Soft and Intermediate Plaques in Coronary Arteries: How Accurately Can We Measure CT Attenuation Using 64-MDCT?

64-MDCT Measurement of Coronary Artery Plaques Cardiac Imaging Original Research Jun Horiguchi 1 Chikako Fujioka 1 Masao Kiguchi 1 Yun Shen 2 Christian E. Althoff 3,4 Hideya Yamamoto 5 Katsuhide Ito 3 Horiguchi J, Fujioka C, Kiguchi M, et al. Keywords: cardiac CT, coronary artery, plaque DOI:10.2214/AJR.07.2296 Received November 15, 06; accepted after revision May 18, 07. Y. Shen is an employee of GE Healthcare. 1 Department of Clinical Radiology, Hiroshima University Hospital, 1-2-3, Kasumi-cho, Minami-ku, Hiroshima, 734-8551, Japan. Address correspondence to J. Horiguchi. 2 CT Lab of Great China, GE Healthcare, Mongkok Kowloon, Hong Kong. 3 Department of Radiology, Division of Medical Intelligence and Informatics, Programs for Applied Biomedicine, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima, Japan. 4 Present address: Institute of Radiology, Universitaetsmedizin-Charité-Berlin, Berlin, Germany. 5 Department of Molecular and Internal Medicine, Division of Clinical Medical Science, Programs for Applied Biomedicine, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima, Japan. AJR 07; 189:981 988 0361 3X/07/1894 981 American Roentgen Ray Society Soft and Intermediate Plaques in Coronary Arteries: How Accurately Can We Measure CT Attenuation Using 64-MDCT? OBJECTIVE. The objective of this study was to validate the accuracy of 64-MDCT densitometry of soft and intermediate plaques. MATERIALS AND METHODS. Acrylonitrile butadiene styrene resin (47 H) and acrylic (110 H) were used to simulate soft and intermediate plaques, respectively, in coronary artery models (diameters of 3 and 4 mm). The variable parameters were heart rate (50, 65,, and 95 beats per minute), reconstruction algorithm (half and segmentation), coronary artery enhancement (150, 250, 350, and 450 H), CT densitometry site (arterial lumen or center), shape of plaque (D-shaped, centric, and eccentric), and level of stenosis due to plaque (25%, 50%, and 75% of arterial diameter). Measured CT attenuation values of soft and intermediate plaques were compared for different combinations of parameters. Repeated measures analysis of variance, Wilcoxon s signed rank, Mann-Whitney U, and Kruskal-Wallis tests were used for statistical analyses. RESULTS. For measuring soft plaque, CT densitometry was accurate at low heart rates with the use of a half reconstruction algorithm (p < 0.01) on intracoronary artery enhancement of 250 H (p < 0.01). For both soft and intermediate plaques, the densitometry measurements near the arterial lumen were overestimated and higher than those at the center (p < 0.01). For plaques that were 50% or more of the arterial diameter, accurate CT densitometry was possible. CONCLUSION. Coronary artery enhancement has a significant impact on 64-MDCT densitometry measurements of coronary artery plaques, especially of soft plaques. A large plaque size, densitometry performed not near the arterial lumen but at the center of the plaque, intracoronary enhancement of 250 H, and a low heart rate increase the accuracy of plaque densitometry. ost patients with acute coronary M syndromes present with unstable angina, acute myocardial infarction, and sudden coronary death. The most common cause of coronary thrombosis is plaque rupture followed by plaque erosion, whereas calcified nodules are infrequently associated with erosion [1]. The term vulnerable plaque was introduced to define plaques susceptible to such ischemic complications [2]. The detection and characterization of vulnerable plaques remain difficult tasks even using invasive techniques such as intravascular sonography [3], optical coherence tomography [4], plaque thermography [5], and angioscopy [6]. Kopp et al. [7] opened the way to plaque characterization using MDCT by calculating CT attenuation. Schroeder et al. [8], in their analysis of the composition of 34 plaques on 4-MDCT and intracoronary sonography, found that soft plaques had a mean attenuation (± SD) of 14 ± 26 H (range, 42 to 47 H) and intermediate plaques, of 91 ± 21 H (range, 61 112 H). Leber et al. [9], in their analysis of 58 plaques using 16-MDCT, showed that soft plaques had a mean attenuation of 49 ± 22 H (range, 14 82 H) and intermediate plaques, of 91 ± 22 H (range, 34 125 H), although the values for soft and intermediate plaques overlapped. In the most recent study to date of 16- MDCT involving 252 plaques, Pohle et al. [10] found that the attenuation values of soft (58 ± 43 H) and intermediate (121 ± 34 H) plaques were statistically different; however, there was substantial overlap. They therefore concluded that the differentiation of vulnerable from stable plaques based on their CT attenuation is doubtful. On the other hand, the results of some ex vivo studies indicate that plaque differentiation may be possible [11 13]. After their analysis of the composition of 34 plaques on 4-MDCT and intracoronary sonography [8], Schroeder et al. [13] suggested the following MDCT attenuation criteria for differentiation of plaques: H, predominantly AJR:189, October 07 981

1 Fig. 1 Cardiac phantom (ALPHA 2, Fuyo Corporation). Photograph shows balloon and support components of cardiac phantom. Balloon attached with coronary artery models was surrounded by water. lipid-rich plaques; 61 119 H, intermediate plaques; and 1 H, predominantly calcific plaques. Using static cardiac phantoms for their study, Schroeder et al. [14] showed that plaque densitometry was accurate on thinslice images (1 vs 2.5 mm), which are highly dependent on surrounding contrast enhancement. Cademartiri et al. [15], showing that CT attenuation of plaque significantly varied (22 ± 22, 50 ± 26, 107 ± 36, 152 ± 67 H) according to luminal enhancement (35 ± 10, 91 ± 7, 246 ± 18, 511 ± 89 H, respectively), concluded that it is difficult to identify absolute ranges of attenuation that relate to specific plaque characteristics. We thought about the need for an experiment using 64-MDCT with a thin collimation and improved temporal and spatial resolution, as pointed out by Nikolaou et al. [16]. For such an experiment, the CT attenuations of plaque models should mimic those of real plaques measured on cadavers arteries [11, 12], and the range of coronary enhancement should simulate that seen in clinical situations. In addition, most important but never reported, to our knowledge, a pulsating cardiac phantom with variable heart rates was validated with different temporal resolutions. The purpose of this study was to validate the accuracy of 64-MDCT densitometry of soft and intermediate plaques using coronary artery plaque models on a pulsating cardiac phantom. Heart Rate (bpm) Materials and Methods Intravascular Enhancement of Clinical Studies: Coronary CT Angiography on 64-MDCT This part of the study was to investigate patients Hounsfield attenuation values for blood enhancement in clinical 64-MDCT coronary angiography and to adjust the attenuation values for the coronary artery fluid used for the phantom study. This study was approved by our institutional review committee. After seven patients with heavy calcifications or a stent or stents in the main coronary branches had been excluded, 30 consecutive patients (19 men and 11 women; mean age ± SD, 64 ± 11 years; age range, 42 85 years) undergoing cardiac MDCT investigation were enrolled. Both this clinical study and the following phantom experiments of retrospective ECG-gated 64-MDCT coronary angiography were performed using the same unit (LightSpeed VCT, GE Healthcare) with the same scanning parameters except tube current and pitch. No β-blockers were administered in the series. After a test injection of 15 ml of nonionic contrast medium (iopamidol [Iopamiron 370, Schering]) was administered, a timing bolus scan was obtained with contrast medium (amount of contrast medium = 0.7 body weight) at an injection time of 10 seconds, followed by a 25-mL saline chaser administered at the same rate. The scan covered the entire volume of the heart during a single breath-hold, consisting of 5 or 6 heartbeats, with simultaneous recording of the ECG trace. Detector collimation was 64 0.625 mm, gantry rotation speed was 350 milliseconds per rotation, and tube voltage was 1 kv at a tube current of 550 750 mas (depending on patient size, with a dose-reduction model using the ECG modulation technique). 0 Cardiac Cycle 50 bpm 65 bpm bpm 95 bpm 50 bpm with shifting 65 bpm with shifting bpm with shifting 95 bpm with shifting 1 2 3 4 5 Fig. 2 Graph shows eight heart rate sequences that were programmed in phantom: sequence 1, 50 beats per minute (bpm); sequence 2, 65 bpm; sequence 3, bpm; sequence 4, 95 bpm; sequence 5, 50 bpm with shifting; sequence 6, 65 bpm with shifting; sequence 7, bpm with shifting; and sequence 8, 95 bpm with shifting. CT pitch factors ranged from 0.18 to 0.24 by heart rate according to the manufacturer s recommendations for a coronary CT angiography protocol. For heart rates < 75 beats per minute (bpm), a half-scan algorithm was applied; for heart rates 75 bpm, a two-segment reconstruction algorithm, offering improved temporal resolution, was used. For image reconstruction, cardiac phase imaging with the center of the temporal window corresponding to 65% of the R-R interval was used. Other cardiac phase images were reconstructed in cases in which the image quality of the phase images was better than that at 65% of the R-R interval. CT attenuation values in the regions of interest (ROIs) were 316 ± 67 H (range, 167 415 H), 309 ± 65 H (range, 173 429 H), 297 ± 66 H (range, 173 4 H), and 310 ± H (range, 168 418 H) on the proximal left main, left anterior descending, left circumflex, and right coronary arteries, respectively. Thereafter, for coronary artery enhancement models, four concentrations of contrast medium and water with CT attenuation values of 150, 250, 350, and 450 H were prepared. Pulsating Cardiac Phantom A prototype cardiac phantom was used (ALPHA 2, Fuyo Corporation). The phantom consists of five components: driver, control, support, rubber balloon, and ECG (Fig. 1). A controller with an ECG-synchronizer drives the balloon. The construction of the phantom is described in detail elsewhere [17, 18]. The main characteristic features of this phantom are programmable variable heart rate sequences and programmable heart movements that mimic natural heart movements. In this study, eight heart rate sequences were programmed (Fig. 2). 982 AJR:189, October 07

64-MDCT Measurement of Coronary Artery Plaques 45 mm 45 mm 45 mm 4/3 mm 4/3 mm 4 mm 15 mm 75% 50% 25% 75% 50% 25% 75% 50% 25% 3 mm 2 mm 1 mm Fig. 3 Coronary artery plaque models. This drawing shows plaque size, plaque shape, and stenosis levels of coronary artery plaque models. Seven plaques were prepared with different combinations of plaque shape, plaque CT attenuation, and coronary artery diameter, respectively: D-shaped, H, 4 mm; D-shaped, H, 3 mm; centric, H, 4 mm; centric, H, 3 mm; eccentric, H, 4 mm; D-shaped, H, 4 mm; and centric, H, 4 mm. Coronary Artery Plaque Models Acrylonitrile butadiene styrene resin (47 H) and acrylic (110 H) were used to represent soft and intermediate plaques, respectively. In addition, three plaque shapes (D-shaped, centric, and eccentric) and two coronary diameters (3 and 4 mm) were prepared. Using different combinations of plaque shape and artery diameter, the manufacturer of the phantom made seven plaque models. All plaque models had three levels of stenosis (Fig. 3). These seven plaque models were attached to the balloon phantom (mimicking the heart), with the long axis of the model corresponding to the z-axis, and were surrounded by water (Fig. 4). Cardiac Phantom Scan Volumetric data of the phantom were obtained. The tube current used was 650 ma. The pitch was set to 0.175 to allow improved temporal resolution by two- and four-segment reconstruction algorithms. Other scanning parameters were the same as those used for the clinical study, as described earlier. A half-scan algorithm provided a temporal resolution of 175 milliseconds; a twosegment, 97 137 milliseconds; and a four-segment, 52 95 milliseconds. The temporal resolution of a two- or four-segment reconstruction algorithm varied on sequences performed with heart rate shifting. Theme 1: Heart Rate and Temporal Resolution for the Imaging of Coronary Artery Plaque Degradation of image quality caused by motion artifacts decreased the detection of coronary artery plaque and the accuracy of CT densitometry. Image quality at a static state and image quality at eight heart rates were compared. A two-segment reconstruction algorithm was applied for heart rates of 65 bpm and a four-segment algorithm was used for heart rates of bpm. The segmentation algorithm was not used at static state or at 50 bpm because of its lack of effectiveness in improving temporal resolution. For intracoronary artery enhancement, 250 and 350 H, which are considered representative levels judging from our clinical study results, were chosen. Details of the settings are summarized in Appendix 1. The plaque used for this theme was a D-shaped model (semicircular shape) with 50% stenosis. Image quality was subjectively evaluated by three reviewers (experience interpreting cardiac CT: 7, 3, and 3 years) who were unaware of the kinds of plaque models and who graded image quality using a 3-point scale: 2, no or minor motion artifacts, recognized as semicircular shape; 1, mild motion artifacts, however recognized as almost semicircular shape; 0, significant motion artifacts, not recognized as semicircular shape. When the grades assigned for image quality differed among reviewers, the consensus of two reviewers was defined as the grade. Images were displayed with fixed window settings (window width, 700 H; window level, 0 H). Theme 2: Does CT Attenuation of Coronary Plaque Change by Coronary Artery Enhancement? CT attenuation values of soft and intermediate plaques were compared at four different intracoronary artery enhancement values (150, 250, 350, and 450 H). ROIs that were 1 mm 2 were set at the center of the plaques (Fig. 5A). Static and low-heart-rate sequences were chosen to avoid or minimize the effect of motion artifacts. Details of the settings are summarized in Appendix 2. CT densitometry was performed in five slices per plaque in each circumstance by one reviewer (3 years of experience). Theme 3: Does Placing the ROI near the Arterial Lumen Affect CT Attenuation? ROIs were placed near the arterial lumen and at the center of the plaque (Fig. 5B), and the CT attenuation values of soft and intermediate plaques were compared between measurement sites. Two different CT attenuation values of intracoronary artery enhancement (250 and 350 H), which were representative in our clinical study, were used. Details of the settings are summarized in Appendix 3. Fig. 4 Balloon phantom. Seven plaque models were attached to balloon phantom (mimicking heart) and were surrounded by water. Balloon was filled with mixture of water and contrast medium (CT attenuation = H). Theme 4: Is CT Attenuation of Coronary Plaque Affected by the Level of Arterial Stenosis? CT attenuation values of soft and intermediate plaques were compared at three levels of stenosis: 25%, 50%, and 75% of the arterial diameter (Fig. 5C). Details of the settings are summarized in Appendix 4. Theme 5: Is CT Attenuation of Coronary Plaque Affected by the Coronary Artery Diameter or the Stenosis Shape? CT attenuation values of soft plaque were compared for the following five combinations of plaque shape and coronary artery diameter: D-shaped and 3 mm, D-shaped and 4 mm, centric and 3 mm, centric and 4 mm, and eccentric and 4 mm. The stenotic ratio was 50% in area. ROIs were set at the center of the plaques (Fig. 5D). Details of the settings are summarized in Appendix 5. Statistical Analysis All statistical analyses were performed using a commercially available software package (Statcel2, OMS Publishing). For statistical analyses, Mann-Whitney U, repeated measures analysisof-variance, Wilcoxon s signed rank, and Kruskal- Wallis tests were used to determine differences. When statistical significance was observed by repeated measures analysis of variance, the results were made post hoc using the Scheffé test for multiple pairwise comparisons. A p value of < 0.05 was considered to identify significant differences. Results Theme 1 Subjective evaluation of image quality of plaques for different combinations of heart AJR:189, October 07 983

A Fig. 5 Region-of-interest (ROI) setting in coronary artery plaque. Drawing shows typical ROI placement (dotted lines) in coronary artery plaques (gray). A D, All ROIs were ovoid and 1 mm 2. ROIs were set at center of plaques in theme 2 (A), near arterial lumen and at center of plaque in theme 3 (B), at center of plaques in three stenotic levels in theme 4 (C), and at center of plaques in various types of plaque shape and coronary artery diameters in theme 5 (D). TABLE 1: Subjective Evaluation of Image Quality with Different Combinations of Heart Rate Sequences and Reconstruction Algorithms Heart Rate (bpm) Reconstruction Algorithm Temporal Resolution (ms) Soft Plaque in Coronary Artery with an Enhancement of Image Quality Grade a Intermediate Plaque in Coronary Artery with an Enhancement of 250 H 350 H 250 H 350 H Static Half 175 2 2 1 2 50 Half 175 2 2 1 2 65 Half 175 1 2 1 2 Two-segment 97 2 2 1 2 Half 175 1 1 1 1 Two-segment 137 1 1 1 1 Four-segment 2 2 1 1 95 Half 175 1 1 0 1 Two-segment 122 1 1 1 1 Four-segment 52 1 2 1 1 50 with shifting b Half 175 2 2 1 2 65 with shifting Half 175 1 2 1 2 Two-segment 1 2 1 2 with shifting Half 175 1 1 0 1 Two-segment 1 2 0 2 Four-segment 1 2 0 1 95 with shifting Half 175 1 1 0 1 Two-segment 1 1 0 1 Four-segment 1 1 0 1 a Grade 2 = no or minor motion artifacts, recognized as semicircular shape; 1 = mild motion artifacts, however recognized as almost semicircular shape; 0 = significant motion artifacts, not recognized as semicircular shape. b 50 bpm with shifting = heart rate shifts 50, 47, 50, and 53 bpm in a cycle and returns to 50 bpm in the next cycle. rate sequence and reconstruction algorithm is summarized in Table 1. Assessable image quality of soft plaque was obtained with all combinations of heart rates and reconstruction algorithms tested, whereas images of intermediate plaque were of poor quality on a B coronary artery enhancement level of 250 H at high-heart-rate sequences. The image quality of intermediate plaques at 350 H was better than that at 250 H (Mann-Whitney U test, p < 0.01). When reconstruction algorithms were compared, the segmentation algorithm with improved temporal resolution tended to yield better image quality. Theme 2 CT attenuation values of soft and intermediate plaques on four levels of intracoronary artery enhancement for four combinations of heart rate and reconstruction algorithm are shown in Figure 6. CT attenuation values of soft plaque were overestimated on intracoronary artery enhancement levels of 350 and 450 H (p < 0.01). The CT densitometry measurements were accurate and lower for the static model and at 50 bpm (p < 0.01). Intermediate plaques were not detectable on intracoronary artery enhancement of 150 H. Theme 3 Plaque CT attenuation values near the arterial lumen and at the center of plaque are shown in Figure 7. In all combinations of plaque (soft and intermediate) and intracoronary enhancement (250 and 350 H), plaque CT attenuation values near the lumen were overestimated and higher than those at the center (p <0.01). Theme 4 Plaque CT attenuation values in three stenotic levels are shown in Figure 8. For all combinations of plaque (soft and intermediate) and intracoronary artery enhancement D C 984 AJR:189, October 07

64-MDCT Measurement of Coronary Artery Plaques 125 70 50 30 Heart rate static / half reconstruction algorithm Heart rate 50 bpm / half reconstruction algorithm Heart rate 65 bpm / half reconstruction algorithm Heart rate 65 bpm / two-segment reconstruction algorithm 150 250 350 450 Intracoronary Artery Enhancement (H) A Fig. 6 Relationship of CT attenuation values between plaque and intracoronary enhancement. A and B, Graphs show results for soft (A) and intermediate (B) plaques. Repeated measures analysis of variance revealed that CT attenuation values of soft plaque were different among intracoronary artery enhancement levels (p < 0.01) and combinations of heart rates and reconstruction algorithms (p < 0.01). Scheffé test for multiple pairwise comparisons revealed that CT attenuation values were significantly different between intracoronary enhancement of 150 and 350 H (p < 0.01), 150 and 450 H (p < 0.01), and 250 and 450 H (p < 0.01) on static cardiac phantom using half reconstruction and between intracoronary enhancement of 150 and 350 H (p < 0.01), 150 and 450 H (p < 0.01), 250 and 350 H (p < 0.01), and 250 and 450 H (p < 0.01) on cardiac phantom at 50 beats per minute (bpm) using half reconstruction algorithm. Scheffé test revealed that CT attenuation values of plaque on intracoronary artery enhancement of 150 H were significantly different between static cardiac phantom with half reconstruction algorithm and 65 bpm with half reconstruction (p < 0.01) and between 50 bpm with half reconstruction and 65 bpm with half reconstruction (p < 0.01). Scheffé test also revealed that CT attenuation values of plaque on intracoronary artery enhancement of 250 H were significantly different between static phantom with half reconstruction and 65 bpm with half reconstruction (p < 0.01) and between 50 bpm with half reconstruction and 65 bpm with half reconstruction (p < 0.01). In contrast, CT attenuation values of intermediate plaque were not statistically different based on intracoronary artery enhancement level (p = 0.09) or combinations of heart rate and reconstruction algorithm (p =0.10). (250 and 350 H), CT attenuation values with 25% stenosis were overestimated and higher than those with 50% or 75% stenosis (p < 0.01). Theme 5 Plaque CT attenuation values in different combinations of shape and coronary artery diameter are shown in Figure 9. Plaque CT attenuation values were different between the combinations (p < 0.01). D-shaped plaque in a coronary artery with a diameter of 4 mm showed lower CT attenuation values than the other combinations and was close to the real CT attenuation. Discussion Although the identification of calcified plaque is straightforward because of its higher CT attenuation, differentiation of noncalcified plaque from calcified plaque is challenging. A lipid core larger than 1 mm 2 or a lipid-core-to-plaque ratio of greater than % and a fibrous cap thinner than 0.7 mm have been shown to correlate well with plaque rupture [19]. The usual location of a lipid core within an atherosclerotic plaque is just below the thin cap. 1 115 110 105 Factors and Circumstances Necessary for Accurate CT Densitometry of Coronary Plaque There is a consensus that a low heart rate is an advantage for cardiac CT examination and that multisegment reconstruction, with its improved temporal resolution, is effective in high heart rates, especially when stable. In the current study, the two- or four-segment reconstruction was sometimes helpful for better delineation of the plaques. However, because of the finding that soft plaque showed higher CT attenuation values on half- and two-segment reconstructions at 65 bpm, plaque CT densitometry seems more accurate and can be performed better on a patient with a lower heart rate. The results of the current study suggest that coronary artery enhancement has a significant impact on the CT attenuation of plaque, especially that near the coronary artery lumen that is, just below the thin cap. This is considered due to partial volume averaging, which is theoretically dominant for low-attenuation plaque surrounded by the high-attenuation coronary artery lumen. The results suggest that it is difficult to predict the presence of soft plaque on the basis of a CT attenuation measured in an ROI set near the coronary lumen. High coronary artery enhancement (350 H) is advantageous for 95 Heart rate static / half reconstruction algorithm Heart rate 50 bpm / half reconstruction algorithm Heart rate 65 bpm / half reconstruction algorithm Heart rate 65 bpm / two-segment reconstruction algorithm 250 350 450 Intracoronary Artery Enhancement (H) better delineation of both soft and intermediate plaques. However, as shown in this study, soft plaque (50% stenosis) was detected on all heart rate sequences on coronary artery enhancement of 250 H. This finding suggests that this level of enhancement might be better for characterization of soft plaque, as suggested by Schroeder et al. [13, 14]. From the results of themes 4 and 5, it seems that a minimal plaque size for example, resulting in luminal stenosis of 50% or more is needed to precisely measure the attenuation of plaque, especially of soft plaque. An appropriate size for an ROI remains unresolved. Because CT densitometry of D-shaped soft plaque causing 50% stenosis within a 3-mm coronary artery failed in yielding accurate measures, it follows that an ROI should be situated at the center of plaque to avoid the influence of coronary artery enhancement. Technical matters for enhancement of the left ventricle remain to be mentioned. We filled a mixture of water and contrast medium ( H) into a balloon (mimicking the heart) instead of filling a mixture with the same enhancement as the coronary arteries. In actual coronary CT angiography, the left ventricle is enhanced to the same level as the coronary ar- B AJR:189, October 07 985

Fig. 7 Differences in CT attenuation values between region-of-interest placement inside plaque. CT attenuation values of 250 and 350 H were chosen as representative levels from our clinical study results. Wilcoxon s signed rank test revealed that CT attenuation values of plaque near lumen (dark gray) were overestimated and higher than those at center (light gray): soft plaque and intracoronary enhancement of 250 H, soft plaque and 350 H, intermediate plaque and 250 H, and intermediate plaque and 350 H (p <0.01). Fig. 8 Relationship between plaque CT attenuation values and stenosis level. CT attenuation values of plaque between three stenotic levels that is, 25% (dark gray), 50% (light gray), and 75% (black) (in diameter). Kruskal-Wallis test revealed that CT attenuation values of plaque were different between stenosis levels: soft plaque and intracoronary enhancement of 250 H, soft plaque and 350 H, intermediate plaque and 250 H, and intermediate plaque and 350 H (p < 0.01). Fig. 9 Plaque CT attenuation values in combinations of plaque shape and coronary artery diameter. Kruskal- Wallis test revealed that CT attenuation values of plaque were statistically different between combinations of soft plaque and coronary artery enhancement of 250 H (gray) and soft plaque and coronary artery enhancement of 350 H (black) (p <0.01). 1 1 1 1 1 1 1 teries, whereas the enhancement of the right ventricle and atrium has already decreased. The wall of the left ventricle does not enhance 1 0 Soft and 250 H Soft and 350 H Intermediate and 250 H Intermediate and 350 H Plaque and Intracoronary Artery Enhancement 0 Soft and 250 H Soft and 350 H Intermediate and 250 H 0 Intermediate and 350 H Plaque and Intracoronary Artery Enhancement D-shaped and 3 mm D-shaped and 4 mm Centric and 3 mm Centric and 4 mm Plaque Shape and Coronary Artery Diameter (mm) Eccentric and 4 mm much at the timing of coronary CT angiography. In addition to these technical matters, the epicardial coronary arteries are some distance from the heart. We therefore think that the CT attenuation of coronary plaque is not influenced much by the enhancement of the heart except in two instances: The plaque in the coronary orifice is influenced by the enhancement of the aorta, and the plaque in the myocardial bridge is influenced by the enhancement of the left ventricle. Study Limitations This study has several limitations. Atherosclerotic plaques have a complex composition that is, lipid-rich, fibrous, and calcified areas coexist and are often intermixed. We prepared only one CT attenuation as a model for each of the soft (47 H) and the intermediate (110 H) plaques, although the previously reported CT attenuation values of the plaques had various ranges. We used plaques with a uniform CT attenuation because the purpose of this study was to investigate the effect of coronary artery enhancement on the CT attenuation of plaque. In addition, the shape of the coronary artery was always round; therefore, positive remodeling was not simulated. Next, we changed coronary artery enhancement levels on static and low-heart-rate sequences and did not investigate coronary artery enhancement on high-heart-rate sequences. This omission needs to be mentioned if the results are translated into an in vivo situation. Finally, we did not simulate large variations of heart rate, such as arrhythmia or premature heart beat, or changes in body posture. In conclusion, coronary artery enhancement has a significant impact on 64-MDCT densitometry of coronary artery plaque, especially of soft plaque. Large plaque size, densitometry performed not near the lumen but at the center of the plaque, intracoronary artery enhancement of 250 H, and low heart rates increase the accuracy of plaque densitometry. References 1. Virmani R, Burke AP, Farb A, Kolodgie FD. Pathology of the vulnerable plaque. J Am Coll Cardiol 06; 47[suppl 8]:C13 C18 2. Naghavi M, Libby P, Falk E, et al. From vulnerable plaque to vulnerable patient: a call for new definitions and risk assessment strategies: part I. Circulation 03; 108:1664 1672 3. Nissen SE, Yock P. Intravascular ultrasound: novel pathophysiological insights and current clinical applications. Circulation 01; 103:4 616 4. Patwari P, Weissman NJ, Boppart SA, et al. Assessment of coronary plaque with optical coherence tomography and high-frequency ultrasound. Am J 986 AJR:189, October 07

64-MDCT Measurement of Coronary Artery Plaques Cardiol 00; 85:641 644 5. Stefanadis C, Diamantopoulos L, Vlachopoulos C, et al. Thermal heterogeneity within human atherosclerotic coronary arteries detected in vivo: a new method of detection by application of a special thermography catheter. Circulation 1999; 99:1965 1971 6. Takano M, Mizuno K, Okamatsu K, Yokoyama S, Ohba T, Sakai S. Mechanical and structural characteristics of vulnerable plaques: analysis by coronary angioscopy and intravascular ultrasound. J Am Coll Cardiol 01; 38:99 104 7. Kopp AF, Schroeder S, Baumbach A, et al. Non-invasive characterisation of coronary lesion morphology and composition by multislice CT: first results in comparison with intracoronary ultrasound. Eur Radiol 01; 11:17 1611 8. Schroeder S, Kopp AF, Baumbach A, et al. Noninvasive detection and evaluation of atherosclerotic coronary plaques with multislice computed tomography. J Am Coll Cardiol 01; 37:1430 1435 9. Leber AW, Knez A, Becker A, et al. Accuracy of multidetector spiral computed tomography in identifying and differentiating the composition of coronary atherosclerotic plaques: a comparative study with intracoronary ultrasound. J Am Coll Cardiol 04; 43:1241 1247 10. Pohle K, Achenbach S, MacNeill B, et al. Characterization of non-calcified coronary atherosclerotic plaque by multi-detector row CT: comparison to IVUS. Atherosclerosis 07; 190:174 1 11. Estes JM, Quist WC, Lo Gerfo FW, et al. Noninvasive characterization of plaque morphology using helical computed tomography. J Cardiovasc Surg (Torino) 1998; 39:527 534 12. Becker CR, Nikolaou K, Muders M, et al. Ex vivo coronary atherosclerotic plaque characterization with multi-detector-row CT. Eur Radiol 03; 13:94 98 13. Schroeder S, Kuettner A, Leitritz M, et al. Reliability of differentiating human coronary plaque morphology using contrast-enhanced multislice spiral computed tomography: a comparison with histology. J Comput Assist Tomogr 04; 28:449 454 14. Schroeder S, Flohr T, Kopp AF, et al. Accuracy of density measurements within plaques located in artificial coronary arteries by X-ray multislice CT: results of a phantom study. J Comput Assist Tomogr 01; 25:900 906 15. Cademartiri F, Mollet NR, Runza G, et al. Influence of intracoronary attenuation on coronary plaque measurements using multislice computed tomography: observations in an ex vivo model of coronary computed tomography angiography. Eur Radiol 05; 15:1426 1431 16. Nikolaou K, Flohr T, Knez A, et al. Advances in cardiac CT imaging: 64-slice scanner. Int J Cardiovasc Imaging 04; :535 5 17. Horiguchi J, Shen Y, Akiyama Y, et al. Electron beam CT versus 16-MDCT on the variability of repeated coronary artery calcium measurements in a variable heart rate phantom. AJR 05; 185:995 0 18. Horiguchi J, Shen Y, Akiyama Y, et al. Electron beam CT versus 16-slice spiral CT: how accurately can we measure coronary artery calcium volume? Eur Radiol 06; 16:374 3 19. Ge J, Chirillo F, Schwedtmann J, et al. Screening of ruptured plaques in patients with coronary artery disease by intravascular ultrasound. Heart 1999; 81:621 627 APPENDIX 1: Design for Theme 1: Heart Rate and Temporal Resolution for the Imaging of Coronary Artery Plaque Coronary artery enhancement values: 250 and 350 H Heart rates: static, 50 bpm, 65 bpm, bpm, 95 bpm, 50 bpm with shifting, 65 bpm with shifting, bpm with shifting, and 95 bpm with shifting Reconstruction algorithm: half for all heart rate sequences, two-segment for 65 bpm or greater, and four-segment for bpm or greater No. of image data sets: 38 No. of plaques: 2 (soft and intermediate) D-shaped with 50% stenosis No. of grade assigned for subjective evaluation of image quality per plaque (from many slices): 1 No. of the assignment for subjective evaluation in the theme: 76 APPENDIX 2: Design for Theme 2: Does CT Attenuation of Coronary Plaque Change by Coronary Artery Enhancement? Coronary artery enhancement values: 150, 250, 350, and 450 H Heart rates: static, 50 bpm, and 65 bpm Reconstruction algorithm: half for static, 50 bpm, and 65 bpm; and two-segment for 65 bpm No. of image data sets: 16 No. of plaques: 2 (soft and intermediate) D-shaped with 50% stenosis Diameter of coronary artery: 4 mm No. of slices for CT densitometry per plaque: 5 Region-of-interest setting: at the center of the plaque, 1 mm 2 APPENDIX 3: Design for Theme 3: Does Placing the Region of Interest (ROI) near the Arterial Lumen Affect CT Attenuation? Coronary artery enhancement values: 250 and 350 H Heart rates: static, 50 bpm, and 65 bpm Reconstruction algorithm: half for static, 50 bpm, and 65 bpm; and two-segment for 65 bpm No. of image data sets: 8 No. of plaques: 2 (soft and intermediate) D-shaped with 50% stenosis Diameter of coronary artery: 4 mm No. of slices for CT densitometry per plaque: 5 ROI setting in plaque: near the arterial lumen and at center of the plaque, 1 mm 2 Appendixes continue on next page AJR:189, October 07 987

APPENDIX 4: Design for Theme 4: Is CT Attenuation of Coronary Plaque Affected by the Level of Arterial Stenosis? Coronary artery enhancement values: 250 and 350 H Heart rates: static, 50 bpm, and 65 bpm Reconstruction algorithm: half for static, 50 bpm, and 65 bpm; and two-segment for 65 bpm No. of image data sets: 8 No. of plaques: 6 (two CT attenuation values [soft and intermediate] 3 stenotic ratios [25%, 50%, and 75% of arterial diameter; 18%, 50%, and 82% in area, respectively]) Plaque shape: D-shaped Diameter of coronary artery: 4 mm No. of slices for CT densitometry per plaque: 5 Region-of-interest setting: at center of the plaque, 1 mm 2 APPENDIX 5: Design for Theme 5: Is CT Attenuation of Coronary Plaque Affected by the Coronary Artery Diameter or the Stenosis Shape? Coronary artery enhancement values: 250 and 350 H Heart rates: static, 50 bpm, and 65 bpm Reconstruction algorithm: half for static, 50 bpm, and 65 bpm; and two-segment for 65 bpm No. of image data sets: 8 CT attenuation of plaque: soft plaque = 47 H No. of different combinations of coronary arterial diameter and plaque shape: 5 (D-shaped plaque in 3-mm-diameter coronary artery, D-shaped and 4 mm, centric and 3 mm, centric and 4 mm, and eccentric and 4 mm) Stenotic ratio of the plaque: 50% in area No. of slices for CT densitometry per plaque: 5 Region-of-interest setting: at center of the plaque, 1 mm 2 988 AJR:189, October 07