- Email: [email protected]
Quantification of coronary artery stenosis with high-resolution CT in comparison with histopathology in an ex vivo study
European Journal of Radiology 82 (2013) 264–269
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Quantification of coronary artery stenosis with high-resolution CT in comparison with histopathology in an ex vivo study Matthias Dettmer a,1 , Nicola Glaser-Gallion b,2 , Paul Stolzmann c,3 , Florian Glaser-Gallion b,2 , Juergen Fornaro b,2 , Gudrun Feuchtner d,4 , Wolfram Jochum e,5 , Hatem Alkadhi c,3 , Simon Wildermuth b,6 , Sebastian Leschka b,∗ a
Department of Pathology and Laboratory Medicine, University of Pittsburgh, S-417 BST 200 Lothrop Street, Pittsburgh, PA 15261, USA Institute of Radiology, Kantonsspital St. Gallen, Rorschacherstrasse 95, 9007 St. Gallen, Switzerland c Institute of Diagnostic and Interventional Radiology, University Hospital Zurich, Rämistrasse 100, 8091 Zurich, Switzerland d Department of Radiology II, Medical University Innsbruck, Anichstr. 35, 6020 Innsbruck, Austria e Institute of Pathology, Kantonsspital St. Gallen, Rorschacherstrasse 95, 9007 St. Gallen, Switzerland b
a r t i c l e
i n f o
Article history: Received 25 June 2012 Received in revised form 15 September 2012 Accepted 21 September 2012 Keywords: Coronary artery plaque Atherosclerosis Computed tomography Histology Stenosis Stary classification
a b s t r a c t Purpose: To investigate the ex vivo performance of high-resolution computed tomography (CT) for quantitative assessment of percentage diameter stenosis in coronary arteries compared to histopathology. Materials and methods: High-resolution CT was performed in 26 human heart specimens after the injection of iodinated contrast media into the coronary arteries. Coronary artery plaques were visually identified on CT images and the grade of stenosis for each plaque was measured with electronic calipers. All coronary plaques were characterized by histopathology according to the Stary classification, and the percentage of stenosis was measured. Results: CT depicted 84% (274/326) of all coronary plaques identified by histology. Missed plaques by CT were of Stary type I (n = 31), type II (n = 16), and type III (n = 5). The stenosis degree significantly correlated between CT and histology (r = 0.81, p < 0.001). CT systematically overestimated the stenosis of calcified plaques (mean difference - 11.0 ± 9.5%, p < 0.01) and systematically underestimated the stenosis of noncalcified plaques (mean difference −6.8 ± 10.4%, p < 0.05), while there was no significant difference for mixed-type plaques (mean difference −0.4 ± 11.7%, p = 0.85). There was a significant underestimation of stenosis degree as measured by CT for Stary II plaques (mean difference −14 ± 9%, p < 0.01) and a significant overestimation for Stary VII plaques (mean difference 9 ± 10%, p < 0.05), but there was no significant difference in stenosis degree between both modalities for other plaque types. Conclusions: High-resolution CT reliably depicts advanced stage coronary plaques with an overall good correlation of stenosis degree compared to histology, however, the degree of stenosis is systematically overestimated in calcified and underestimated in non-calcified plaques. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
∗ Corresponding author. Tel.: +41 71 494 2270; fax: +41 71 494 6479. E-mail addresses: [email protected] (M. Dettmer), [email protected] (N. Glaser-Gallion), [email protected] (P. Stolzmann), fl[email protected] (F. Glaser-Gallion), [email protected] (J. Fornaro), [email protected] (G. Feuchtner), [email protected] (W. Jochum), [email protected] (H. Alkadhi), [email protected] (S. Wildermuth), [email protected] (S. Leschka). 1 Tel.: +1 412 864 3351. 2 Tel.: +41 71 494 1111; fax: +41 71 494 6479. 3 Tel.: +41 44 255 1111; fax: +41 44 255 1819. 4 Tel.: +43 512 504 81898; fax: +43 512 504 24029. 5 Tel.: +41 71 494 2102; fax: +41 71 494 2894. 6 Tel.: +41 71 494 2181; fax: +41 71 494 6479. 0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejrad.2012.09.021
Identification of coronary atherosclerotic plaques improves the risk stratification with regard to the occurrence of coronary events [1]. Assessing patients with arterial stenosis with different imaging techniques is important for determining optimal therapy [2]. Traditionally, imaging of coronary artery disease (CAD) was based on catheter coronary angiography. However, catheter angiography is limited by luminographic visualization of the coronary artery and its inability to visualize coronary atherosclerotic plaques [3]. Intravascular ultrasound (IVUS) is considered as the clinical reference modality for most accurate characterization and detection of coronary atherosclerotic plaques, regarding plaque morphology, and regarding quantification of arterial stenosis [3]. However, IVUS is an invasive procedure and thus, its broad application in larger
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
patient populations for risk stratification is limited. It is apparent that a non-invasive technique for the early detection and characterization of the content of coronary atherosclerotic plaques is highly desirable. Cardiac computed tomography (CT) is an emerging technique that allows for the accurate, non-invasive assessment of coronary artery stenosis and coronary plaques [3,4]. Being a cross-sectional imaging modality, CT allows for the evaluation of the arterial wall in addition to the assessment of luminal narrowing [6]. Several studies have quantified the grade of stenosis in coronary arteries with CT in comparison with IVUS or conventional coronary angiography (CCA) [5,6]. Some studies comparing CT with conventional coronary angiography with a good overall correlation between the two modalities for cross-sectional and longitudinal vessel reconstructions have been done with even improved diagnostic accuracy using image postprocessing methods [8]. However, these data are questioned by others who could not find a correlation, and reliability for CT to detect coronary artery stenoses precisely in comparison to angiography [5]. A better correlation between CT and IVUS than between angiography and IVUS has been shown [9]. These data suggest that catheter angiography – albeit being considered as gold standard – has limitations [10]. To overcome the limitations of an imaging technique as a gold standard, we compared the CT measurements of coronary stenosis degree with histopathology. As a matter of fact, some studies have tried to determine the morphology of coronary atherosclerotic plaques in human heart specimens in comparison with histopathology [6,7,12] and in patients in comparison with IVUS [5,7] and angiography [13]. Recent CT studies comparing coronary plaques with histopathology using 4-slice [11] and 16-slice technology [12] have shown promising results. However, the depiction of coronary artery plaques was limited by the relatively low spatial and temporal resolution of the former CT systems used. In addition, the sample sizes in those studies were small, including only up to 50 plaques [11]. One of the most recent innovations in CT technology, i.e., dual-source CT, is characterized by a spatial resolution of 0.4 mm × 0.4 mm × 0.4 mm and a temporal resolution of 83 ms. Thus, the purpose of this study was to investigate the performance of high-resolution CT to depict and quantify coronary artery stenosis in comparison to histopathology in human heart specimens. 2. Materials and methods 2.1. Study objects This study was approved by the local ethics committee. The hearts of 26 patients (8 women, 18 men; mean age 70 ± 12 years; age range 38–85 years) were included. Inclusion criteria were age over 35 years and an autopsy had to be intended. All cause of death was accepted for study inclusion. Cause of death were pulmonary embolism (n = 7), end stage malignancy (n = 6), sepsis (n = 5), acute myocardial infarction (n = 4), acute respiratory distress syndrome (n = 2), ventricular fibrillation (n = 1), and myocardial rupture due to subacute myocardial infarction (n = 1). 2.2. Specimen preparation The autopsies were performed according to standard techniques. After explanation from the corpse, the hearts were washed in water without further fixation or staining, cooled on crushed iced, and immediately transferred to the CT imaging suite. The ostia of the left and the right coronary artery were cannulated with a 5-French balloon catheter (6.0 mm × 2.0 mm) and the coronary arteries were filled under pressure with a mixture of iodinated
265
contrast medium (Ultravist 370, 370 mg/mL, Bayer Schering Pharma, Berlin, Germany) and saline solution to achieve a target attenuation of 300 Hounsfield units (HU) within the coronary lumen. Thereafter, the balloons were blocked to avoid contrast rinsing out of the coronary arteries. 2.3. CT protocol The prepared specimen was placed in the middle of the CT scanner. All examinations were performed with a dual-source CT scanner (Somatom Definition, Siemens Healthcare, Forchheim, Germany). The following parameters were used for imaging: detector collimation 2 mm × 32 mm × 0.6 mm, slice acquisition 2 mm × 64 mm × 0.6 mm, gantry rotation time 330 ms, pitch 0.2, tube potential 120 kV, and tube current time product 330 mAs per rotation. This scan protocol is similar to that commonly used for the in vivo imaging of the coronary arteries [4]. The data was reconstructed with a slice thickness of 0.6 mm, a reconstruction increment of 0.4 mm, and using a soft-tissue convolution kernel (B30) and a sharp kernel (B46). The field of view was set to 205 mm × 205 mm (matrix: 512 × 512 pixels) to achieve an inplane spatial resolution of 0.4 mm × 0.4 mm. 2.4. CT data analysis The presence of stenosis in the coronary arteries was evaluated immediately after the CT scan on the contrast-enhanced CT dataset by one experienced cardiovascular radiologist. The exact position of the plaques was noted by distance measurements from the coronary ostium in order to assure accurate histopathological sectioning. For each plaque reconstructions perpendicular to the vessel centerline were performed and used for data analysis. The radiologist analyzed the perpendicular reconstructions and selected the image demonstrating the highest degree of stenosis of each plaque. In each plaque one free-hand region of interest was circled around the contrast filled lumen and one region of interest around the vessel outline and the area of both circles were noted. Both measurements were performed twice. The mean of the repeated measurements was used to calculate the degree of stenosis by dividing the lumen area with the vessel area. Each plaque was classified as being non-calcified (absence of any calcified deposits), calcified (absence of any non-calcified components), and mixed-type (presence of non-calcified and calcified components). Measurements for all plaque types (e.g., non-calcified, mixed plaques, and calcified plaques) were performed on softtissue and sharp kernel reconstructions. 2.5. Histopathologic analysis After CT scanning, standard pathological sectioning of the heart was performed. All histopathologic analyses were performed by the same experienced pathologist. Transverse sections of the coronary arteries were obtained for all coronary artery plaques detected by CT and also for the remaining part of the coronary tree. Coronary artery stenosis was identified by measuring the distance of the plaques from the coronary ostium. Coronary arteries were fixed in 10% buffered formalin, decalcified if necessary and embedded in paraffin. Two-micron thick sections were mounted on glass slides, deparaffinized, rehydrated and stained with hematoxylin-eosin and elastica-van gieson using standard histological techniques. To quantify stenosis, the percentage of the area of the coronary artery around the elastica interna and the lumen was measured. To quantify stenosis, the area of the coronary artery around the lumen, the lamina elastica interna and around the lamina elastica externa was measured and a relationship was calculated. All plaques were classified according to
266
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
the Stary classification [14] using a lightmicroscope (Zeiss Axiophot, Carl Zeiss, Jena, Germany) and Imageaccess© . Stary type I–III plaques were considered as early stage plaques, and Stary type IV–VIII plaques were considered as advanced stage plaques. Histopathological classification was considered the standard of reference. 2.6. Statistical analysis A p value of <.05 was considered statistically significant for all tests. Continuous variables were expressed as means ± standard deviations and categorical variables as frequencies or percentages. Overall differences in measurement of stenosis degree between CT and histology were compared by using Pearson’s correlation analysis and Bland–Altman scatterplot. The Wilcoxon signed-rank test was used for pair-wise comparisons. The analysis of variance test was used to statistically compare measurement differences between CT and histology in the different Stary type plaques. Statistical significance of differences in stenosis degree measurement between CT and histology were calculated by using a one-sample t test against zero. 3. Results Histological analysis depicted 326 coronary atherosclerotic plaques in the 26 heart specimens: 31 plaques (10%) were classified as Stary type I, 26 plaques (8%) as Stary II, 27 plaques (8%) as Stary III, 29 plaques (9%) as Stary IV, 73 plaques (22%) as Stary V, 22 plaques (7%) as Stary VI, 62 plaques (19%) as Stary VII, and 56 plaques (17%) as Stary VIII. A total of 274 coronary plaques were identified in the CT coronary angiograms. The 52 plaques missed by CT were classified as Stary type I (n = 31/31, 100%), type II (n = 16/26, 62%), and type III (n = 5/27, 17%). Thus, the overall depiction rate of CT for Stary type I-VIII plaques was 84% (274/326). 3.1. Overall accuracy of stenosis degree quantification by CT in comparison to histology
Fig. 1. Linear regression plot of stenosis degree measured by histology (y-axis) against stenosis degree measured by CT (x-axis). Dashed lines represent 95% confidence limits. Linear correlation indicates significant correlation of stenosis degree measurement between CT and histology (Pearson correlation; r = 0.81, p < 0.001). The different Stary type plaques are indicated by markers.
image noise, and measurements in sharp kernel reconstructions were used for calcified plaques to compensate for blooming artifacts. Comparing the stenosis degree measurements between CT and histology, CT significantly underestimated the stenosis degree for non-calcified plaques (p < 0.05) and overestimated the stenosis
For the 274 coronary plaques identified by both CT and histology, the overall stenosis degree measurement significantly correlated between CT and histology (r = 0.81, p < 0.001, Fig. 1). Bland–Altman analysis revealed a minor overestimation of stenosis degree measurement by CT (mean 0.5%) and a wide range of the levels of agreement (−24.3% to 25.3%, Fig. 2). 3.2. Differences in stenosis degree measurement for plaques classified as non-calcified, mixed or calcified by CT Image examples of stenosis degree measurements for noncalcified, mixed, and calcified plaques are demonstrated in Fig. 3. CT classified 29% (80/274) of the coronary artery plaques as noncalcified, 46% (127/274) as mixed, and 25% (67/274) as calcified. The stenosis degree was significantly higher for calcified plaques compared to non-calcified (p < 0.01) and mixed plaques (p < 0.05, Table 1). Mixed plaques had a significantly higher stenosis degree than non-calcified plaques (p < 0.05). There was no significant difference in the stenosis degree between soft tissue and sharp reconstruction kernel for noncalcified (p = 0.76) and mixed plaque (p = 0.51), while the stenosis degree of calcified plaques was significantly lower for sharp compared to soft tissue kernel reconstruction (p < 0.01). The difference of stenosis degree measured by CT and histology for calcified plaques was lower for the sharp kernel reconstruction (mean difference 9 ± 14%) as when using a soft tissue kernel (mean difference 20 ± 19%). Thus, measurements in soft-tissue kernel reconstructions were used for non-calcified and mixed plaques to avoid higher
Fig. 2. Bland–Altman plot of the agreement of stenosis degree measurements between CT and histology. There is a minor overestimation of stenosis degree measurement by CT (mean 0.5%, bold line) with high levels of agreement (−24.3% to 25.3%, dashed lines). The different Stary type plaques are indicated by markers.
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
267
Fig. 3. Measurement of degree of stenosis with histology and CT in non-calcified (A), mixed (B), and calcified plaques (C). In the non-calcified plaque, area stenosis was 23% at histology and 17% at CT. In the mixed plaque, area stenosis was 82% with histology and 79% with CT. In the calcified plaque, area stenosis was 55% with histology. Using the B30 kernel for CT area stenosis was 79% while area stenosis was 62% with the B46 kernel.
degree for calcified plaques (p < 0.01). The stenosis degree measurements for mixed plaques were comparable between CT and histology. 3.3. Differences in stenosis degree measurement between CT and histology for the different Stary type plaques Table 2 summarizes stenosis degree measurements of CT and histology for the different Stary type plaques. The difference in stenosis degree measurements between CT and histology were not significantly different among the different Stary type plaques
(p = 0.16). The stenosis degree as measured by CT revealed a significant underestimation for Stary II plaques (mean difference −14 ± 9%, p < 0.01) and a significant overestimation for Stary VII plaques (mean difference 9 ± 10%, p < 0.05). There was a significantly higher stenosis degree for advanced stage plaques (i.e., Stary IV–VIII) compared to early stage plaques (i.e., Stary II and III) for both CT and histology (each p < 0.05). The stenosis degree as measured by CT revealed a significant underestimation for early stage plaques (mean difference −8 ± 15%, p < 0.05) and a significant overestimation for advanced stage plaques (mean difference 5 ± 12%, p < 0.05).
Table 1 Measurement of stenosis degree by CT and histology for non-calcified, mixed, and calcified plaques. Stenosis degree measured by CT
b
Non-calcified plaques Mixed plaquesb Calcified plaquesb a b
Soft tissue kernel (B30)
Sharp kernel (B46)
52 ± 19% 73 ± 14% 95 ± 10%
51 ± 20% 71 ± 16% 86 ± 12%
Stenosis degree measured by histology
Difference in measurements
p valuea
59 ± 18% 73 ± 14% 75 ± 14%
−6.8 ± 10.4% −0.4 ± 11.7% 11.0 ± 9.5%
<0.05 0.85 <0.01
The p values were calculated using a one-sample t test against zero. Classification in non-calcified, mixed, and calcified according to CT plaques composition.
268
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
Table 2 Measurement of stenosis degree by CT and histology for the different Stary type plaques. Stenosis degree measured by CT b
Stary type II Stary type IIIb Stary type IVb Stary type Vb Stary type VIb Stary type VIIb Stary type VIIIb Early stage Advanced stage a b
24 41 56 70 77 87 75 30 76
± ± ± ± ± ± ± ± ±
7% 14% 18% 14% 14% 13% 12% 14% 19%
Stenosis degree measured by histology 38 45 61 72 76 78 71 38 71
± ± ± ± ± ± ± ± ±
13% 14% 16% 16% 12% 13% 12% 15% 16%
Difference in measurements
p valuea
−13.9 ± 8.5% −3.8 ± 13.9% −4.5 ± 12.1% −2.7 ± 11.4% 1.0 ± 12.2% 9.3 ± 10.1% 4.1 ± 10.5% −8.1 ± 14.6% 4.5 ± 12.1%
<0.01 0.34 0.11 0.52 0.74 <0.05 0.41 <0.05 <0.05
The p values were calculated using a one-sample t test against zero. Histopathologic classification of coronary atherosclerotic plaques were done according to Stary.
4. Discussion 4.1. Overall accuracy of stenosis degree measurement by CT in comparison to histology The reliable detection of coronary artery stenosis by CT is a challenge and the first attempts to quantify the stenosis degree have been made in 2000, at that time with a minor correlation [5]. It could be shown, that coronary computed tomography angiography is able to reliable predict high grade lesions of more than >70% grade of stenosis and that it is possible to measure an excellent correlation of stenosis degree using angiography as a standard of reference [15]. There is also a good correlation between CT in comparison with IVUS [7,15]. Atherosclerosis is a process which is generalized and not limited to a special segment of whatever artery in the human body. This means for the coronary arteries, that they are all affected by the disease, some segments more than others. A main problem for the conventional imaging techniques is the fact, that most of them compare the segment before and after the stenosis of interest with the stenosis itself. It is often forgotten, that these segments already themselves are affected by atherosclerotic changes. These limitations do not exist for histopathology as we could measure the actual grade of stenosis much more precise. CT is rarely compared with histopathology. The first attempt to detect coronary artery stenosis by quantification of calcifications in comparison to histology was in the early nineties, where no correlation between the two modalities could be found [16]. Modern CT systems provide a high spatial resolution of 0.4 mm3 . We could show a significant correlation of the overall stenosis degree for the 274 coronary plaques identified by the two modalities. 4.2. Differences in stenosis degree measurement for plaques classified as non-calcified, mixed or calcified by CT Coronary artery plaques are widely classified by CT in noncalcified, mixed and calcified plaques and the degree of coronary calcifications has a strong predictive value concerning cardiovascular events in the future of patients [17]. The classifications are based on the presence or absence of calcifications, by density measurements of the plaque, or by both methods combined. However, it has to be admitted that the characterization in “vulnerable” and “stable” plaques based on their CT density measurements is limited [18]. A consisting problem while trying to quantify the grade of stenosis is the systematic underestimation of non-calcified plaques and mixed plaques whereas calcified plaques are systematically overestimated in comparison to IVUS [18]. Our results show a better agreement than previous 64-slice CT [7] and are about in the range observed by Dewey et al. [19]. On the other hand, it has to be admitted that CT systematically
overestimates the degree of stenosis in calcified and underestimates it in non-calcified plaque, while CT is accurate in stenosis quantification in mixed plaques. 4.3. Differences in stenosis degree measurement between CT and histology for the different Stary type plaques Up to now, there have no attempts been made to correlate the degree of stenosis, measured by CT with the different Stary-type plaques. The Stary-classification has been correlated with 16-sliceCT, but a classification beyond the three-tired system non-calcified, mixed, calcified was not possible and the grade of stenosis was not assessed [20]. Compared to histology, Becker et al. [11] reported a moderate correlation of the three-tired classification scheme with the histological Stary classification, but a grade of stenosis was also not evaluated. In our study, there were no differences in stenosis degree measurements between CT and histology for the Stary type plaques III–VI and type VIII while we found a significant difference between the two modalities for the early non-calcified lesion Stary II and the heavy calcified lesion Stary VII. Revealed by Bland–Altman, the stenosis degree of the early and the mixed lesions (e.g. Stary I–IV) were more likely to be underestimated in contrast to the advanced lesions (Stary V–VIII) which were more likely to be overestimated. The underestimation of early lesions is due to the limited spatial resolution while it is known that calcification of plaques lead to overestimation of stenosis. 4.4. Study limitations First, no moving or beating of the heart specimens was simulated. Thus, our results represent a “best case scenario” which cannot be fully transferred to an in vivo clinical situation. Even though the CT system used in our study provides the highest currently available temporal resolution, artifacts deriving from cardiac motion causing blurring of the vessel wall may still occur [4]. Second, we did not test the possibility of dual-source CT with acquisition of data with two different energy levels. This technique has shown to allow for an improved differentiation of various tissue components. Third, we observed a partial luminal collapse during tissue fixation on the histological slides. This phenomenon may partially explain an overestimation of stenosis by histology. All histological slides were blinded reviewed twice by the same experienced pathologist to eliminate a learning curve. Finally, the CT evaluation was performed by only one radiologist. Thus, interobserver agreement for stenosis grading could not be assessed. 5. Conclusion High-resolution CT reliably depicts advanced stage coronary plaques with an overall good correlation of stenosis degree
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
compared to histology, however, the degree of stenosis is systematically overestimated in calcified and underestimated in non-calcified plaques. Quantification of stenosis degree by CT is only been reliable for mixed plaques. Conflict of interest There is no conflict of interest.
[10]
[11]
[12]
References [13] [1] Taylor AJ, Merz CN, Udelson JE. 34th Bethesda Conference: executive summary—can atherosclerosis imaging techniques improve the detection of patients at risk for ischemic heart disease? Journal of the American College of Cardiology 2003;41(11):1860–2. [2] Ota H, Takase K, Rikimaru H, et al. Quantitative vascular measurements in arterial occlusive disease. Radiographics 2005;25(5):1141–58. [3] Raggi P, Taylor A, Fayad Z, et al. Atherosclerotic plaque imaging: contemporary role in preventive cardiology. Archives of Internal Medicine 2005;165(20):2345–53. [4] Alkadhi H, Scheffel H, Desbiolles L, et al. Dual-source computed tomography coronary angiography: influence of obesity, calcium load, and heart rate on diagnostic accuracy. European Heart Journal 2008;29(6):766–76. [5] Leber AW, Knez A, von Ziegler F, et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrasound. Journal of the American College of Cardiology 2005;46(1):147–54. [6] Harish A, Khatri P, Priyadarshini H, et al. Accuracy of 64-slice coronary CT angiography in predicting percentage diameter stenosis. Indian Heart Journal 2008;60(4):296–301. [7] Caussin C, Larchez C, Ghostine S, et al. Comparison of coronary minimal lumen area quantification by sixty-four-slice computed tomography versus intravascular ultrasound for intermediate stenosis. American Journal of Cardiology 2006;98(7):871–6. [8] Ferencik M, Ropers D, Abbara S, et al. Diagnostic accuracy of image postprocessing methods for the detection of coronary artery stenoses by using multidetector CT. Radiology 2007;243(3):696–702. [9] Okabe T, Weigold WG, Mintz GS, et al. Comparison of intravascular ultrasound to contrast-enhanced 64-slice computed tomography to assess the significance
[14]
[15]
[16]
[17] [18]
[19]
[20]
269
of angiographically ambiguous coronary narrowings. American Journal of Cardiology 2008;102(8):994–1001. Joshi SB, Okabe T, Roswell RO, et al. Accuracy of computed tomographic angiography for stenosis quantification using quantitative coronary angiography or intravascular ultrasound as the gold standard. American Journal of Cardiology 2009;104(8):1047–51. Becker CR, Nikolaou K, Muders M, et al. Ex vivo coronary atherosclerotic plaque characterization with multi-detector-row CT. European Radiology 2003;13(9):2094–8. Halliburton SS, Schoenhagen P, Nair A, et al. Contrast enhancement of coronary atherosclerotic plaque: a high-resolution, multidetector-row computed tomography study of pressure-perfused, human ex vivo coronary arteries. Coronary Artery Disease 2006;17(6):553–60. Chen Y, Han P, Liang B, et al. Comparative study on 16-slice CT coronary angiography vs conventional coronary angiography – a report of 38 cases. Journal of Huazhong University of Science and Technology [Medical Sciences] 2008;28(1):110–3. Stary HC. Natural history and histological classification of atherosclerotic lesions: an update. Arteriosclerosis, Thrombosis, and Vascular Biology 2000;20(5):1177–8. Dragu R, Kerner A, Gruberg L, et al. Angiographically uncertain left main coronary artery narrowings: correlation with multidetector computed tomography and intravascular ultrasound. International Journal of Cardiovascular Imaging 2008;24(5):557–63. Simons DB, Schwartz RS, Edwards WD, Sheedy PF, Breen JF, Rumberger JA. Noninvasive definition of anatomic coronary artery disease by ultrafast computed tomographic scanning: a quantitative pathologic comparison study. Journal of the American College of Cardiology 1992;20(5):1118–26. Achenbach S, Daniel WG. Current role of cardiac computed tomography. Herz 2007;32(2):97–107. Hur J, Kim YJ, Lee HJ, et al. Quantification and characterization of obstructive coronary plaques using 64-slice computed tomography: a comparison with intravascular ultrasound. Journal of Computer Assisted Tomography 2009;33(2):186–92. Dewey M, Zimmermann E, Deissenrieder F, et al. Noninvasive coronary angiography by 320-row computed tomography with lower radiation exposure and maintained diagnostic accuracy. Comparison of results with cardiac catheterization in a head-to-head pilot investigation. Circulation 2009. Schroeder S, Kuettner A, Wojak T, et al. Non-invasive evaluation of atherosclerosis with contrast enhanced 16 slice spiral computed tomography: results of ex vivo investigations. Heart 2004;90(12):1471–5.
Contents lists available at SciVerse ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Quantification of coronary artery stenosis with high-resolution CT in comparison with histopathology in an ex vivo study Matthias Dettmer a,1 , Nicola Glaser-Gallion b,2 , Paul Stolzmann c,3 , Florian Glaser-Gallion b,2 , Juergen Fornaro b,2 , Gudrun Feuchtner d,4 , Wolfram Jochum e,5 , Hatem Alkadhi c,3 , Simon Wildermuth b,6 , Sebastian Leschka b,∗ a
Department of Pathology and Laboratory Medicine, University of Pittsburgh, S-417 BST 200 Lothrop Street, Pittsburgh, PA 15261, USA Institute of Radiology, Kantonsspital St. Gallen, Rorschacherstrasse 95, 9007 St. Gallen, Switzerland c Institute of Diagnostic and Interventional Radiology, University Hospital Zurich, Rämistrasse 100, 8091 Zurich, Switzerland d Department of Radiology II, Medical University Innsbruck, Anichstr. 35, 6020 Innsbruck, Austria e Institute of Pathology, Kantonsspital St. Gallen, Rorschacherstrasse 95, 9007 St. Gallen, Switzerland b
a r t i c l e
i n f o
Article history: Received 25 June 2012 Received in revised form 15 September 2012 Accepted 21 September 2012 Keywords: Coronary artery plaque Atherosclerosis Computed tomography Histology Stenosis Stary classification
a b s t r a c t Purpose: To investigate the ex vivo performance of high-resolution computed tomography (CT) for quantitative assessment of percentage diameter stenosis in coronary arteries compared to histopathology. Materials and methods: High-resolution CT was performed in 26 human heart specimens after the injection of iodinated contrast media into the coronary arteries. Coronary artery plaques were visually identified on CT images and the grade of stenosis for each plaque was measured with electronic calipers. All coronary plaques were characterized by histopathology according to the Stary classification, and the percentage of stenosis was measured. Results: CT depicted 84% (274/326) of all coronary plaques identified by histology. Missed plaques by CT were of Stary type I (n = 31), type II (n = 16), and type III (n = 5). The stenosis degree significantly correlated between CT and histology (r = 0.81, p < 0.001). CT systematically overestimated the stenosis of calcified plaques (mean difference - 11.0 ± 9.5%, p < 0.01) and systematically underestimated the stenosis of noncalcified plaques (mean difference −6.8 ± 10.4%, p < 0.05), while there was no significant difference for mixed-type plaques (mean difference −0.4 ± 11.7%, p = 0.85). There was a significant underestimation of stenosis degree as measured by CT for Stary II plaques (mean difference −14 ± 9%, p < 0.01) and a significant overestimation for Stary VII plaques (mean difference 9 ± 10%, p < 0.05), but there was no significant difference in stenosis degree between both modalities for other plaque types. Conclusions: High-resolution CT reliably depicts advanced stage coronary plaques with an overall good correlation of stenosis degree compared to histology, however, the degree of stenosis is systematically overestimated in calcified and underestimated in non-calcified plaques. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
∗ Corresponding author. Tel.: +41 71 494 2270; fax: +41 71 494 6479. E-mail addresses: [email protected] (M. Dettmer), [email protected] (N. Glaser-Gallion), [email protected] (P. Stolzmann), fl[email protected] (F. Glaser-Gallion), [email protected] (J. Fornaro), [email protected] (G. Feuchtner), [email protected] (W. Jochum), [email protected] (H. Alkadhi), [email protected] (S. Wildermuth), [email protected] (S. Leschka). 1 Tel.: +1 412 864 3351. 2 Tel.: +41 71 494 1111; fax: +41 71 494 6479. 3 Tel.: +41 44 255 1111; fax: +41 44 255 1819. 4 Tel.: +43 512 504 81898; fax: +43 512 504 24029. 5 Tel.: +41 71 494 2102; fax: +41 71 494 2894. 6 Tel.: +41 71 494 2181; fax: +41 71 494 6479. 0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejrad.2012.09.021
Identification of coronary atherosclerotic plaques improves the risk stratification with regard to the occurrence of coronary events [1]. Assessing patients with arterial stenosis with different imaging techniques is important for determining optimal therapy [2]. Traditionally, imaging of coronary artery disease (CAD) was based on catheter coronary angiography. However, catheter angiography is limited by luminographic visualization of the coronary artery and its inability to visualize coronary atherosclerotic plaques [3]. Intravascular ultrasound (IVUS) is considered as the clinical reference modality for most accurate characterization and detection of coronary atherosclerotic plaques, regarding plaque morphology, and regarding quantification of arterial stenosis [3]. However, IVUS is an invasive procedure and thus, its broad application in larger
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
patient populations for risk stratification is limited. It is apparent that a non-invasive technique for the early detection and characterization of the content of coronary atherosclerotic plaques is highly desirable. Cardiac computed tomography (CT) is an emerging technique that allows for the accurate, non-invasive assessment of coronary artery stenosis and coronary plaques [3,4]. Being a cross-sectional imaging modality, CT allows for the evaluation of the arterial wall in addition to the assessment of luminal narrowing [6]. Several studies have quantified the grade of stenosis in coronary arteries with CT in comparison with IVUS or conventional coronary angiography (CCA) [5,6]. Some studies comparing CT with conventional coronary angiography with a good overall correlation between the two modalities for cross-sectional and longitudinal vessel reconstructions have been done with even improved diagnostic accuracy using image postprocessing methods [8]. However, these data are questioned by others who could not find a correlation, and reliability for CT to detect coronary artery stenoses precisely in comparison to angiography [5]. A better correlation between CT and IVUS than between angiography and IVUS has been shown [9]. These data suggest that catheter angiography – albeit being considered as gold standard – has limitations [10]. To overcome the limitations of an imaging technique as a gold standard, we compared the CT measurements of coronary stenosis degree with histopathology. As a matter of fact, some studies have tried to determine the morphology of coronary atherosclerotic plaques in human heart specimens in comparison with histopathology [6,7,12] and in patients in comparison with IVUS [5,7] and angiography [13]. Recent CT studies comparing coronary plaques with histopathology using 4-slice [11] and 16-slice technology [12] have shown promising results. However, the depiction of coronary artery plaques was limited by the relatively low spatial and temporal resolution of the former CT systems used. In addition, the sample sizes in those studies were small, including only up to 50 plaques [11]. One of the most recent innovations in CT technology, i.e., dual-source CT, is characterized by a spatial resolution of 0.4 mm × 0.4 mm × 0.4 mm and a temporal resolution of 83 ms. Thus, the purpose of this study was to investigate the performance of high-resolution CT to depict and quantify coronary artery stenosis in comparison to histopathology in human heart specimens. 2. Materials and methods 2.1. Study objects This study was approved by the local ethics committee. The hearts of 26 patients (8 women, 18 men; mean age 70 ± 12 years; age range 38–85 years) were included. Inclusion criteria were age over 35 years and an autopsy had to be intended. All cause of death was accepted for study inclusion. Cause of death were pulmonary embolism (n = 7), end stage malignancy (n = 6), sepsis (n = 5), acute myocardial infarction (n = 4), acute respiratory distress syndrome (n = 2), ventricular fibrillation (n = 1), and myocardial rupture due to subacute myocardial infarction (n = 1). 2.2. Specimen preparation The autopsies were performed according to standard techniques. After explanation from the corpse, the hearts were washed in water without further fixation or staining, cooled on crushed iced, and immediately transferred to the CT imaging suite. The ostia of the left and the right coronary artery were cannulated with a 5-French balloon catheter (6.0 mm × 2.0 mm) and the coronary arteries were filled under pressure with a mixture of iodinated
265
contrast medium (Ultravist 370, 370 mg/mL, Bayer Schering Pharma, Berlin, Germany) and saline solution to achieve a target attenuation of 300 Hounsfield units (HU) within the coronary lumen. Thereafter, the balloons were blocked to avoid contrast rinsing out of the coronary arteries. 2.3. CT protocol The prepared specimen was placed in the middle of the CT scanner. All examinations were performed with a dual-source CT scanner (Somatom Definition, Siemens Healthcare, Forchheim, Germany). The following parameters were used for imaging: detector collimation 2 mm × 32 mm × 0.6 mm, slice acquisition 2 mm × 64 mm × 0.6 mm, gantry rotation time 330 ms, pitch 0.2, tube potential 120 kV, and tube current time product 330 mAs per rotation. This scan protocol is similar to that commonly used for the in vivo imaging of the coronary arteries [4]. The data was reconstructed with a slice thickness of 0.6 mm, a reconstruction increment of 0.4 mm, and using a soft-tissue convolution kernel (B30) and a sharp kernel (B46). The field of view was set to 205 mm × 205 mm (matrix: 512 × 512 pixels) to achieve an inplane spatial resolution of 0.4 mm × 0.4 mm. 2.4. CT data analysis The presence of stenosis in the coronary arteries was evaluated immediately after the CT scan on the contrast-enhanced CT dataset by one experienced cardiovascular radiologist. The exact position of the plaques was noted by distance measurements from the coronary ostium in order to assure accurate histopathological sectioning. For each plaque reconstructions perpendicular to the vessel centerline were performed and used for data analysis. The radiologist analyzed the perpendicular reconstructions and selected the image demonstrating the highest degree of stenosis of each plaque. In each plaque one free-hand region of interest was circled around the contrast filled lumen and one region of interest around the vessel outline and the area of both circles were noted. Both measurements were performed twice. The mean of the repeated measurements was used to calculate the degree of stenosis by dividing the lumen area with the vessel area. Each plaque was classified as being non-calcified (absence of any calcified deposits), calcified (absence of any non-calcified components), and mixed-type (presence of non-calcified and calcified components). Measurements for all plaque types (e.g., non-calcified, mixed plaques, and calcified plaques) were performed on softtissue and sharp kernel reconstructions. 2.5. Histopathologic analysis After CT scanning, standard pathological sectioning of the heart was performed. All histopathologic analyses were performed by the same experienced pathologist. Transverse sections of the coronary arteries were obtained for all coronary artery plaques detected by CT and also for the remaining part of the coronary tree. Coronary artery stenosis was identified by measuring the distance of the plaques from the coronary ostium. Coronary arteries were fixed in 10% buffered formalin, decalcified if necessary and embedded in paraffin. Two-micron thick sections were mounted on glass slides, deparaffinized, rehydrated and stained with hematoxylin-eosin and elastica-van gieson using standard histological techniques. To quantify stenosis, the percentage of the area of the coronary artery around the elastica interna and the lumen was measured. To quantify stenosis, the area of the coronary artery around the lumen, the lamina elastica interna and around the lamina elastica externa was measured and a relationship was calculated. All plaques were classified according to
266
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
the Stary classification [14] using a lightmicroscope (Zeiss Axiophot, Carl Zeiss, Jena, Germany) and Imageaccess© . Stary type I–III plaques were considered as early stage plaques, and Stary type IV–VIII plaques were considered as advanced stage plaques. Histopathological classification was considered the standard of reference. 2.6. Statistical analysis A p value of <.05 was considered statistically significant for all tests. Continuous variables were expressed as means ± standard deviations and categorical variables as frequencies or percentages. Overall differences in measurement of stenosis degree between CT and histology were compared by using Pearson’s correlation analysis and Bland–Altman scatterplot. The Wilcoxon signed-rank test was used for pair-wise comparisons. The analysis of variance test was used to statistically compare measurement differences between CT and histology in the different Stary type plaques. Statistical significance of differences in stenosis degree measurement between CT and histology were calculated by using a one-sample t test against zero. 3. Results Histological analysis depicted 326 coronary atherosclerotic plaques in the 26 heart specimens: 31 plaques (10%) were classified as Stary type I, 26 plaques (8%) as Stary II, 27 plaques (8%) as Stary III, 29 plaques (9%) as Stary IV, 73 plaques (22%) as Stary V, 22 plaques (7%) as Stary VI, 62 plaques (19%) as Stary VII, and 56 plaques (17%) as Stary VIII. A total of 274 coronary plaques were identified in the CT coronary angiograms. The 52 plaques missed by CT were classified as Stary type I (n = 31/31, 100%), type II (n = 16/26, 62%), and type III (n = 5/27, 17%). Thus, the overall depiction rate of CT for Stary type I-VIII plaques was 84% (274/326). 3.1. Overall accuracy of stenosis degree quantification by CT in comparison to histology
Fig. 1. Linear regression plot of stenosis degree measured by histology (y-axis) against stenosis degree measured by CT (x-axis). Dashed lines represent 95% confidence limits. Linear correlation indicates significant correlation of stenosis degree measurement between CT and histology (Pearson correlation; r = 0.81, p < 0.001). The different Stary type plaques are indicated by markers.
image noise, and measurements in sharp kernel reconstructions were used for calcified plaques to compensate for blooming artifacts. Comparing the stenosis degree measurements between CT and histology, CT significantly underestimated the stenosis degree for non-calcified plaques (p < 0.05) and overestimated the stenosis
For the 274 coronary plaques identified by both CT and histology, the overall stenosis degree measurement significantly correlated between CT and histology (r = 0.81, p < 0.001, Fig. 1). Bland–Altman analysis revealed a minor overestimation of stenosis degree measurement by CT (mean 0.5%) and a wide range of the levels of agreement (−24.3% to 25.3%, Fig. 2). 3.2. Differences in stenosis degree measurement for plaques classified as non-calcified, mixed or calcified by CT Image examples of stenosis degree measurements for noncalcified, mixed, and calcified plaques are demonstrated in Fig. 3. CT classified 29% (80/274) of the coronary artery plaques as noncalcified, 46% (127/274) as mixed, and 25% (67/274) as calcified. The stenosis degree was significantly higher for calcified plaques compared to non-calcified (p < 0.01) and mixed plaques (p < 0.05, Table 1). Mixed plaques had a significantly higher stenosis degree than non-calcified plaques (p < 0.05). There was no significant difference in the stenosis degree between soft tissue and sharp reconstruction kernel for noncalcified (p = 0.76) and mixed plaque (p = 0.51), while the stenosis degree of calcified plaques was significantly lower for sharp compared to soft tissue kernel reconstruction (p < 0.01). The difference of stenosis degree measured by CT and histology for calcified plaques was lower for the sharp kernel reconstruction (mean difference 9 ± 14%) as when using a soft tissue kernel (mean difference 20 ± 19%). Thus, measurements in soft-tissue kernel reconstructions were used for non-calcified and mixed plaques to avoid higher
Fig. 2. Bland–Altman plot of the agreement of stenosis degree measurements between CT and histology. There is a minor overestimation of stenosis degree measurement by CT (mean 0.5%, bold line) with high levels of agreement (−24.3% to 25.3%, dashed lines). The different Stary type plaques are indicated by markers.
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
267
Fig. 3. Measurement of degree of stenosis with histology and CT in non-calcified (A), mixed (B), and calcified plaques (C). In the non-calcified plaque, area stenosis was 23% at histology and 17% at CT. In the mixed plaque, area stenosis was 82% with histology and 79% with CT. In the calcified plaque, area stenosis was 55% with histology. Using the B30 kernel for CT area stenosis was 79% while area stenosis was 62% with the B46 kernel.
degree for calcified plaques (p < 0.01). The stenosis degree measurements for mixed plaques were comparable between CT and histology. 3.3. Differences in stenosis degree measurement between CT and histology for the different Stary type plaques Table 2 summarizes stenosis degree measurements of CT and histology for the different Stary type plaques. The difference in stenosis degree measurements between CT and histology were not significantly different among the different Stary type plaques
(p = 0.16). The stenosis degree as measured by CT revealed a significant underestimation for Stary II plaques (mean difference −14 ± 9%, p < 0.01) and a significant overestimation for Stary VII plaques (mean difference 9 ± 10%, p < 0.05). There was a significantly higher stenosis degree for advanced stage plaques (i.e., Stary IV–VIII) compared to early stage plaques (i.e., Stary II and III) for both CT and histology (each p < 0.05). The stenosis degree as measured by CT revealed a significant underestimation for early stage plaques (mean difference −8 ± 15%, p < 0.05) and a significant overestimation for advanced stage plaques (mean difference 5 ± 12%, p < 0.05).
Table 1 Measurement of stenosis degree by CT and histology for non-calcified, mixed, and calcified plaques. Stenosis degree measured by CT
b
Non-calcified plaques Mixed plaquesb Calcified plaquesb a b
Soft tissue kernel (B30)
Sharp kernel (B46)
52 ± 19% 73 ± 14% 95 ± 10%
51 ± 20% 71 ± 16% 86 ± 12%
Stenosis degree measured by histology
Difference in measurements
p valuea
59 ± 18% 73 ± 14% 75 ± 14%
−6.8 ± 10.4% −0.4 ± 11.7% 11.0 ± 9.5%
<0.05 0.85 <0.01
The p values were calculated using a one-sample t test against zero. Classification in non-calcified, mixed, and calcified according to CT plaques composition.
268
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
Table 2 Measurement of stenosis degree by CT and histology for the different Stary type plaques. Stenosis degree measured by CT b
Stary type II Stary type IIIb Stary type IVb Stary type Vb Stary type VIb Stary type VIIb Stary type VIIIb Early stage Advanced stage a b
24 41 56 70 77 87 75 30 76
± ± ± ± ± ± ± ± ±
7% 14% 18% 14% 14% 13% 12% 14% 19%
Stenosis degree measured by histology 38 45 61 72 76 78 71 38 71
± ± ± ± ± ± ± ± ±
13% 14% 16% 16% 12% 13% 12% 15% 16%
Difference in measurements
p valuea
−13.9 ± 8.5% −3.8 ± 13.9% −4.5 ± 12.1% −2.7 ± 11.4% 1.0 ± 12.2% 9.3 ± 10.1% 4.1 ± 10.5% −8.1 ± 14.6% 4.5 ± 12.1%
<0.01 0.34 0.11 0.52 0.74 <0.05 0.41 <0.05 <0.05
The p values were calculated using a one-sample t test against zero. Histopathologic classification of coronary atherosclerotic plaques were done according to Stary.
4. Discussion 4.1. Overall accuracy of stenosis degree measurement by CT in comparison to histology The reliable detection of coronary artery stenosis by CT is a challenge and the first attempts to quantify the stenosis degree have been made in 2000, at that time with a minor correlation [5]. It could be shown, that coronary computed tomography angiography is able to reliable predict high grade lesions of more than >70% grade of stenosis and that it is possible to measure an excellent correlation of stenosis degree using angiography as a standard of reference [15]. There is also a good correlation between CT in comparison with IVUS [7,15]. Atherosclerosis is a process which is generalized and not limited to a special segment of whatever artery in the human body. This means for the coronary arteries, that they are all affected by the disease, some segments more than others. A main problem for the conventional imaging techniques is the fact, that most of them compare the segment before and after the stenosis of interest with the stenosis itself. It is often forgotten, that these segments already themselves are affected by atherosclerotic changes. These limitations do not exist for histopathology as we could measure the actual grade of stenosis much more precise. CT is rarely compared with histopathology. The first attempt to detect coronary artery stenosis by quantification of calcifications in comparison to histology was in the early nineties, where no correlation between the two modalities could be found [16]. Modern CT systems provide a high spatial resolution of 0.4 mm3 . We could show a significant correlation of the overall stenosis degree for the 274 coronary plaques identified by the two modalities. 4.2. Differences in stenosis degree measurement for plaques classified as non-calcified, mixed or calcified by CT Coronary artery plaques are widely classified by CT in noncalcified, mixed and calcified plaques and the degree of coronary calcifications has a strong predictive value concerning cardiovascular events in the future of patients [17]. The classifications are based on the presence or absence of calcifications, by density measurements of the plaque, or by both methods combined. However, it has to be admitted that the characterization in “vulnerable” and “stable” plaques based on their CT density measurements is limited [18]. A consisting problem while trying to quantify the grade of stenosis is the systematic underestimation of non-calcified plaques and mixed plaques whereas calcified plaques are systematically overestimated in comparison to IVUS [18]. Our results show a better agreement than previous 64-slice CT [7] and are about in the range observed by Dewey et al. [19]. On the other hand, it has to be admitted that CT systematically
overestimates the degree of stenosis in calcified and underestimates it in non-calcified plaque, while CT is accurate in stenosis quantification in mixed plaques. 4.3. Differences in stenosis degree measurement between CT and histology for the different Stary type plaques Up to now, there have no attempts been made to correlate the degree of stenosis, measured by CT with the different Stary-type plaques. The Stary-classification has been correlated with 16-sliceCT, but a classification beyond the three-tired system non-calcified, mixed, calcified was not possible and the grade of stenosis was not assessed [20]. Compared to histology, Becker et al. [11] reported a moderate correlation of the three-tired classification scheme with the histological Stary classification, but a grade of stenosis was also not evaluated. In our study, there were no differences in stenosis degree measurements between CT and histology for the Stary type plaques III–VI and type VIII while we found a significant difference between the two modalities for the early non-calcified lesion Stary II and the heavy calcified lesion Stary VII. Revealed by Bland–Altman, the stenosis degree of the early and the mixed lesions (e.g. Stary I–IV) were more likely to be underestimated in contrast to the advanced lesions (Stary V–VIII) which were more likely to be overestimated. The underestimation of early lesions is due to the limited spatial resolution while it is known that calcification of plaques lead to overestimation of stenosis. 4.4. Study limitations First, no moving or beating of the heart specimens was simulated. Thus, our results represent a “best case scenario” which cannot be fully transferred to an in vivo clinical situation. Even though the CT system used in our study provides the highest currently available temporal resolution, artifacts deriving from cardiac motion causing blurring of the vessel wall may still occur [4]. Second, we did not test the possibility of dual-source CT with acquisition of data with two different energy levels. This technique has shown to allow for an improved differentiation of various tissue components. Third, we observed a partial luminal collapse during tissue fixation on the histological slides. This phenomenon may partially explain an overestimation of stenosis by histology. All histological slides were blinded reviewed twice by the same experienced pathologist to eliminate a learning curve. Finally, the CT evaluation was performed by only one radiologist. Thus, interobserver agreement for stenosis grading could not be assessed. 5. Conclusion High-resolution CT reliably depicts advanced stage coronary plaques with an overall good correlation of stenosis degree
M. Dettmer et al. / European Journal of Radiology 82 (2013) 264–269
compared to histology, however, the degree of stenosis is systematically overestimated in calcified and underestimated in non-calcified plaques. Quantification of stenosis degree by CT is only been reliable for mixed plaques. Conflict of interest There is no conflict of interest.
[10]
[11]
[12]
References [13] [1] Taylor AJ, Merz CN, Udelson JE. 34th Bethesda Conference: executive summary—can atherosclerosis imaging techniques improve the detection of patients at risk for ischemic heart disease? Journal of the American College of Cardiology 2003;41(11):1860–2. [2] Ota H, Takase K, Rikimaru H, et al. Quantitative vascular measurements in arterial occlusive disease. Radiographics 2005;25(5):1141–58. [3] Raggi P, Taylor A, Fayad Z, et al. Atherosclerotic plaque imaging: contemporary role in preventive cardiology. Archives of Internal Medicine 2005;165(20):2345–53. [4] Alkadhi H, Scheffel H, Desbiolles L, et al. Dual-source computed tomography coronary angiography: influence of obesity, calcium load, and heart rate on diagnostic accuracy. European Heart Journal 2008;29(6):766–76. [5] Leber AW, Knez A, von Ziegler F, et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrasound. Journal of the American College of Cardiology 2005;46(1):147–54. [6] Harish A, Khatri P, Priyadarshini H, et al. Accuracy of 64-slice coronary CT angiography in predicting percentage diameter stenosis. Indian Heart Journal 2008;60(4):296–301. [7] Caussin C, Larchez C, Ghostine S, et al. Comparison of coronary minimal lumen area quantification by sixty-four-slice computed tomography versus intravascular ultrasound for intermediate stenosis. American Journal of Cardiology 2006;98(7):871–6. [8] Ferencik M, Ropers D, Abbara S, et al. Diagnostic accuracy of image postprocessing methods for the detection of coronary artery stenoses by using multidetector CT. Radiology 2007;243(3):696–702. [9] Okabe T, Weigold WG, Mintz GS, et al. Comparison of intravascular ultrasound to contrast-enhanced 64-slice computed tomography to assess the significance
[14]
[15]
[16]
[17] [18]
[19]
[20]
269
of angiographically ambiguous coronary narrowings. American Journal of Cardiology 2008;102(8):994–1001. Joshi SB, Okabe T, Roswell RO, et al. Accuracy of computed tomographic angiography for stenosis quantification using quantitative coronary angiography or intravascular ultrasound as the gold standard. American Journal of Cardiology 2009;104(8):1047–51. Becker CR, Nikolaou K, Muders M, et al. Ex vivo coronary atherosclerotic plaque characterization with multi-detector-row CT. European Radiology 2003;13(9):2094–8. Halliburton SS, Schoenhagen P, Nair A, et al. Contrast enhancement of coronary atherosclerotic plaque: a high-resolution, multidetector-row computed tomography study of pressure-perfused, human ex vivo coronary arteries. Coronary Artery Disease 2006;17(6):553–60. Chen Y, Han P, Liang B, et al. Comparative study on 16-slice CT coronary angiography vs conventional coronary angiography – a report of 38 cases. Journal of Huazhong University of Science and Technology [Medical Sciences] 2008;28(1):110–3. Stary HC. Natural history and histological classification of atherosclerotic lesions: an update. Arteriosclerosis, Thrombosis, and Vascular Biology 2000;20(5):1177–8. Dragu R, Kerner A, Gruberg L, et al. Angiographically uncertain left main coronary artery narrowings: correlation with multidetector computed tomography and intravascular ultrasound. International Journal of Cardiovascular Imaging 2008;24(5):557–63. Simons DB, Schwartz RS, Edwards WD, Sheedy PF, Breen JF, Rumberger JA. Noninvasive definition of anatomic coronary artery disease by ultrafast computed tomographic scanning: a quantitative pathologic comparison study. Journal of the American College of Cardiology 1992;20(5):1118–26. Achenbach S, Daniel WG. Current role of cardiac computed tomography. Herz 2007;32(2):97–107. Hur J, Kim YJ, Lee HJ, et al. Quantification and characterization of obstructive coronary plaques using 64-slice computed tomography: a comparison with intravascular ultrasound. Journal of Computer Assisted Tomography 2009;33(2):186–92. Dewey M, Zimmermann E, Deissenrieder F, et al. Noninvasive coronary angiography by 320-row computed tomography with lower radiation exposure and maintained diagnostic accuracy. Comparison of results with cardiac catheterization in a head-to-head pilot investigation. Circulation 2009. Schroeder S, Kuettner A, Wojak T, et al. Non-invasive evaluation of atherosclerosis with contrast enhanced 16 slice spiral computed tomography: results of ex vivo investigations. Heart 2004;90(12):1471–5.