| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Technical Developments |
1 From the Departments of Radiology (F.C., R.P., K.C., W.C., G.M., Y.N.) and Molecular and Vascular Biology (Y.S., N.N.), University Hospitals, K. U. Leuven, Herestraat 49, B-3000 Leuven, Belgium. Received October 13, 2003; revision requested January 7, 2004; revision received February 6; accepted March 2. Address correspondence to Y.N. (e-mail: yicheng.ni@med.kuleuven.ac.be).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and Methods Results DiscussionREFERENCES |
|---|
© RSNA, 2004
Index terms: Animals • Magnetic resonance (MR), diffusion study • Magnetic resonance (MR), experimental studies • Magnetic resonance (MR), perfusion study
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods Results DiscussionREFERENCES |
|---|
Since the middle cerebral artery (MCA) territory is the most frequently affected area for human focal cerebral infarcts due to poor collateral blood supply, stroke occurrence in animal models most often involves MCA occlusion of various extent (1). Among several methods described in rats, photochemically induced thrombosis of MCA is a basic approach of inducing various types of MCA occlusion that closely resembles various types of stroke in humans (6,15–19). Actually, the extent and severity of focal cerebral ischemic lesions can be different, depending on whether a distal (6,17–19) or proximal (15,16) MCA occlusion is induced.
The photochemically induced thrombosis model of proximal MCA occlusion is believed to be advantageous over other models in terms of its possibilities in both permanent ischemia and postischemic reperfusion and is therefore ideal for studies of new antistroke medications (16,20,21), such as thrombolytic microplasmin (22). With conventional invasive autopsy techniques, however, a large number of animals is required, since only limited information can be obtained from one sacrificed animal at a single time point. Furthermore, these techniques are not ideal for precise investigation of pathologic and therapeutic mechanisms because of uncertain initial infarct size in individual animals. Thus, the purpose of our study was to report the feasibility of using clinical MR imaging devices for depiction of stroke in a rat model.
|
Materials and Methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods Results DiscussionREFERENCES |
|---|
The left MCA was occluded with a transorbital approach by using a surgical microscope. Briefly, a vertical incision was made between the left orbit and the external auditory canal. To expose the proximal section of the left MCA, the scalp and temporalis muscle were reflected, and a subtemporal craniotomy was performed by using a dental drill to open an elliptic bone window (major axis, 3 mm). The MCA was occluded by a photochemically induced thrombotic reaction (19,20).
The window was irradiated to thrombotically occlude the main trunk of the left MCA and small branches at the olfactory tract. Photoillumination with green light (wave length, 540 nm; bandwidth, 80 nm) was achieved by using a xenon lamp (model L-4887; Hamamatsu Photonics, Hamamatsu City, Japan) with heat-absorbing and green filters. The irradiation at an intensity of 0.68 W/cm2 was directed with a 3-mm optic fiber, the head of which was placed at a distance 2 mm above the MCA. Photoillumination was performed for 10 minutes after intravenous injection of the photosensitizer rose bengal (20 mg per kilogram of body weight) through a tail vein. After MCA occlusion, the air-filled cavities caused by surgical dissection were filled with a gelatin sponge to avoid susceptibility artifacts during MR imaging. The temporalis muscle and skin were closed in layers, and anesthesia was discontinued.
The presence and degree of stroke-induced brain damage were assessed by one of two authors (F.C., Y.S.) by using the following criteria, described by Bederson et al (24), which are used to systematically score neurologic and behavioral aspects of the rat: 0, rats extend straight both forelimbs, no observable deficit; 1, rats attach the right forelimb to breast and extend straight the left forelimb; 2, rats decrease resistance to lateral push in addition to behavior as in score 1 without circling; and 3, rats twist the upper half of their bodies in addition to behavior in score 2.
MR Imaging Techniques
Rats were placed supinely into a plastic holder and connected to the same inhalation anesthesia system used in surgery. The penile vein of rats was cannulated for contrast agent injection.
MR imaging was performed with a 1.5-T imager (Sonata; Siemens, Erlangen, Germany) by using a commercially available 4-channel phased-array wrist coil (MRI Devices Corporation, Waukesha, Wis). Sagittal, coronal, and transverse pilot images were obtained for positioning of the subsequent MR imaging acquisitions.
For each imaging sequence, 12 coronal images were acquired with a section thickness of 2 mm gapped at 0.2 mm. T1-weighted spin-echo MR images were acquired with a repetition time msec/echo time msec of 500/15, field of view of 55 x 34.4 mm, image acquisition matrix of 256 x 120, and three signals acquired, which led to a total imaging time of 3 minutes 21 seconds. T2-weighted turbo spin-echo MR images were obtained with 5860/100, field of view of 55 x 31 mm, acquisition matrix of 256 x 146, and four signals acquired, which led to a total measurement time of 3 minutes 9 seconds.
Diffusion-weighted MR images were obtained by using a two-dimensional spin-echo echo-planar sequence with a field of view of 140 x 70 mm and an acquisition matrix size of 192 x 96, which led to an in-plane resolution of 0.7 x 0.7 mm. To reduce susceptibility artifacts and imaging time, a parallel imaging technique—namely, a generalized autocalibrating partially parallel acquisition (GRAPPA; Siemens) with an acceleration factor of two was applied. The parameters of 3000/80 and 12 signals acquired resulted in a total acquisition time of 4 minutes 20 seconds, including the repeated measurements for three different b values (0, 500, and 1000 sec/mm2) to allow calculation of quantitative maps of the apparent diffusion coefficient (ADC). For diffusion-weighted MR imaging, three directions (x, y, and z) were measured and averaged for the calculation of the isotropic ADC value.
Perfusion-weighted MR images were acquired by using a T2*-weighted (2000/40) echo-planar imaging sequence in combination with the GRAPPA technique (Siemens) to increase spatial homogeneity in the images. A dynamic image series of 60 measurements resulted in a total imaging time of 2 minutes 36 seconds, with a field of view of 156 x 43.5 mm and an image acquisition matrix of 128 x 64, which led to an in-plane resolution of 0.7 x 0.7 mm. During the dynamic series, a triple-dose (0.3 mmol/kg) bolus injection of contrast material (Omniscan; Amersham, Oslo, Norway) was started after the 30th acquired volume to ensure a sufficient precontrast baseline. The bolus injection was administered manually in less than 1 second without saline flush. During postprocessing, a 25-mm2 region of interest was placed contralaterally (F.C., Y.N.), covering the origin of the right MCA to allow measurement of the arterial input function, from which we selected the pixels representative of the right MCA branch. With this technique, brain perfusion maps showing relative cerebral blood flow and volume were derived automatically with the built-in software (Siemens). Local dynamic mean signal intensity curves were obtained by using a region of interest (F.C.) ipsi- and contralaterally.
Delayed (10–20 minutes) postcontrast T1-weighted MR imaging was conducted with the same acquisition parameters described earlier to detect any enhancement of the cerebral ischemic lesions due to breakdown of the blood-brain barrier.
Study Design
According to the duration after MCA occlusion, the animals were randomly assigned to the following two groups: All 24 rats were subjected to the MR imaging protocol described earlier at the superacute phase (approximately 1 hour) after MCA occlusion. Twelve of the rats were then sacrificed after completion of MR imaging and were considered to constitute the superacute group (n = 12). The other 12 rats were followed up with MR imaging at both the acute phase (12 hours) and the subacute phase (24 hours). They were sacrificed afterward and were considered to constitute the subacute group (n = 12). The calculated statistical power of the present study was .95.
CT, Radiographic Microangiography, and Histochemical Staining
For rats in the superacute group (n = 12), we performed microangiography and three-dimensional CT to validate reliability of perfusion-weighted MR imaging for mapping of abnormal perfusion of the ischemic lesion and to assess the vasculature and access of the contrast agents to the ischemic lesions. Technically, after euthanasia of the animals with an intravenous overdose of pentobarbital, the left ventricle of the heart was infused with a 1:5 Evans blue barium suspension until barium became visible subcutaneously at the extremities. The excised brain was first scanned with multi–detector row CT (Sensery 16; Siemens) with a section thickness of 0.75 mm and reconstruction thickness of 0.2 mm and displayed with a volume-rendered technique. After that, the brain was radiographed with a mammographic soft x-ray machine, serially sliced in the same plane as on MR images, and imaged again with a microfocus radiographic technique (25–26 kV, 2.8–5.0 mAs).
For the rats in the subacute group (n = 12), a half-hour before sacrifice, 1 mL of 1% Evans blue solution was injected intravenously for assessment of postmortem blood-brain barrier integrity. The euthanized rats were cryoprotected at –20°C. Their skulls were sliced with a bone saw (Dedoc beta 200; Medoc, Pol. Cantabria Logroño, Spain) into 12 2-mm-thick slices oriented similar to MR imaging views. The slices were incubated at 37°C in 2% triphenyl tetrazolium chloride (TTC) solution for 15 minutes, and normal brain tissue was stained brick red while cerebral infarct remained pale. The brain sections were fixed with 10% formalin and processed with hematoxylin-eosin staining for microscopy.
Image Analysis
For the quantification of lesion size, the area of a lesion was determined in consensus by three authors (F.C., Y.N., N.N.) by means of an operator-defined region of interest on each of the lesion-containing sections, and the areas were then summed for each animal. The entire brain areas were also measured by using the same method. To facilitate the comparison of lesion sizes between animals and imaging methods, the relative lesion size (lesion size divided by brain size) was calculated.
Digitized photographs of pathologic sections, microangiographic findings, and specimens with TTC staining were transferred to a personal computer for planimetric analysis by using a commercial program (Adobe Photoshop 7; Adobe, San Jose, Calif). The perfusion defect and normal brain, as well as negatively and positively TTC-stained areas, were contoured manually (F.C., Y.N.) on the computer screen. By using the same formula as in MR imaging analysis, relative sizes (expressed in percentages) on microangiographic or pathologic and TTC-stained sections were also obtained as postmortem standards of reference for lesion size in super- and subacute phases.
Statistical Analysis
Numeric data were reported as means ± standard deviations. Statistical analysis was performed with the SPSS for Windows software package (version 11.5; SPSS, Chicago, Ill). A general linear-model univariate test was applied for global comparisons between groups and techniques, followed by Bonferroni tests for multiple comparisons of lesion size. A significant difference was indicated by a P value of less than .05.
|
Results |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods Results DiscussionREFERENCES |
|---|
The MR measurement parameters were optimized in the present rat experiment, given the limitations imposed by the clinical imager.
MR Imaging Findings as Verified with Postmortem Techniques
Despite certain magnetic susceptibility artifacts on diffusion- and perfusion-weighted MR images, which resulted in some extent of image distortion, the focal cerebral ischemic lesions were distinguishable in all rats, and their temporal evolution could be monitored and verified with the different imaging protocols (Figs 1–3).
|
|
|
|
Stroke in acute and subacute phases.—In acute and subacute groups 12 and 24 hours after occlusion of MCA, with time, T2- and diffusion-weighted MR imaging exhibited expanding hyperintense regions in the striatum and left temporal cortex (Fig 1, Table). The signal intensity of the lesion on T2- and diffusion-weighted MR images, the decrease of ADC values, and the extent of ischemic lesion reached maximum levels at 24 hours (Fig 1, Table). The lesion on perfusion-, T2-, and diffusion-weighted MR images and ADC maps at 24 hours was matched well with that seen at TTC staining (P > .05), indicating the real infarct size (Fig 1, Table). Postcontrast T1-weighted MR images showed focal contrast enhancement in the ischemic lesions, which corresponded to disintegrated blood-brain barrier, as labeled with Evans blue dye (Fig 1). The lesion sizes defined at T2- and diffusion-weighted MR imaging and on ADC maps at this phase were significantly larger than those in the superacute phase (P < .0001, Table). However, there was no significant difference in lesion size on perfusion-weighted MR images between these two phases (P > .05, Table).
|
Discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods Results DiscussionREFERENCES |
|---|
Other factors, including rectangular field of view, relatively small bandwidth, and large number of signals acquired all contributed to a reasonable imaging time and quality. When compared with studies that have been performed with dedicated high-field-strength animal MR imagers, the maximal obtainable spatial resolution in our clinical setup is smaller (typically by a factor of five). This lower spatial resolution limits the possibilities of use of our setup to only certain stroke-related protocols in rats, where we expect relatively large affected areas. The use of low-field-strength clinical MR imagers also has its advantages, however, since it would be easier to extrapolate the obtained results with those of clinical human studies. The geometric distortion and susceptibility artifacts with these low-field-strength MR imagers are much lower compared with those with high-field-strength MR imagers. This and the ability to use parallel imaging techniques therefore allow us to use very fast echo-planar imaging sequences in the entire brain for perfusion and diffusion studies with a high temporal resolution. With high-field-strength dedicated MR imagers, a large geometric distortion is observed on echo-planar images; therefore, other acquisition sequences will be used that have a lower signal-to-noise ratio, a lower temporal resolution, or only a few echo-planar sections in combination with localized section shimming.
In contrast to the hyperintense lesion on T2-weighted MR images that resulted from interstitial edema and/or tissue necrosis during relatively late phases (2,17), the high signal intensity on diffusion-weighted MR images could be concordant with the fact that in the superacute phase, the formation of cytotoxic edema—that is, cell swelling due to partial relocation of water from extra- to intracellular compartment—is dominant, which subsequently results in the decrease of ADC of water in ischemic areas of the brain. This finding is consistent with results published previously (2,3).
Soon after MCA obstruction, the hyperintense area on diffusion-weighted MR images observed in our study was also recognized as the ischemic core by others who used animal imagers (1,3), whereas the broader area of hemodynamic disorders in MCA-supplied territory could only be detected with perfusion-weighted MR imaging as a perfusion defect. The mismatched region that resulted from subtracting the diffusion-weighted image by the perfusion-weighted image suggests a reversible ischemic margin that may provide a platform for assessment of the therapeutic effect of candidate antistroke drugs.
When compared with conventional studies involving a large number of experimental groups and/or animals, an MR imaging–monitored experiment can be performed in a longitudinal fashion without the need to sacrifice a large number of animals yet with more accurate results. Such study with more efficient economic and ecologic study designs would be more widely acceptable, especially in accordance with the current trend of animal and environmental protection. To evaluate noninvasively the therapeutic effects of new thrombolytic agents, such as microplasmin (22), on limiting the extent of ischemia, comparison of stroke status at early and end time points is mandatory on an individual basis.
As shown in the present study, contrast material–enhanced perfusion-weighted MR imaging proves most sensitive for identification of ischemia, which can occur as early as immediately after the beginning of an effective MCA obstruction. In the literature, perfusion-weighted MR imaging maps in rats were usually obtained with very-high-field-strength MR imagers (1). The results of our study demonstrated that—though less sensitive than the dedicated animal imagers, mainly in terms of spatial resolution—use of the 1.5-T clinical MR imager with common techniques, including perfusion- and diffusion-weighted imaging, turned out to be feasible and reliable in the evaluation of certain aspects of stroke in rats. The data become more convincing, especially after further verification both qualitatively and quantitatively with other postmortem standard-of-reference techniques, including microangiography, CT, and histomorphology, which are also clinically available.
Although the lesions defined at T2- and diffusion-weighted MR imaging and according to ADC in the subacute phase were significantly larger than those in the superacute phase (P < .0001), the lesion sizes on perfusion-weighted MR images remained statistically equivalent between these two phases (P > .05). Furthermore, the ischemic area derived from perfusion-weighted MR imaging and postmortem microangiography at 1 hour is comparable to the lesion defined with TTC staining, as well as T2- and diffusion-weighted MR imaging and ADC at 24 hours (P > .05), which confirmed the previous result that the final infarct size, without therapeutic intervention, can be predicted with early perfusion-weighted MR imaging (25). However, to what extent the results of perfusion-weighted MR imaging can be influenced by the pathologic process of compromised blood-brain barrier, as suggested in clinical studies (26,27), warrants further investigation in rodents.
The photochemically induced thrombosis model used in this study has some features that are different from photochemical models used elsewhere. First, dissimilar to distal MCA or end-artery photochemically induced thrombosis models, including the ring stroke model, the proximal left MCA at the olfactory tract was thrombotically occluded in rats. Second, the ischemic lesion—usually including the cortex and striatum—was relatively larger than that in other photochemical models (6,17–19,28). Third, criticisms of the photochemical model in the literature include its end-arterial occlusive nature and absence of a penumbra, which apparently refer to the photochemical models in which rat brain ischemic lesions were restricted only to the cortex (6,17–19). On the contrary, this photochemically induced thrombosis model of proximal MCA occlusion was considered to have a reperfusionlike phenomenon (15), and a rescuable region may exist.
Unlike other photochemical models (6,17–19), the evolution and imaging characteristics of rat brain ischemic lesions in this model have not been evaluated yet with noninvasive MR imaging. The encouraging results from the present study should be beneficial to the experimental research on neuroprotective effects of compounds and other potential diagnostic and therapeutic regimes with use of this model.
One limitation in the present study with the photochemically induced thrombosis model was relatively severe magnetic susceptibility artifact, which caused image distortion to some extent on diffusion- and perfusion-weighted MR images and would be even more problematic in the higher-field-strength dedicated animal systems. This resulted mainly from residual air-filled cavities left subtemporally after surgery. The magnetic susceptibility artifacts can be reduced by first filling the cavity with gelatin sponge to cancel the air-tissue interface effect and by applying a parallel imaging technique (GRAPPA; Siemens), a new k-space–based reconstruction algorithm that doubles the speed and improves signal-to-noise ratio (29). Although the use of GRAPPA (Siemens) decreased the absolute signal-to-noise ratio values, we could reclaim some signal-to-noise ratio by decreasing the echo time of the diffusion-weighted images in this case. Moreover, the most important effect was that the sequence became less prone to susceptibility artifacts, which led to better quality of the images—that is, less image distortion and less signal loss in some areas at the edge of the brain. Indeed, most of the optimized parameters in the present study had met the limits of the imager; further improvements of the imaging parameters and imaging quality are still possible with, for instance, implementation of a segmented echo-planar sequence, which was not yet available for our 1.5-T MR imager at the time this study was conducted.
In summary, focal cerebral ischemic lesions in rats could be examined and evaluated with a 1.5-T MR imager in conjunction with other clinical imaging modalities. The present photochemically induced thrombosis model of proximal MCA occlusion in rats appears to be ideal for stroke studies. With this experimental setting, we were able to perform a double-blind placebo-controlled study on the neuroprotective effects of microplasmin in rats and obtained encouraging results. This fact may justify the utility of the presently introduced method.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: ADC = apparent diffusion coefficient, MCA = middle cerebral artery, TTC = triphenyl tetrazolium chloride
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, Y.N., G.M.; study concepts, F.C., G.M., Y.N.; study design, F.C., Y.N.; literature research, F.C., Y.S.; experimental studies, F.C., Y.N., Y.S., N.N.; data acquisition and analysis/interpretation, all authors; statistical analysis, F.C., P.R.; manuscript preparation, F.C., Y.N.; manuscript definition of intellectual content, all authors; manuscript editing, F.C., Y.N.; manuscript revision/review, F.C., Y.N., P.R.; manuscript final version approval, Y.N., G.M.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods Results DiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F CHEN, Q LIU, H WANG, Y SUZUKI, N NAGAI, J YU, G MARCHAL, and Y NI Comparing two methods for assessment of perfusion-diffusion mismatch in a rodent model of ischaemic stroke: a pilot study Br. J. Radiol., March 1, 2008; 81(963): 192 - 198. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Chen, Y. Suzuki, N. Nagai, X. Sun, H. Wang, J. Yu, G. Marchal, and Y. Ni Microplasmin and Tissue Plasminogen Activator: Comparison of Therapeutic Effects in Rat Stroke Model at Multiparametric MR Imaging Radiology, August 1, 2007; 244(2): 429 - 438. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
