| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Departments of Diagnostic Radiology (A.C.L., S.G., N.H., W.S.R.), Pathology (R.M.B., S.v.G.), and Cardiology (G.W., H.H.), University of Giessen, Langhansstrasse 10, D-35392 Giessen, Germany. Received December 18, 2002; revision requested February 24, 2003; final revision received August 28; accepted October 21. Address correspondence to R.M.B. (e-mail: rainer.bohle@patho.med.uni-giessen.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: With micro CT, two-dimensional transverse images were generated from 10 human autopsy specimens of coronary arteries (2.5–3.5 cm long), with section thickness of 6 µm. Vessel wall perimeter, plaque area, calcified lesion area, media area, and lumen area were determined by three experienced radiologists. Results were compared with those obtained from a detailed conventional histomorphometric analysis of corresponding cross sections. Hotelling T2 test (a multivariate generalization of the univariate Student t test) and Pearson correlation coefficient were used to assess the correlation between micro CT findings and conventional histologic measurements. The significance of differences in gray-scale measurements was tested with analysis of variance.
RESULTS: Micro CT provided quantitative information about plaque morphology equivalent to that provided with histomorphometric analysis. Hotelling T2 test revealed significantly smaller values for vessel wall perimeter and lumen area with histologic sections (P < .001). Gray-scale measurements were established with which lesions could be categorized after histologic classification.
CONCLUSION: Micro CT is feasible for analysis of the coronary artery wall.
© RSNA, 2004
Index terms: Arteriosclerosis, 54.76 • Computed tomography (CT), microscopy, 54.1211, 54.76 • Computed tomography (CT), experimental studies, 54.1211, 54.76 • Coronary vessels, CT, 54.12119 • Coronary vessels, stenosis or obstruction, 54.76
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Clinical two-dimensional imaging studies (ie, classic angiography), in contrast, have limited ability for visualization of the three-dimensional vessel wall, as opposed to the lumen. Thus, histopathologic microscopy of human autopsy specimens and explanted tissue has been established as the standard for analysis of human arteries in the past decades. Conventional histomorphometry, however, also has its limitations. Histologic analysis as a method for evaluation of vessel wall architecture is substantially limited: (a) Serial sectioning of multiple thin slices from the tissue specimen is slow and costly. (b) Once sliced, the intact volume is lost so that further examinations with other methods are difficult if not impossible. (c) Histologic analysis does not provide three-dimensional information and does not allow continuous longitudinal measurements.
In the past, micro computed tomography (CT) became a powerful technique in laboratory investigation as technical advances in computer speed and memory enabled micro CT systems to generate thin-section images of small specimens (9–16). Although investigators in studies of the early implementation of three-dimensional micro CT focused on the technical and methodological aspects of this system, other researchers emphasized the practical aspects of this technology. So far, the technique has been successfully used in the visualization of vasculature in the following: intact isolated rodent organs (11); myocardial, renal, and hepatic vasculature (12,17–22); and trabecular bones (9,10) in surgical bone specimens.
To our knowledge, micro CT technology has not yet been investigated to assess its accuracy in the characterization of human atherosclerotic plaques and there have been no studies of the correlation between micro CT findings and histomorphometric data. Thus, the purpose of our study was to evaluate the feasibility of micro CT for the analysis of the coronary artery wall.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
During autopsy, the left and right coronary arteries were cannulated and perfused with a solution of 0.9% normal saline and heparin until the perfusate was free of blood. Unprocessed arteries were cut in tissue blocks at 2.5–3.5-cm intervals along the entire length of the vessel. Segments were stored in sealing film (Parafilm M; Pechiney Plastic Packaging, Chicago, Ill) to prevent distortion during scanning.
These sections were scanned with an 80-kV micro CT unit (SkyScan 1072; SkyScan, Aartselaar, Belgium). The x-ray system was based on a microfocus tube (20–80 kV, 0–100 µA) that had a minimum spot size of 8 µm at 8 W for generation of projection images and that produced x rays in a cone-beam geometric formation. Similar systems have been described in detail before (14–16). For our system, the manufacturer gives a maximum spatial resolution of 8 µm at 10% modulation transfer function, which was used in our study and is an accepted method for definition of the experimentally determined resolution and performance of an optical system (23). The x-ray detector consisted of a 12-bit digital water-cooled charge-coupled device high-resolution (1,024 x 1,024-pixel) camera, with fiberoptic coupling to an x-ray scintillator and digital frame grabber in a ratio of 3.7:1.0.
In our experimental setting, samples were positioned on a computer-controlled rotation stage and scanned 180° around the vertical axis in rotation steps of 0.675° at 40 kV. Acquisition time for each view was 2.4 seconds. Relative position of the object from the source determined geometric magnification, and thus, the pixel size was defined on the basis of the cone-beam geometry of the system. Maximum possible magnification was also limited by the specimen size, and the specimen had to be positioned so that the horizontal diameter was within the cone beam. We used magnifications to x80, which led to a pixel size of 2 µm. Raw data were reconstructed with a modified Feldkamp cone-beam reconstruction modus (24), which resulted in two-dimensional 8-bit gray-scale images consisting of cubic voxels. Image reconstruction was performed with a dual 1,800-MHz processor (Intel Xeon; Intel, Santa Clara, Calif) equipped with 1,024 MByte of RAM. With a general acquisition time of approximately 10 cross sections in 1 minute, 43,831 images were generated. The 43,831 images resulted from the section thickness (6 µm) and the specimen size (2.5–3.5 cm long).
Histopathologic Findings
After standard embedding, routine hematoxylin-eosin staining with elastica counterstaining was performed in histologic cross sections obtained from 10 coronary artery segments (2.5–3.5 cm long). In addition, immunohistochemical analysis was performed. To exclude any possible influence of micro CT on subsequent immunochemical staining, immunohistochemical analysis was performed in specimens investigated with micro CT and in specimens processed for immediate immunohistochemical analysis. For immunohistochemical analysis, 5-µm-thick sections were cut from each paraffin-embedded tissue block every 5 mm, which resulted in 65 cross sections, and these sections were transferred to 75-µm glass slides (ChemMate capillary gap microscope slides; Dako, Glostrup, Denmark). After deparaffinization in xylene, acetone, and a mixture of acetone and Tris-buffered saline (Roth, Karlsruhe, Germany) at each step for 10 minutes, the sections were washed twice with Tris-buffered saline. Staining was performed in an automated immunostainer (TechMate 500; Dako) with a detection kit (ChemMate Detection Kit APAAP Mouse; Dako).
The following mouse monoclonal antibodies, working dilutions, and antibody diluents from the manufacturer of the detection kit were used: anti–smooth-muscle actin, 1:100 (clone 1A4; Dako); anti–von Willebrand factor, 1:50 (clone F8/86; Dako); anti-CD68, 1:50 (clone PG-M1; Dako); and anti-CD34, 1:100 (clone QBEND10; Immunotech, Marseille, France). For staining of smooth-muscle actin, heat-induced antigen retrieval was performed in a microwave oven (heating three times at 600 W for 5 minutes each) with citrate buffer at pH 6.0. For CD34 staining, heat-induced antigen retrieval was performed with ethylenediaminetetraacetic acid buffer at pH 8.0. For CD68 staining, retrieval was performed after 30 minutes with a trypsin solution containing 100 µL trypsin (GibcoBRL Life Technologies, Paisley, Scotland) in 100 mL of Tris-buffered saline (Roth). All incubation steps lasted 25 minutes. Immunohistologic sections were counterstained with hematoxylin and mounted in gelatin. Negative controls with tissue sections from human atherosclerotic coronary arteries (n = 3) were prepared by using 100 µL of mouse anti-rabbit immunoglobulin (clone MR12/52; Dako) as the primary antibody, with a working dilution of 1 µg immunoglobulin in 1 mL antibody diluent. Histologic cross sections were digitized, and image analysis was performed with a software package (Analyze, version 4.0; Biomedical Imaging Resource, Mayo Foundation, Rochester, Minn).
Comparison of Imaging and Histopathologic Findings
Image analysis was performed with the same commercial software package mentioned before. This software yielded various three-dimensional visualization tools. We basically used spatial filtering, multiplanar reformation, volume rendering, object extraction, and region-of-interest measurement. Volume measurements of objects were obtained by adding the voxel size to the header. Histologic slides were digitized after application of a scale of 2 mm imported into the software, and pixel size was added to the header, which provided the possibility of performance of measurements in these sections. The corresponding section for each histologic section was located in the three-dimensional data set with a visualization tool that provided oblique views. In-plane measurements of plaque diameters and areas were performed and compared with data derived with histomorphometric analysis. Volumetric data about plaque sizes, as well as gray-scale measurements in the plaques, were analyzed.
For correlation of micro CT findings and histomorphometric data, measurements of atherosclerotic lesions were performed on digitized histologic images and the corresponding micro CT images, respectively. Images were analyzed quantitatively by two experienced pathologists (R.M.B., 20 years of experience; S.v.G., 7 years of experience) for plaque area, lumen area, media area, vessel wall perimeter, and total calcified area. We obtained a total number of 65 pairs of micro CT images and histologic cross sections. Histopathologic analysis was performed by the pathologists independently, and they were blinded to the micro CT analysis performed by three radiologists (A.C.L., W.S.R., S.G.) who each had 4 years of experience with micro CT.
Next, gray-scale attenuation of micro CT images was compared with histologically stained plaque components. Because no calibration block was implemented in our micro CT scanner, classic Hounsfield units could not be declared. For determination of relative attenuation values, square regions of interest with a side length of 0.5 mm (range, 0–255 mm) were placed by one author (A.C.L.) and were established manually in various areas of plaque, with visual assessment of gray-scale differences. Gray-scale attenuation was measured in five areas in the plaque on each micro CT image by using the same software package mentioned before. Differences in gray-scale values in the plaques were compared with conventional cross sections histologically classified as lipid-rich plaques, fibrous plaques (with or without calcification), and atheroma with rare smooth-muscle cells (with or without calcification).
Statistical Analysis
Hotelling T2 test (a multivariate generalization of the univariate Student t test) was used to assess the correlation between micro CT and conventional histologic measurements. The distributional assumption of multivariate normality must be checked to assess the validity of the Hotelling T2 test. This checking was performed by using two tests developed by Mardia et al (25), one based on empirical skewness and the other based on empirical kurtosis.
All continuous variables are expressed as the mean ± standard error of the mean. Analysis of variance was performed to test the statistical significance of differences in gray-scale measurements. A P value of less than .05 was considered to indicate a significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Comparisons between Micro CT and Conventional Histologic Findings
As illustrated in Figure 1, images obtained at micro CT closely approximated those obtained at histologic microscopy and, with spatial resolution of 8 µm, allowed precise plaque visualization. Significant correlations between results at micro CT image analysis and conventional histomorphometric findings were obtained for lesion area, calcified lesion area, and media area (Table; Figs 1, 2). Measurements performed in histologic sections revealed smaller values for vessel wall perimeter and lumen area than did measurements performed at micro CT (Hotelling T2 test).
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
First, micro CT proved to be a rapid method for qualitative and quantitative evaluation of atherosclerotic lesions. With a mean acquisition time of approximately 10 cross sections per minute, acquisition time was faster with micro CT than was the processing time with the mechanical method of conventional histologic analysis. This technical benefit might become more important, with consideration of findings in a recent study (1). In that study, morphologic predictors of arterial remodeling in coronary atherosclerosis were evaluated with 2,885 coronary cross sections at histologic examination. In the future, such labor-intensive and time-consuming investigations might be augmented with use of micro CT.
Second, in the present study, micro CT emerged as a technique (at least equal to histologic analysis) that could be used for detection of atherosclerotic lesions in coronary vessels. Micro CT was shown to help in the detection of even small early lesions that had been missed at conventional histologic examination. This finding may underscore the superiority of continuous visualization of the vessel segment by using micro CT compared with the segmental examination of tissue specimens during histologic cross-sectioning.
Third, with micro CT images, accurate characterization of lesion morphology was obtained. Measurements of morphometric indexes performed with micro CT led to the same results obtained with conventional histologic analysis. This finding indicates that micro CT is an appropriate tool for quantitative analysis of atherosclerotic plaques. In addition, on the basis of differences in gray-scale attenuation, micro CT helped in the correct identification of atherosclerotic lesions that were histologically classified as fibrous plaques, calcified lesions, fibroatheromas, and lipid-rich lesions. It seems conceivable that with future technical advances, micro CT might enable visualization of distinct cellular plaque composition.
Fourth, in the present study, micro CT was performed in unprocessed coronary arteries. Because micro CT analysis was performed prior to histologic processing (eg, fixation, dehydration, or decalcification before paraffin embedding), mechanical or physiochemical artifacts, such as shrinkage of the specimen, were prevented. The observed statistically significant differences in vessel wall perimeter and lumen area are within a few micrometers of each other and presumably reflect the well-known shrinkage phenomenon provoked by histologic procedures. Conventional histologic analysis also is limited in that once tissue is sliced the intact volume is lost so that further examinations of the specimens with other methods are difficult if not impossible. Micro CT neither restrained subsequent conventional histomorphologic analyses nor interfered with specific immunohistologic staining performed on the same coronary vessel segments. Since further histologic analysis can be performed in the same specimens, the use of micro CT may be of great advantage in clinical or experimental studies.
Fifth, to the present, quantitative analyses of atherosclerotic lesions typically are performed at light microscopy in thin histologic sections. One problem with this approach is the difficulty of obtaining adjacent sections and the occurrence of distortions during the sectioning process. Histologic microscopy also is limited in that it does not provide three-dimensional information; as a destructive technique, it does not allow continuous measurements; and it limits tissue quantification to a small number of two-dimensional sections. In addition, only a few micrometers of the vessel wall can be analyzed histologically. With micro CT, a complete digital data set of the vessel wall is now available. As a nondestructive approach, micro CT allows analysis of the various plaque characteristics such as plaque area, lumen area, calcified lesion area, or lipid core area over the entire length of the investigated vessel segment. By using tomographic reconstruction algorithms, three-dimensional images of the vessel wall can be generated that allow total stereoscopic visualization of the three-dimensional microarchitecture of the plaques. Since tomography allows convenient extraction of appropriate sections from the three-dimensional images, micro CT also can be used to analyze distinct regions of interest more accurately (eg, high-risk lesions with thin fibrous caps can be detected prior to dissection of the specimen). Thus, "steering" of the pathologist’s knife to distinct local alterations within the tissue specimen is accomplished.
In our study, micro CT was used for quantification of vascular structures, but only structures delineated with sufficient contrast material can be visualized and quantified exactly. Here, histologic analysis shows clear advantages over micro CT. Thus, micro CT should be considered as an additional tool besides histologic analysis because each has its own advantages and disadvantages. Neither of these methods can fully replace the other. Issues of particular concern include those related to the scanner characteristics, such as spatial and gray-scale resolution. A more detailed classification of different plaque types according to gray-scale attenuation is the focus of ongoing work.
Practical application: To our knowledge, this is the first time that micro CT technology has been applied to describe and quantify atherosclerotic lesions in human coronary arteries. We demonstrated the feasibility of using micro CT to assess morphologic characteristics of atherosclerotic lesions in a manner that approaches and, in some aspects, clearly exceeds the results with histologic microscopy. Thus, micro CT should be considered as an additional tool for ex vivo assessment of atherosclerosis, and it has the potential to become a standard technique in many laboratories.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, A.C.L., H.H., R.M.B., W.S.R.; study concepts, A.C.L., R.M.B.; study design, A.C.L., H.H., R.M.B.; literature research, H.H., R.M.B., G.W.; experimental studies, A.C.L., R.M.B., H.H., W.S.R.; data acquisition, S.G., S.v.G., N.H.; data analysis/interpretation, G.W., H.H., R.M.B.; statistical analysis, A.C.L., G.W.; manuscript preparation, A.C.L., W.S.R.; manuscript definition of intellectual content, W.S.R., R.M.B.; manuscript editing, R.M.B., A.C.L.; manuscript revision/review, H.H., R.M.B., S.v.G.; manuscript final version approval, A.C.L., R.M.B., H.H.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Goel, B. R. Schrank, S. Arora, B. Boylan, B. Fleming, H. Miura, P. J. Newman, R. C. Molthen, and D. K. Newman Site-Specific Effects of PECAM-1 on Atherosclerosis in LDL Receptor-Deficient Mice Arterioscler. Thromb. Vasc. Biol., November 1, 2008; 28(11): 1996 - 2002. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X.-Y. Zhu, M. D. Bentley, A. R. Chade, E. L. Ritman, A. Lerman, and L. O. Lerman Early changes in coronary artery wall structure detected by microcomputed tomography in experimental hypercholesterolemia Am J Physiol Heart Circ Physiol, September 1, 2007; 293(3): H1997 - H2003. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Vengrenyuk, S. Carlier, S. Xanthos, L. Cardoso, P. Ganatos, R. Virmani, S. Einav, L. Gilchrist, and S. Weinbaum A hypothesis for vulnerable plaque rupture due to stress-induced debonding around cellular microcalcifications in thin fibrous caps PNAS, October 3, 2006; 103(40): 14678 - 14683. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Langheinrich, A. Michniewicz, D. G. Sedding, G. Walker, P. E. Beighley, W. S. Rau, R. M. Bohle, and E. L. Ritman Correlation of Vasa Vasorum Neovascularization and Plaque Progression in Aortas of Apolipoprotein E-/-/Low-Density Lipoprotein-/- Double Knockout Mice Arterioscler. Thromb. Vasc. Biol., February 1, 2006; 26(2): 347 - 352. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Serfaty, M. Nonent, N. Nighoghossian, F. Rouhart, L. Derex, C. Rotaru, P. Chirossel, G. Thabut, B. Guias, J-F Heautot, et al. Plaque density on CT, a potential marker of ischemic stroke Neurology, January 10, 2006; 66(1): 118 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Savai, J. C. Wolf, S. Greschus, B. G. Eul, R. T. Schermuly, J. Hanze, R. Voswinckel, A. C. Langheinrich, F. Grimminger, H. Traupe, et al. Analysis of Tumor Vessel Supply in Lewis Lung Carcinoma in Mice by Fluorescent Microsphere Distribution and Imaging with Micro- and Flat-Panel Computed Tomography Am. J. Pathol., October 1, 2005; 167(4): 937 - 946. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Watz, A. Breithecker, W. S. Rau, and A. Kriete Micro-CT of the Human Lung: Imaging of Alveoli and Virtual Endoscopy of an Alveolar Duct in a Normal Lung and in a Lung with Centrilobular Emphysema--Initial Observations Radiology, September 1, 2005; 236(3): 1053 - 1058. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Mohr, F. Roemer, H. Genant, A. C. Langheinrich, and R. M. Bohle Analysis of Atherosclerosis with Micro-CT * Drs Langheinrich and Bohle respond: Radiology, April 1, 2005; 235(1): 338 - 339. [Full Text] [PDF]
|
|
|
|
|
|
P. J. Pickhardt, R. B. Halberg, A. J. Taylor, B. Y. Durkee, J. Fine, F. T. Lee Jr., and J. P. Weichert Microcomputed tomography colonography for polyp detection in an in vivo mouse tumor model PNAS, March 1, 2005; 102(9): 3419 - 3422. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
