Magnetic Resonance Imaging of Regional Cerebral Blood Oxygenation Changes Under Acetazolamide in Carotid Occlusive Disease
Background Gradient-echo magnetic resonance imaging can demonstrate changes in cerebral blood oxygenation with high spatiotemporal resolution. We have previously shown that this technique allows monitoring of autoregulatory responses under vasodilatory stress in the healthy human brain. Here the approach has been extended to assess impairment of the autoregulatory reserve capacity in patients with carotid occlusive disease.
Summary of Report We studied four patients with unilateral occlusion of the internal carotid artery on a 2.0-T clinical high-field magnetic resonance system. Oxygenation-sensitive imaging was based on long-echo-time, gradient-echo sequences (repetition time, 62.5 milliseconds; echo time, 30 milliseconds) with low flip angles (10°) to emphasize changes in blood oxygenation rather than flow velocity. Dynamic recording monitored signal intensities before and after injecting 1 g of acetazolamide. In sections covering the hand area of the primary sensorimotor cortex, acetazolamide-induced magnetic resonance signal increases were attenuated in the vascular territories of occluded arteries. Lateralization of responses in the left and right hemispheric parts of the section corresponded to decreased hemodynamic reserve capacity as measured globally by transcranial Doppler ultrasonography.
Conclusions The present findings indicate that magnetic resonance imaging can demonstrate exhaustion of the autoregulatory reserve capacity when monitoring cerebral blood oxygenation changes during vasodilatory stress. We suggest that this method can help to evaluate regional cerebral hemodynamics in patients with carotid occlusive disease.
Rapid changes in reaction to functional activation or pharmacological manipulation are a characteristic feature of cerebral perfusion. Such responses are mediated by adapting vasomotor tone, which presents a considerable autoregulatory reserve capacity. Hemodynamic compromise as in the case of carotid occlusive disease stresses this reserve capacity and can be assessed by determining the ratio between cerebral blood flow and volume or by evaluating responses to additional vasodilatory stress as, for example, after application of the carbonic anhydrase inhibitor acetazolamide. In patients with cerebrovascular disease, the reserve capacity is currently monitored by transcranial Doppler ultrasonography (TCD), single-photon emission computed tomography (SPECT), positron emission tomography (PET), or dynamic or xenon computed tomography.1
Oxygen consumption is closely coupled to cerebral blood flow in the resting state but will “uncouple” at least temporarily during functional brain activation.2 Recently, the resulting rise in regional cerebral blood oxygenation (rCBO) has been exploited in magnetic resonance imaging (MRI) for mapping human brain function.3 Postcapillary “hyperoxgenation” occurring during brain activation can be visualized by long-echo-time, gradient-echo MRI sequences that are sensitive to the endogeneous contrast change provided by the concomitant decrease in paramagnetic deoxyhemoglobin.4 We have modified respective MRI techniques to more specifically reflect CBO changes by reducing artifactual influences from flow velocity changes or subject motion.5 Based on this improvement, we have shown in healthy subjects that CBO changes after acetazolamide injection can be monitored at high spatiotemporal resolution.6 The aim of this study was to demonstrate the changing pattern of CBO under vasodilatory stress in patients with unilateral occlusion of the internal carotid artery (ICA).
Subjects and Methods
Four patients (aged 31, 52, 55, and 70 years) were studied after they gave informed written consent. Except for transient symptoms in their history, they were neurologically asymptomatic with no evidence of previous brain infarction or other intracranial pathology on MRI. In all patients, conventional angiography demonstrated unilateral ICA occlusion without high-grade stenoses in the other brain-supplying arteries. All were normotensive without vasoactive medication and presented no medical problems that precluded the application of acetazolamide.
Dynamic CBO studies were carried out on a clinical high-field MRI system (2.0 T, Siemens Magnetom); the standard head coil was used. Anatomic images were acquired by means of three-dimensional T1-weighted MRI (radiofrequency-spoiled FLASH; repetition time [TR], 15 milliseconds; gradient echo time [TE], 6 milliseconds; radiofrequency pulse flip angle α, 20°; 32 contiguous partitions of 4-mm thickness). Oxygenation-sensitive images were spin density–weighted with T2* sensitivity (radiofrequency-spoiled FLASH; TR, 62.5 milliseconds; TE, 30 milliseconds; α, 10°; 96×256 matrix in conjunction with a rectangular 150×200-mm field of view corresponding to an in-plane resolution of 0.78×1.56 mm; section thickness 4 mm; 6 seconds per image; 65-Hz bandwidth per pixel; first-order motion compensation for the frequency-encoding and slice selection gradient). The use of a very low flip angle ensures that changes of blood flow in vessels crossing the imaging plane are virtually eliminated as a direct source of signal contrast.5
In this study global pharmacological and focal functional stimulation were compared in identical sections (results of the latter are to be reported elsewhere). Therefore, the acetazolamide-induced effects were measured in sections parallel to the bicommissural plane covering the hand area of primary sensorimotor cortex bilaterally. The pharmacological stimulation protocol consisted of dynamic MRI over 5 minutes with a bolus of 1.0 g IV acetazolamide after 96 seconds. The patients had ear plugs and were blindfolded.
Regional MRI signal intensities were compared before and after acetazolamide injection. Data evaluation was both quantitative and topographic and used regions of interest (ROI) and difference maps from a mean of 15 consecutive images before and 2 minutes after acetazolamide. ROI analyses followed dynamic responses in the original data set but also quantified relative signal intensity changes in the left and right hemispheric parts of the selected section. The ratio of relative signal changes in the hemispheres ipsilateral to the occluded and patent ICA, respectively, was defined as an index (O/P-Index) reflecting impairment of autoregulatory reserve capacity. The observer performing MR data analysis was blind for the side of occlusion.
For simultaneous insonation of the left and right middle cerebral artery (MCA), we used a refined TCD device containing two synchronized 2-MHz pulsed-wave Doppler instrumentations (Multidop X, DWL). Axial width of the sample volume was set at 9 mm. The MCA was insonated in a depth ranging between 50 and 60 mm. The Doppler probes were affixed with a head tape to the lateral temporal region such that insonation through the skull was possible. The envelope curve of the Fourier-transform spectra (128 points in each vertical line) was digitally stored with a sample frequency of 250 Hz for off-line analysis.
For the estimation of acetazolamide-induced cerebral blood flow velocity (CBFV) changes the mean velocity (CBFVmean) was calculated from the envelope curve during rest and 5 minutes after bolus injection of 1 g IV acetazolamide over an averaging period of 5 minutes each according to
with venv(ti) the velocity value of the envelope curve at time ti and N the number of points (5 minutes=5×60×250=75 000 points). The relative CBFVmean increase (%) due to acetazolamide was determined for each side, and response lateralization was expressed as an O/P-Index similar to the evaluation of the MRI data.
Resting MRI signal intensities displayed no left-right asymmetries in any of the patients. Acetazolamide administration induced an overall rise in signal intensity in normally perfused hemispheres in accordance with previous results in healthy subjects.6 Regional changes were greatest in macroscopic pial vessels and underlying gray matter, with considerably lower and mostly nonsignificant values in white matter. In contrast, signal intensity changes in the hemispheres ipsilateral to the occluded ICA ranged from widespread unresponsiveness (patients 1 and 2) to rises parallel to those in the normally perfused hemisphere (patient 4). The findings for patient 1 are illustrated in Fig 1⇓, showing ROI time courses in small veins within the central sulcus local to the left and right sensorimotor hand area. While ROI data demonstrate the dynamic profile of signal alterations, the topographic pattern of the acetazolamide-induced increase in CBO is visualized in difference maps of the “resting” and “stimulated” state. Resulting maps for patients 2 and 3 are shown in Fig 2⇓, together with the corresponding flow-sensitized anatomic images. The Table⇓ summarizes quantified evaluations of response lateralization. MRI-derived values in these patients indicate degrees of autoregulatory compromise similar to those deduced from the TCD recordings, in which relative MCA blood flow velocity changes were analyzed correspondingly. Sample size was too small to allow for reasonable statistical analysis.
The physiological response of vasomotor tone to vasodilatory stress from either functional activation or pharmacological manipulation is one of the prime features of intact cerebral perfusion. While corresponding changes in rCBO have recently been visualized in healthy subjects by means of MRI,6 this pilot study has demonstrated the potential of such an approach to assess regional compromise of the autoregulatory reserve capacity in ICA occlusion.
First, it is not surprising that we found no asymmetry of resting signal intensities. A lower signal intensity in the vascular territory of the occluded ICA would reflect a rise in fractional oxygen extraction with a consecutive increase in postcapillary deoxyhemoglobin. This may occur if the vasomotor reserve is exhausted but was not observed in the patients studied since even the most pronounced cases retained a weak MRI response to acetazolamide. Second, despite inconspicuous basal values, vasodilatory stress served to unmask a pathological compromise of the autoregulatory reserve capacity. Since oxygen metabolism remains stable under acetazolamide,7 changes in CBO are directly related to changes in blood flow. Therefore, reduced increases in CBO (or reduced decreases in local deoxyhemoglobin concentrations) are due to reduced stimulated blood flow during pharmacological manipulation. This finding implies a low vasomotor tone in the resting state with a consecutively attenuated dynamic response range. Teleologically, such abnormal vasomotor relaxation has the effect of maintaining regional blood flow despite low arterial perfusion pressure.
Noninvasive monitoring of parameters linked to perfusion and CBO in humans clearly extends beyond theoretical interest. In patients with ICA occlusion, an exhausted autoregulatory reserve capacity has been found to indicate increased risk for subsequent ischemic stroke.8 9 Extracranial-intracranial bypass surgery can improve the cerebrovascular reserve capacity in such patients.10 11 Hence, the issue of whether in strictly defined cases of “misery perfusion” clinical outcome may be improved by this procedure despite disappointing results in a more broadly defined patient group is not resolved.12 In addition, similar considerations may apply for asymptomatic patients presenting with high-grade ICA stenoses.
Within the ensemble of techniques currently applied for assessing cerebral hemodynamics, PET is considered the gold standard because it has the potential to separately quantify cerebral blood flow, cerebral blood volume, and fractional oxygen extraction.13 Although the correlation between findings obtained with different techniques remains moderate,13 14 TCD has been suggested as a useful clinical tool with practical advantages.14 15 However, since an uncontrolled change in collateral inflow during vasodilatory stress may limit the validity of TCD,14 16 future applications of the MRI approach introduced here should incorporate comparison with established tomographic measurements providing estimates of regional blood flow. Before widespread clinical application, the further development of an MRI-based vasodilatory stress test will also need to clarify whether the potential for increased spatial and temporal resolution of autoregulatory processes provides additional insights into cerebrovascular disease.
At present, despite the encouraging finding that MRI can detect pathological alterations in cerebral autoregulation, several caveats have to be borne in mind.
(1) While certain MRI techniques can be sensitized to changes in CBO, they do not allow absolute quantitation of oxygenation levels. Progress may arise from mapping differential changes in the effective spin-spin relaxation rate (1/T2*). When calibrated, these provide a direct measure of blood oxygenation in the microvasculature independent of interindividual signal differences and experimental settings, as has recently been demonstrated under respiratory challenges in rat brain in vivo.17 Nevertheless, for studies of unilateral carotid disease absolute rCBO values are not necessarily required since the contralateral hemisphere may serve as a “healthy” internal reference similar to its use in SPECT.14
(2) In this study the measurements were restricted to single sections of 4-mm thickness. However, volume coverage is not principally limited because data acquisition may readily be expanded to a multislice protocol (A.K., et al, unpublished data, 1994). Furthermore, the use of very high spatiotemporal resolution (implemented here mainly because of accompanying hand activation studies) is probably not necessary to evaluate acetazolamide responses and may be traded for simultaneous volume coverage.
(3) An additional issue is sensitivity. By comparison with flow-sensitized high-resolution anatomic images, rCBO changes are maximal in and around vessels (Fig 2⇑). While cortical responses are weaker but still achieve significance, this is not always the case in white matter.6 Although the technique in its present form might therefore fail to detect critical reduction of perfusion in the so-called internal border zone, there are technical means to enhance sensitivity to changes in the deoxyhemoglobin concentration. These include prolonged gradient echo times as well as larger voxel sizes as, for example, are required for improved volume coverage. Here, results derived from averaged data from hemispheric ROIs that included more white than gray matter were well in line with the TCD measurements.
(4) Finally, as with other techniques, it remains debatable whether vasodilatory stress tests or a refined analysis of the resting situation bears greater meaningfulness. In principle, MRI could be used to assess both cerebral blood volume and flow, thereby defining hemodynamic status. Such studies appear to be complementary approaches to determine hemodynamic compromise.
This study was supported by the Hermann & Lilly Schilling Foundation, Essen, Germany (Dr Steinmetz). We thank M. Requardt for assistance in data evaluation and A. Nachtmann for expert technical assistance. Figs 1⇑ and 2⇑ were generated by using the facilities of the Gesellschaft für wissenschaftliche Datenverarbeitung Göttingen.
- Received August 1, 1994.
- Revision received October 13, 1994.
- Accepted October 13, 1994.
- Copyright © 1995 by American Heart Association
Schmiedek P, Einhäupl K, Kirsch CM, eds. Stimulated Cerebral Blood Flow. Berlin, Germany: Springer-Verlag; 1992.
Fox PT, Raichle ME. Focal physiological uncoupling of cerebral blood flow and oxidative metabolism during somatosensory stimulation in human subjects. Proc Natl Acad Sci U S A. 1986;83:1140-1144.
Society of Magnetic Resonance in Medicine. Functional MRI of the brain. In: Syllabus of a workshop; June 17-19, 1993; Berkeley, Calif.
Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A. 1990;87:9868-9872.
Vorstrup S, Henriksen L, Paulson OB. Effect of acetazolamide on cerebral blood flow and cerebral metabolic rate for oxygen. J Clin Invest. 1984;74:1634-1639.
Kleiser B, Widder B. Course of carotid artery occlusions with impaired cerebrovascular reactivity. Stroke. 1992;23:171-174.
Karnik R, Valentin A, Ammerer HP, Donath P, Slany J. Evaluation of vasomotor reactivity by transcranial Doppler and acetazolamide test before and after extracranial-intracranial bypass in patients with internal carotid artery occlusion. Stroke. 1992;23:812-817.
Herold S, Brown MM, Frackowiak RSJ, Mansfield AO, Thomas DJ, Marshall J. Assessment of cerebral haemodynamic reserve: correlation between PET parameters and CO2 reactivity measured by the intravenous 133xenon injection technique. J Neurol Neurosurg Psychiatry. 1988;51:1045-1050.
Dahl A, Russell D, Nyberg-Hansen R, Rootwelt K, Bakke SJ. Cerebral vasoreactivity in unilateral carotid artery disease. Stroke. 1994;25:621-626.
Ringelstein EB, Sievers C, Ecker S, Schneider PA, Otis SM. Noninvasive assessment of CO2-induced cerebral vasomotor response in normal individuals and patients with internal carotid artery occlusion. Stroke. 1988;19:963-969.