methods can be applied to assess the effects of acute drug challenge on cerebral hemodynamics as a surrogate for changes in the underlying neuronal activity. This approach has been widely applied to study central effects of drugs on the central nervous system (CNS) in humans and animal models [1,2]. However, many of these drugs can also induce significant peripheral effects, including severe alterations of cardiovascular parameters. Under physiological conditions, mechanisms of autoregulation keep cerebral blood flow (CBF) relatively constant in the presence of changes in mean arterial blood pressure (MABP). However, general anesthetics, widely used in preclinical phMRI studies to avoid head motion and to better control animal physiology, may affect the central vasoadaptive response to peripheral MABP changes, thus 4 Corresponding author. Tel.: +39 0458219233; fax: +39 0458218073. E-mail address: [email protected] (A. Gozzi). 0730-725X/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.mri.2007.03.003

making it difficult to predict the influence of systemic vasopressive effects on cerebral hemodynamics. Moreover, large and rapid changes in MABP may cause a breakdown in the autoregulatory mechanisms that control brain microcirculation, thus introducing potential confounds in the interpretation of phMRI data. While blood-oxygen-level-dependent (BOLD) signals are most often measured in humans, relative cerebral blood volume (rCBV) has been widely used in phMRI studies in small laboratory animals due to the increased sensitivity afforded by rCBV measurements with intravascular contrast agents over BOLD [3]. However, dilation and constriction of cerebral blood vessels are thought to modulate vascular resistance in order to maintain CBF relatively constant in the presence of changes in perfusion pressure [4]. As a consequence, CBV might be sensitive to MABP changes even in the presence of intact autoregulation. Several attempts to correlate the magnitude of systemic MABP changes with the central hemodynamic res-

A. Gozzi et al. / Magnetic Resonance Imaging 25 (2007) 826 – 833

ponses in the rodent brain have been published. Zaharchuk et al. [5] did not observe significant CBF, CBV or BOLD changes as MABP was gradually decreased ( 1 mmHg/min) by continuous arterial blood withdrawal over the range of maximally effective autoregulation in the halothane-anesthetized rat. However, in a more recent study, Kalisch et al. [6] argued that slow and gradual decreases in MABP may not be representative of the abrupt changes typically observed in pharmacological MR experiments. Indeed, the same authors reported a significant correlation between BOLD signal time courses and MABP changes following rapid arterial blood withdrawal– reinfusion under three different anesthetic regimes in the rat (isoflurane, halothane and propofol). However, the use of the blood withdrawal–reinfusion method presents potential drawbacks such as the need to account for hemodilution, the risk of hemorrhagic shock-like complications and the lack of a stable and normotensive prestimulus MABP baseline. Other investigators have measured the BOLD signal changes produced by pharmacologically evoked MABP alterations. Tuor et al. [7] and, more recently, Wang et al. [8] reported significant correlations between BOLD signal and the MABP changes induced by norepinephrine (NE), a nonbrain-penetrant vasopressor, in rats anesthetized with achloralose. Luo et al. [9] observed significant fMRI responses under urethane anesthesia following acute challenge with cocaine but not with cocaine methiodide, a nonbrain-penetrant cocaine analogue, at doses that increased MABP up to 180 mmHg, thus suggesting that potentially confounding peripheral effects were negligible in that specific protocol. However, the BOLD response may result from changes in several metabolic and hemodynamic parameters whose contributions cannot be easily disentangled, and these conclusions cannot be extended to phMRI methods based on rCBV measurements. Here, we have applied laser Doppler flowmetry (LDF) and MRI to measure changes in CBF and microvascular CBV induced by increasing doses of an intravenous NE challenge in the halothane-anesthetized rat. The rCBV protocol employed has been used by us as well as by other groups to map the central hemodynamic response to a number of neuroactive compounds, including amphetamine [10–13], cocaine [14,15], apomorphine [16,17] and nicotine [18,19]. Following Tuor’s approach, we explored increasing doses of NE in order to correlate the magnitude of the cardiovascular response with the corresponding changes in CBV, to assess the potentially confounding effects of MABP changes on CBV-based phMRI data. The use of a pharmacological vasopressor has the advantage of circumventing the limitations of the blood withdrawal and reinfusion method while reproducing the abrupt MABP changes that are typically observed upon drug injection. By measuring LDF changes, we were able to assess the integrity of CBF autoregulatory mechanisms in our model. The independent measurement of CBF and CBV enabled

827

us to investigate the interplay of these two parameters in the range of effective vasoadaptive response and under autoregulation breakdown. 2. Methods 2.1. Animal preparation All experiments were carried out in accordance with Italian regulations governing animal welfare and protection. Protocols were also reviewed and consented to by a local animal care committee, in accordance with the guidelines of the Principles of Laboratory Animal Care (NIH publication 86-23, revised 1985). These studies were performed on male Sprague–Dawley rats (250–350 g; Charles River, Como, Italy). Animals had free access to standard rat chow and tap water and were housed in groups of five in solid bottom cages with sawdust litter. Room temperature (20–228C), relative humidity (45–65%) and dark–light cycles (12 h each, lights on at 0600 h) were automatically controlled. After arrival, rats were allowed to acclimatize for at least 5 days. 2.2. rCBV measurements Animal preparation/monitoring and MRI acquisition in each phMRI study were similar to those in previous studies [18]. Briefly, rats were anesthetized with 3% halothane in a 30%:70% O2:N2 gas mixture, tracheotomized and artificially ventilated with a mechanical respirator. The ventilation volume was adjusted to maintain physiological levels of p aO2 and p aCO2 according to arterial blood gases’ measurements performed during the study. The left femoral artery and vein were cannulated, and the animal was paralyzed with a 0.25-mg/kg iv bolus of d-tubocurarine followed by a continuous infusion of 0.25 mg/kg/h through the artery. All wounds were infiltrated with 1% lidocaine before incision. After surgery, the rat was secured into a customized stereotaxic holder (Bruker, Ettlingen, Germany), and the halothane level was set to 1% (1 MAC [20]). An MR-compatible thermocouple probe was used to measure rectal temperature. The body temperature of all subjects remained within physiological range (37F1.58C). MABP was monitored continually throughout the MRI experiment. At the end of the experiment, the animals were euthanized with an overdose of anesthetic followed by cervical dislocation. MRI data were acquired using a Bruker Avance 4.7-T system, a 72-mm birdcage resonator for RF transmit and a Bruker curved bRat BrainQ quadrature receive coil. The MR acquisition for each subject comprised RARE T 2weighted anatomical images (TReff, 5000 ms; TEeff, 76 ms; RARE factor, 32; FOV, 40 mm; 256256 matrix; 16 contiguous 1-mm slices) followed by a time series acquisition with the same spatial coverage and similar parameters (TReff, 2700 ms; TEeff, 110 ms), but with a lower in-plane spatial resolution (128128), giving a functional pixel volume of ~0.1 mm3. Following five

828


reference images, 2.67 ml/kg of the blood-pool contrast agent Endorem (Guerbet, France) was injected so that subsequent signal changes would reflect alterations in rCBV [21,22]. Four successive scans were averaged for a resulting time resolution of 80 s. Signal intensity changes in the time series were converted into fractional rCBV on a pixel-wise basis, using a constrained exponential model of the gradual elimination of contrast agent from the blood pool [23]. The rCBV time series data for each experiment were analyzed on both an image and volume of interest (VOI) basis within the framework of the general linear model. Individual subjects in each study were spatially normalized by a 9-df affine transformation, mapping their T 2weighted anatomical images to a stereotaxic rat brain MRI template set [16] and applying the resulting transformation matrix to the accompanying rCBV time series (FSL/ FLIRT v.5.2). RCBV time series were calculated covering a 6-min 40-s (10 time points) preinjection baseline and 12 min postinjection (18 time points), normalized to a common injection time point. This time window captured the relatively rapid rCBV signal changes observed following injection of NE (see Section 3). Image-based time series analysis was carried out using FEAT (fMRI Expert Analysis Tool) Version 5.43, part of FSL (FMRIB’s Software Library, www.fmrib.ox.ac.uk/fsl), with a 0.6-mm spatial smoothing, and using a model function identified by wavelet cluster analysis across all animals in the cohort, capturing the temporal profile of the postinjection signal change [24,25]. The design matrix also included the temporal derivative of this regressor and a linear ramp (both orthogonalized to the regressor of interest). The coefficients of the model function thus provided a map of rCBV response amplitude for each injection in each subject. Higher-level group comparisons were carried out using ordinary least squares simple mixed effects; Z (Gaussianized T/F) statistic images were thresholded using clusters determined by Z N 2.3 and a (corrected) cluster significance threshold of P =.05 [26]. VOI time courses were extracted from unsmoothed rCBV time series data using a 3D digital reconstruction of a rat brain atlas [27] coregistered with the MRI template [16], using custom in-house software written in IDL (Research Systems Inc., Boulder, CO). For each VOI time course, the average rCBV over a 4-min time window covering the peak response (from 1 min 20 s to 5 min 20 s postinjection; 7 time points) was used as a summary statistic of the relative change. The statistical significance of NE’s effects was assessed versus the postinjection response in vehicle-treated animals, as this reflects standard practice in phMRI. Group rCBV response from VOIs was compared between vehicle and NE-challenged groups by a oneway ANOVA followed by a Dunnett’s test for multiple comparisons. Threshold for statistical significance was considered as P = .05. Results are quoted and displayed as meanFS.E.M. unless otherwise indicated.

2.3. LDF measurements An Oxylite/LDF system (Oxford Optronix, Oxford, UK) was used to measure blood flow in the rat brain. Oxylite technology uses an optical fiber (200 Am diameter) for the LDF measurement. All probes are precalibrated by the manufacturer, and the calibration parameters are scanned into the system prior to each experiment. Animals were prepared using the same animal preparation protocol described above for phMRI experiments. At the end of surgery, the animals were placed in a stereotaxic frame for the insertion of the LDF probe. A hole was drilled through the parietal bone, and the probe was inserted in the right S1FL cortex perpendicular to the brain surface using the following coordinates from dura mater [28]: AP, +2.2 mm; ML, +2.8 mm; DV, 2.5 mm. S1FL is a region that ensures stable and reliable LDF measurements and in which high doses of NE induced significant rCBV changes (see Section 3). LDF and MABP were concurrently monitored throughout the experiment using a multichannel MP150 Biopac data-acquisition system (Biopac Systems Inc., Goweta, USA). The body temperature was monitored with a rectal probe and maintained at 37.5F18C using a heating pad. The ventilation volume was adjusted to maintain physiological levels of p aO2 and p aCO2 according to arterial blood gases’ measurements performed during the study. LDF values were recorded for 5 min before and 15 min after NE injection. At the end of the experiment, the animals were euthanized with an overdose of anesthetic followed by cervical dislocation. 2.4. Experiments and drugs RCBV and LDF were measured in two separate groups of animals. In the rCBV experiment, after 20 min of stabilization, animals were intravenously challenged with increasing doses of NE [l-( )-norepinephrine-(+)-bitartrate, CALBIOCHEM; 0.125 Ag/kg (n = 5), 0.5 Ag/kg (n =5), 2 Ag/kg (n =5) or 8 Ag/kg (n = 5)] or its vehicle (saline; n = 5). In the LDF study, animals were challenged with NE after 2 h of stabilization. In this case, group sizes were as follows: vehicle (n = 4), 0.125 Ag/kg (n =4), 0.5 Ag/kg (n =4), 2 Ag/kg (n = 4) or 8 Ag/kg (n = 4). NE doses refer to the salt form of the compound. NE or saline was administered over 80 s in a final volume of 1 ml, followed by 400 Al of saline to flush the intravenous line. 2.5. Statistical analysis Time courses for LDF were exported from the Biopac software at a time resolution of 10 s per data point. For each animal, the average LDF over a 50-s time window covering the peak response (1 min 20 s to 2 min postinjection) was used as a summary statistic of the relative change in each parameter. LDF time courses were normalized to the mean of a 5-min baseline period prior to the challenge and, thus, expressed as a fractional change. Statistical analysis of LDF was performed on the fractional changes relative to one. LDF values were compared between vehicle and NE-


829

Fig. 1. (A) Temporal profiles of MABP. NE or saline was injected at 0 min. (B) MABP group mean response to NE. (C) Time profile of LDF in the S1FL cortex. NE or saline was injected at 0 min. The horizontal line above the x-axis indicates the infusion time. (D) Group mean response of LDF changes in the motor cortex (statistical significance of one-way ANOVA with Dunnett’s correction for multiple comparisons: *P b.05, **P b.01, ***P b.0001 vs. vehicle). The horizontal line above the x-axis indicates the infusion time. The error bars of the temporal profiles have been omitted to improve the clarity of display.

challenged groups by a one-way ANOVA followed by a Dunnett’s test for multiple comparisons. The arterial blood pressure (BP) probe was calibrated with an external reference, and MABP values were expressed as millimeters of mercury. MABP time courses were exported at a resolution of 10 s per data point, and the average MABP of six time points covering the peak response (1 min 54 s to 2 min 6 s postinjection) was used as summary statistic. Preinjection baseline MABP values (averaged over 5 min preceding the injection) in the rCBV and LDF cohorts were similar (88.6F1.2 and 93.8F3.1 mmHg, respectively; P =.15). Since the profile and magnitude of the MABP response to NE were also equivalent in the two cohorts and no statistically significant difference was observed at any of the NE doses tested (two-tailed Student’s t test, P z.52), MABP data were pooled together to simplify the description of the results. MABP values were compared between vehicle and NE-challenged groups using a one-way ANOVA followed by a Dunnett’s test for multiple comparisons. Correlation coefficients for basal, pre- or post-p aCO2 and LDF or rCBV were calculated by linear regression analysis, with a level of significance of .05. To investigate the integrity of autoregulation, a scatter plot of all the individual time points (n = 2340) recorded during the simultaneous measurement of arterial BP and LDF was created. Arterial BP values were binned on the basis of 10-mmHg subdivisions, to improve the clarity of the plot. Linear regression analysis (least-squares fit) was performed for data within the autoregulation range (60–120 mmHg).

3. Results Intravenous administration of saline did not affect baseline MABP values. NE (0.125, 0.5, 2 and 8 Ag/ml) produced fast-onset dose-dependent rises in MABP (98.1F5.5, 115.7F4.3, 141.2F6.8 and 159.8F5.9 mmHg at peak, respectively; Fig. 1A). At the three highest doses, the effect reached statistical significance ( P b.0001 vs. saline, Fig. 1B). The changes were abrupt and short-lived, with return to preinjection baseline values typically within 5 min. No significant changes in mean LDF were observed at the two lower doses of NE, despite peak MABP changes up

Fig. 2. Relative LDF versus BP (meanF0.95 confidence intervals). The broken line shows the linear least-squares fit to the LDF data between 60 and 120 mmHg.

830


Fig. 3. (A) Group mean response of microvascular rCBV from representative VOIs at different NE doses (statistical significance of one-way ANOVA with Dunnett’s correction for multiple comparisons: *P b.05, **P b.01 vs. vehicle). (B) Temporal profiles of microvascular rCBV following NE challenge at different doses measured in the hippocampus and S1FL cortex. NE or saline were injected at 0 min.

to 115 mmHg (NE 0.5 Ag/kg). Transient increases in mean LDF (+30.6F10.1% and +37.4F12% at peak, respectively; P b.05 vs. vehicle baseline, Fig. 1C and D) were observed with larger doses of NE (2 and 8 Ag/kg; P b.05 and P b.01, respectively), corresponding to MABP responses in excess of 120 mmHg. The correlation of LDF and arterial BP is shown in Fig. 2. The small LDF reactivity (0.17% per mmHg, P =.00001) for BP values between 60 and 120 mmHg

suggests that autoregulation is preserved in this range. As BP exceeded 120 mmHg, a clear breakdown in CBF autoregulation was observed, with LDF becoming highly dependent on BP (0.48% per mmHg). A similar trend was also noticeable with BP values below 60 mmHg. The microvascular rCBV response to NE was consistent with that of LDF. Saline injection induced a slight, transient signal decrease. This effect is probably the result of temporary dilution of the blood-pool intravascular contrast

Fig. 4. Statistical parametric maps of the rCBV changes induced by increasing doses of NE versus vehicle, thresholded using clusters determined by Z N 2.3 and a (corrected) cluster significance threshold of P = .05. The figure also illustrates the region of insertion of the LDF probe within the S1FL cortex.

A. Gozzi et al. / Magnetic Resonance Imaging 25 (2007) 826 – 833 Table 1 Mean p aCO2 values before and after intravenous injections of NE Before first injection After last injection

CBV group (n = 5)

LDF group (n = 4)

36.6F3.9 39.9F2.6

35.8F1.0 33.5F1.6

p aCO2, partial pressure of arterial CO2. Values are presented as meanFS.E.M.

agent caused by the intravenous bolus. At the two lower doses of NE, no significant changes were observed. Shortlived microvascular rCBV increases started to appear in some of the VOIs at 2 Ag/kg (Fig. 3A), while at the higher dose of NE (8 Ag/kg, corresponding to a peak MABP of 165 mmHg), the rCBV increments were robust and statistically significant in all the VOIs examined (Fig. 3). Statistical parametric maps were generated to evaluate the spatial distribution of the microvascular rCBV changes across the brain. No significant effects were observed at the two lowest doses. At 2 Ag/kg, focal areas of significant activation were apparent in the cingulate and retrosplenial cortices alongside the sagittal sinus. The spatial distribution of the microvascular rCBV changes at the highest dose was much more widespread and also involved subcortical structures, including the hippocampus and the dorsomedial thalamus. The effect was prominent in cortical areas, probably reflecting the higher vascular density of these regions (Fig. 4). Basal p aCO2 levels were not significantly different in the LDF and the CBV groups ( P = .83). Mean p aCO2 values before and after the administration of NE were similar and not significantly different in both groups ( P = .51 and P =.61, respectively; Table 1). Linear regression analysis did not show any significant correlation between rCBV or LDF and pre- or post-NE p aCO2 levels ( P N.21, r b.67, for all VOIs). 4. Discussion and conclusion Whether and how anesthesia affects cerebrovascular reactivity have been a contentious matter in recent literature. Results have often been inconsistent, possibly because blood flow autoregulation is sensitive to the specific experimental conditions of the study. The assessment of the effects of anesthesia on brain circulation is of pivotal importance in the context of phMRI, as several psychoactive drugs are known to induce profound cardiovascular effects. Here, we have independently measured MABP, CBF and CBV changes induced by acute challenge with NE, a vasopressive drug that does not cross the blood– brain barrier and, therefore, is not expected to elicit changes in neuronal activity. The challenge was administered intravenously, thus inducing a rapid cardiovascular response that closely mimics the profile of MABP changes observed in phMRI experiments. No changes in either LDF or microvascular rCBV were observed over the 60- to 120-mmHg MABP range, while with larger MABP

831

changes (130 mmHg and above), both LDF and microvascular rCBV showed transient but significant increases. These results have important implications for the interpretation of phMRI studies. Firstly, our LDF data provide evidence that under our experimental conditions, blood flow autoregulation was maintained in the face of steeply increasing BP. Both the upper limit and the lower limit of the autoregulation range are very consistent with previous rodent studies performed under similar conditions [5], and the presence of a gentle slope in the autoregulation bplateauQ is a finding that has been reported by several contributors using different techniques [5,29]. It should be noted that the autoregulation range identified in the present study is wide enough to encompass the robust cardiovascular changes produced by strong psychostimulant agents such as nicotine, cocaine or amphetamine, at doses that elicit robust and reproducible central rCBV changes [10,14,18], thus supporting the use of this protocol for the functional investigation of vasoactive CNS drugs. A second important finding is that MABP changes within the blood flow autoregulatory range did not produce significant microvascular rCBV alterations. Vascular diameter is thought to vary in order to control vascular resistance, thus maintaining CBF despite changes in perfusion pressure. Therefore, changes in rCBV may occur even within the effective autoregulatory range. However, previous studies suggest that microvasculature contributes marginally to vascular resistance, which is thought to be dominated by arterioles [30]. For example, in the anesthetized cat, 39% of total vascular resistance resides in pial arteries with diameters larger than 400 Am and only 10% in smaller arteries [31]. Similar results have been published for other species [4]. Moreover, it was previously reported that during pharmacologically induced hypertension, small ( b150 Am) arteries and arterioles showed no changes in diameter until 170 to 190 mmHg, at which point they started to dilate with increases in MABP [29]. In our experiments, we used a spin-echo imaging sequence for the acquisition of the phMRI time series. Spin-echo techniques sensitize the image to rCBV changes in capillaries and greatly reduce the contribution of larger vessels to image contrast [32,33] with N 95% of the microvascular CBV signal theoretically arising from vessels with sizes between 4 and 30 Am [34]. Consistent with this, we did not observe significant changes in our rCBV data in the entire range of autoregulation. Our data are also in very good agreement with a previous study by Zaharchuk et al. [5] who used MR methods to measure total and microvascular CBV, CBF and BOLD signals during hemorrhagic hypotension in the halothane-anesthetized rat. In their study, no significant total or microvascular CBV changes were observed when the MABP was decreased from 140 to 50 mmHg, thus suggesting that CBV adjustments within autoregulatory range are modest and substantially smaller than previously reported [35]. To explain their findings, Zaharchuk et al. postulated that CBV changes do occur only in a very small subset of vessels (the

832


arterioles), thought to comprise less than 5% of overall cerebral blood volume. The results of the present study both substantiate and expand Zaharchuk et al.’s findings, by showing that microvascular CBV changes are negligible even in presence of abrupt MABP increases like those induced by NE. As the dose of NE was increased and MABP exceeded 130 mmHg, autoregulation did break down, and measurable changes in both LDF and microvascular CBV were observed. The onset of precipitous CBV increases following acute hypertension exceeding autoregulation is an established phenomenon reflecting a forced vasodilation consequent to an excessive perfusion pressure [4,29]. While the overall spatial distribution of the microvascular CBV changes in this regime was quite widespread across the brain, a certain degree of regionalization was observed particularly in cortical structures, consistent with the higher vascular density of these regions [6–8]. Alternatively, a nonuniform response to increased perfusion pressure could result from the systemic increases in MABP being transmitted in the cerebrovascular bed heterogeneously, depending on the local vascular resistance and responsiveness of the vessels to hypertension, as recently suggested by Wang et al. [8]. It should also be noted that very large MABP changes may trigger hypertensive encephalopathy accompanied by temporary disruption of the blood–brain barrier [36]. As NE is a potent, centrally active neurotransmitter, its extravasation into the CNS might induce direct neuronal activation. However, disruption of the blood–brain barrier is only observed for MABP increases in excess of 170 mmHg, outside the range investigated in the present study. Therefore, the observed LDF and CBV changes were most likely the result of a breakdown of autoregulation. In conclusion, we have shown that, in the widely used phMRI model of the halothane-anesthetized rat, CBF autoregulation is maintained over an MABP range of 60– 120 mmHg. Within this range, no significant microvascular rCBV responses were observed in the face of increasing MABP, thus suggesting that within the autoregulatory range, microvascular rCBV changes are negligible. However, larger MABP changes exceeding this range might confound the interpretation of the phMRI results, and care should be taken in selecting appropriate doses of drugs that affect cardiovascular parameters. The autoregulatory range and cerebrovascular response to peripheral MABP changes may vary with anesthetic agents and regimens and should, therefore, be assessed in the specific protocol used in the phMRI experiment.

References [1] Jenkins BG, Chen Y-CI, Mandeville JB. Pharmacological magnetic resonance imaging (phMRI). In: van Bruggen N, Roberts T, editors. Biomedical imaging in experimental neuroscience. New York7 CRC Press; 2003. p. 155 – 209.

[2] Honey G, Bullmore E. Human pharmacological MRI. Trends Pharmacol Sci 2004;25:366 – 74. [3] Chen Y-CI, Mandeville JB, Nguyen TV, Talele A, Cavagna F, Jenkins BG. Improved mapping of pharmacologically induced neuronal activation using the IRON technique with superparamagnetic blood pool agents. J Magn Reson Imaging 2001;14:517 – 24. [4] Chillon JM, Baumbach GL. Autoregulation: arterial and intracranial pressure. In: Edvinsson L, Krause DN, editors. Cerebral blood flow and metabolism. Philadelphia7 Lippincott Williams & Wilkins; 2002. p. 395 – 412. [5] Zaharchuk G, Mandeville JB, Bogdanov Jr AA, 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. [6] Kalisch R, Elbel GK, Gossl C, Czisch M, Auer DP. Blood pressure changes induced by arterial blood withdrawal influence bold signal in anesthesized rats at 7 tesla: implications for pharmacologic MRI. Neuroimage 2001;14:891 – 8. [7] Tuor UI, McKenzie E, Tomanek B. Functional magnetic resonance imaging of tonic pain and vasopressor effects in rats. Magn Reson Imaging 2002;20:707 – 12. [8] Wang R, Foniok T, Wamsteeker JI, Qiao M, Tomanek B, Vivanco RA, et al. Transient blood pressure changes affect the functional magnetic resonance imaging detection of cerebral activation. Neuroimage 2006;31:1 – 11. [9] Luo F, Wu G, Li Z, Li S-J. Characterisation of effects of mean arterial blood pressure induced by cocaine and cocaine methiodide on BOLD signals in the rat brain. Magn Reson Med 2003; 49:264 – 70. [10] Schwarz A, Gozzi A, Reese T, Bertani S, Crestan V, Hagan J, et al. Selective dopamine D(3) receptor antagonist SB-277011-A potentiates phMRI response to acute amphetamine challenge in the rat brain. Synapse 2004;54:1 – 10. [11] Chen YC, Galpern WR, Brownell AL, Matthews RT, Bogdanov M, Isacson O, et al. Detection of dopaminergic neurotransmitter activity using pharmacologic MRI: correlation with PET, microdialysis, and behavioral data. Magn Reson Med 1997;38:389 – 98. [12] Dixon AL, Prior M, Morris PM, Shah YB, Joseph MH, Young AM. Dopamine antagonist modulation of amphetamine response as detected using pharmacological MRI. Neuropharmacology 2005; 48:236 – 45. [13] Chen YI, Choi JK, Jenkins BG. Mapping interactions between dopamine and adenosine A2a receptors using pharmacologic MRI. Synapse 2005;55:80 – 8. [14] Schwarz AJ, Zocchi A, Reese T, Gozzi A, Garzotti M, Varnier G, et al. Concurrent pharmacological MRI and in situ microdialysis of cocaine reveal a complex relationship between the central hemodynamic response and local dopamine concentration. Neuroimage 2004; 23:296 – 304. [15] Marota JJA, Mandeville JB, Weisskoff R, Moskowitz MA, Rosen B, Kosofsky BE. Cocaine activation discriminates dopaminergic projections by temporal response: an fMRI study in rat. Neuroimage 2000; 11:13 – 23. [16] Schwarz AJ, Danckaert A, Reese T, Gozzi A, Paxinos G, Watson C, et al. A stereotaxic MRI template set for the rat brain with tissue class distribution maps and co-registered anatomical atlas: application to pharmacological MRI. Neuroimage 2006;32:538–50. [17] Nguyen TV, Brownell A-L, Chen Y-CI, Livni E, Coyle JT, Rosen B, et al. Detection of the effects of dopamine receptor supersensitivity using pharmacological MRI and correlations with PET. Synapse 2000;36:57 – 65. [18] Gozzi A, Schwarz AJ, Reese T, Bertani S, Crestan V, Bifone A. Region-specific effects of nicotine on brain activity: a pharmacological MRI study in the drug-naRve rat. Neuropsychopharmacology 2006;31:1690 – 703.

A. Gozzi et al. / Magnetic Resonance Imaging 25 (2007) 826 – 833 [19] Choi JK, Mandeville JB, Chen YI, Kim YR, Jenkins BG. High resolution spatial mapping of nicotine action using pharmacologic magnetic resonance imaging. Synapse 2006;60:152 – 7. [20] Steffey MA, Brosnan RJ, Steffey EP. Assessment of halothane and sevoflurane anesthesia in spontaneously breathing rats. Am J Vet Res 2003;64:470 – 4. [21] Mandeville JB, Marota JJA, Kosofsky BE, Keltner JR, Weissleder R, Rosen B, et al. Dynamic functional imaging of relative cerebral blood volume during rat forepaw stimulation. Magn Reson Med 1998; 39:615 – 24. [22] Reese T, Bjelke B, Porszasz R, Baumann D, Bochelen D, Sauter A, et al. Regional brain activation by bicuculline visualized by functional magnetic resonance imaging. Time-resolved assessment of bicuculline-induced changes in local cerebral blood volume using an intravascular contrast agent. NMR Biomed 2000;13:43 – 9. [23] Schwarz AJ, Reese T, Gozzi A, Bifone A. Functional MRI using intravascular contrast agents: detrending of the relative cerebrovascular (rCBV) time course. Magn Reson Imaging 2003;21:1191 – 200. [24] Whitcher B, Schwarz AJ, Barjat H, Smart SC, Grundy RI, James MF. Wavelet-based cluster analysis: data-driven grouping of voxel time courses with application to perfusion-weighted and pharmacological MRI of the rat brain. Neuroimage 2005;24:281 – 95. [25] Schwarz AJ, Whitcher B, Gozzi A, Reese T, Bifone A. Study-level wavelet cluster analysis and data-driven signal models in pharmacological MRI. J Neurosci Methods 2006;159:346 – 60. [26] Worsley KJ, Evans AC, Marrett S, Neelin P. A three-dimensional statistical analysis for CBF activation studies in human brain. J Cereb Blood Flow Metab 1992;12:900 – 18.

833

[27] Paxinos G, Watson C. The rat brain in stereotaxic coordinates. 4th ed. San Diego7 Academic Press; 1998. [28] Paxinos G, Watson C. The rat brain in stereotaxic coordinates. 5th ed. Amsterdam: Elsevier; 2005. [29] Kontos HA, Wei EP, Navari RM, Levasseur JE, Rosenblum WI, Patterson Jr JL. Responses of cerebral arteries and arterioles to acute hypotension and hypertension. Am J Physiol Heart Circ Physiol 1978;234:H371 – 83. [30] Mellander S, Johansson B. Control of resistance, exchange, and capacitance functions in the peripheral circulation. Pharmacol Rev 1968;20:117 – 96. [31] Shapiro HM, Stromberg DD, Lee DR, Wiederhielm CA. Dynamic pressures in the pial arterial microcirculation. Am J Physiol 1971;221:279 – 83. [32] Gao JH, Miller I, Lai S, Xiong J, Fox PT. Quantitative assessment of blood inflow effects in functional MRI signals. Magn Reson Med 1996;36:314 – 9. [33] Kennan RP, Zhong J, Gore JC. Intravascular susceptibility contrast mechanisms in tissues. Magn Reson Med 1994;31: 9 – 21. [34] Boxerman JL, Hamberg LM, Rosen BR, Weisskoff RM. MR contrast due to intravascular magnetic susceptibility perturbations. Magn Reson Med 1995;34:555 – 66. [35] Powers WJ. Cerebral hemodynamics in ischemic cerebrovascular disease. Ann Neurol 1991;29:231 – 40. [36] Nukhet TA, Ziya ZY. Protection of blood–brain barrier breakdown by nifedipine in adrenaline-induced acute hypertension. Int J Neurosci 2004;114:517 – 28.