Author + information
- Received March 10, 2015
- Revision received June 3, 2015
- Accepted June 26, 2015
- Published online October 1, 2015.
- Kenji Matsumoto, MD, PhD∗,
- Shoichi Ehara, MD, PhD∗∗ (, )
- Takao Hasegawa, MD, PhD∗,
- Mikumo Sakaguchi, MD∗,
- Kenichiro Otsuka, MD, PhD∗,
- Junichi Yoshikawa, MD, PhD† and
- Kenei Shimada, MD, PhD∗
- ∗Department of Cardiovascular Medicine, Osaka City University Graduate School of Medicine, Osaka, Japan
- †Nishinomiya Watanabe Cardiovascular Center, Hyogo, Japan
- ↵∗Reprint requests and correspondence:
Dr. Shoichi Ehara, Department of Cardiovascular Medicine, Osaka City University Graduate School of Medicine, 1-4-3 Asahi-machi, Abeno-ku, Osaka 545-8585, Japan.
Objectives This study sought to investigate the relationship between localization of high-intensity signals (HISs) on T1-weighted imaging (T1WI) with the noncontrast magnetic resonance technique and plaque morphology detected on optical coherence tomography, and the clinical severity of angina pectoris.
Background Since the introduction of the T1WI noncontrast magnetic resonance technique for plaque imaging, some groups have reported that HISs in the coronary artery on T1WI are associated with a vulnerable morphology and future cardiac events. However, the association between the localization of HISs, such as coronary intrawall or intraluminal, and plaque morphology has not been investigated.
Methods One hundred lesions with either stable or unstable angina were included and divided into 3 groups according to the following criteria using T1WI. First, the plaques with the ratio between the signal intensities of coronary plaque and cardiac muscle ≤1.0 were classified as non-HISs (n = 39). Then, HISs with the ratio between the signal intensities of coronary plaque and cardiac muscle >1.0 were classified into 2 types by using cross-sectional T1WI. Those localized within the coronary wall when the lumen was identified were defined as intrawall HISs (n = 37), whereas those occupying the lumen when the lumen was not, or even if only partly, identified, were defined as intraluminal HISs (n = 24).
Results Multivariate analysis revealed that intrawall HISs were associated with macrophage accumulation and the absence of calcification assessed by using optical coherence tomography. In contrast, thrombus and intimal vasculature were independent factors associated with intraluminal HISs. Furthermore, 50% of patients with intraluminal HISs experienced rest angina, such as Braunwald class II or III.
Conclusions This study shows that intrawall and intraluminal HISs on T1WI in patients with angina are related to the different types of vulnerable plaque morphology and the clinical severity.
- angina pectoris
- intraplaque hemorrhage
- magnetic resonance imaging
- optical coherence tomography
Plaque rupture or erosion of the endothelial surface with subsequent thrombus formation is recognized as the most important mechanism in acute coronary syndromes (1). Moreover, histopathologic reports have demonstrated that coronary intraplaque hemorrhage could lead to an increase in plaque burden, in addition to being a reflection of the biological activity of the lesion (2–4). However, the association between intraplaque hemorrhage and the onset of clinical symptoms resulting from coronary artery disease is still a controversial issue.
Since the introduction of the T1-weighted imaging (T1WI) noncontrast magnetic resonance (MR) technique for plaque imaging, many researchers have shown that high-intensity signals (HISs) in the carotid arterial wall on T1WI indicate the presence of intraplaque hemorrhage containing methemoglobin (5–7). Furthermore, some groups, including ours, have reported that coronary artery HISs on T1WI are associated with a vulnerable morphology and future cardiac events (8–11). However, the association between the localization of HISs, such as coronary intrawall or intraluminal, and plaque morphology was not investigated in previous studies.
Optical coherence tomography (OCT) was recently developed as a high-resolution imaging device for plaque characterization. Several studies have already shown that OCT allows the identification not only of plaque rupture, fibrous cap thickness, and intraluminal thrombus, but also of macrophages and intimal vasculature within atherosclerotic plaques in vivo (12–14). Therefore, we hypothesized that OCT might allow us to assess in detail the characteristics of coronary intrawall and intraluminal HISs on noncontrast T1WI. In this study, the term “localization” was used as a differentiation between intrawall and intraluminal, not suggesting the longitudinal distribution of HISs along the course of the coronary artery. The aim of this study was to investigate the relationship between localization of coronary HISs on noncontrast T1WI and plaque morphology detected by OCT, and the clinical severity of angina pectoris.
One hundred twenty-six consecutive patients with angina were prospectively enrolled in this study between September 2010 and January 2014. Patients with prior percutaneous coronary intervention, coronary artery bypass grafting, an occluded coronary vessel, or contraindications for MR were excluded from the study. Patients eligible for an early invasive strategy according to the American College of Cardiology Foundation/American Heart Association Guideline (elevated levels of cardiac biomarkers and signs or symptoms of heart failure) were also excluded. All patients underwent MR within 24 h before the day on which invasive coronary angiography (CAG) and OCT were performed. Of the 126 patients initially enrolled, 26 were excluded from the analysis for technical reasons as follows: 23 patients did not undergo OCT examination before percutaneous coronary intervention (11 had failure of OCT crossing, 11 were eligible for coronary artery bypass grafting, and 1 had an angiographically significant left main coronary stenosis), and 2 had MR images and 1 had OCT images that were of poor image quality and thus could not be analyzed.
Taking into account intrapatient correlations when evaluating the data, 1 culprit lesion was used in the analysis for any patient with more than 2 lesions. Thus, 100 lesions from 100 patients who had angiographically documented narrowing of at least 50% of the luminal diameter of a major coronary artery on CAG were examined in this study. Unstable angina pectoris (UAP) was diagnosed in 58 patients according to Braunwald criteria. Class I indicates new-onset severe or accelerated exertional angina within 2 months (n = 26); classes II (n = 20) and III (n = 12) indicate angina at rest during the previous month. The remaining stable angina pectoris group was made up of 42 patients with chest pain typical of cardiac ischemia on exertion that was clinically unchanged for >2 months. The culprit vessel was identified based on clinical, scintigram stress test and angiographic data. Oral aspirin (100 mg) and clopidogrel (75 mg) were administered on admission. Patients at high risk were also treated with intravenous heparin, but no patient was given thrombolytic agents.
The study was approved by the hospital ethics committee, and informed consent was obtained from all patients before the study.
MR coronary plaque image acquisition
Coronary plaque imaging was performed using a 1.5-T MR imager (Achieva, Philips Medical Systems, Best, the Netherlands) with a 5-element cardiac coil. Nitroglycerin (0.3 mg) was administered sublingually immediately before image acquisition to obtain high-quality MR images. Initial survey images were focused around the heart, following which the reference images were obtained for the sensitivity of parallel imaging. Transaxial cine MR images were then acquired using a steady-state free-precession sequence with breath holding, to determine the trigger delay time when the motion of the right coronary artery was minimal.
First, to obtain detailed information on the location of the target lesion, free-breathing, steady-state, free-precession, whole-heart coronary MR angiographic images were obtained (repetition time, 3.7 ms; echo time, 1.8 ms; flip angle, 80°; SENSE factor, 2.0; number of excitations, 1; navigator gating window of ± 2.0 mm with diaphragm drift correction; field of view, 300 × 255 × 120 mm [rectangular field of view, 85%]; acquisition matrix, 240 × 240; reconstruction matrix, 512 × 512 × 160, resulting in an acquired spatial resolution of 1.25 × 1.25 × 1.5 mm reconstructed to 0.6 × 0.6 × 0.75 mm) (10).
Next, coronary plaque images were obtained while the patients were breathing freely, by using a 3-dimensional T1WI, inversion-recovery, gradient-echo technique with fat-suppressed and radial k-space sampling in the Y-Z plane (repetition time, 4.4 ms; echo time, 2.0 ms; flip angle, 20°; SENSE factor, 2.5; number of excitations, 2; navigator gating window of ± 1.5 mm with diaphragm drift correction; field of view, 300 × 240 × 120 mm [rectangular field of view, 80%]; acquisition matrix, 224 × 224; reconstruction matrix, 512 × 512 × 140, resulting in an acquired spatial resolution of 1.34 × 1.34 × 1.7 mm reconstructed to 0.6 × 0.6 × 0.85 mm) (8–11). The patient-specific inversion time (409 ± 50 ms) of the inversion-recovery sequence was adjusted to null blood signal by using a Look-Locker sequence. The mean acquisition times for MR angiography and plaque imaging were 10 ± 3 min and 16 ± 3 min, respectively.
MR coronary plaque image analysis and classification
The location of the target lesion was determined by carefully comparing the CAG and MR angiographic images, by using fiduciary points, such as side branches. Once the target lesion had been confirmed with the coronary MR angiography, the areas corresponding to the above site in the coronary T1WI were carefully matched according to the surrounding cardiac and chest wall structures. Then, the ratio between the signal intensities of coronary plaque and cardiac muscle was calculated (PMR, defined as the highest signal intensity of the coronary plaque divided by the signal intensity of the left ventricular muscle near the coronary plaque, measured by placing a freehand circular region of interest on a standard console of the clinical MR system) (8,10,11). Areas with PMR ≤1.0 were classified as non-HISs. HISs with PMR >1.0 were then classified into 2 types, according to the localization of HIS, by using cross-sectional coronary T1WI.
Intrawall HIS was defined as a signal localized within the coronary wall, with the lumen identified as a black hole through the HIS site; intraluminal HIS was defined as a signal occupying the lumen, when the lumen was not, or even if partly, identified (Figure 1). The MR coronary image dataset was analyzed by 2 experienced cardiologists who were blinded to the plaque information obtained by OCT and the clinical data. In case of disagreement, consensus was reached by an additional joint reading. The interobserver and intraobserver coefficients of variation for measurement of the PMR were reported in our previous study (10), and the κ values for interobserver and intraobserver agreements with regard to classification of HIS types as intrawall or intraluminal were 0.89 and 0.93, respectively.
OCT image acquisition and analysis
OCT images were obtained of the culprit lesion, as previously defined. Thrombolysis or thrombectomy was not performed for any patient. The OCT images were acquired using a commercially available time-domain (M2 Cardiology Imaging system, LightLab Imaging, Inc., Westford, Massachusetts) or frequency-domain (ILUMIEN/ILUMIEN OPTIS OCT Intravascular Imaging System, St. Jude Medical, St. Paul, Minnesota) OCT system, as previously described (12–14).
OCT image analysis was performed by 2 experienced observers who were blinded to the clinical information, using previously established criteria for OCT plaque characterization (12–14). The presence of lipid, thin-cap fibroatheroma (TCFA), plaque rupture, calcification, thrombus, macrophage, and intimal vasculature on OCT images was evaluated within a 10-mm-long culprit lesion segment (5 mm proximal and 5 mm distal to the culprit lesion site with the smallest lumen cross-sectional area), in accordance with previous reports (10,13). If there was a discordance between the diagnosis of the 2 observers, a consensus diagnosis was obtained by using repeated off-line readings. When lipid characterized by signal-poor regions with diffuse borders was present over ≥90° in any of the cross-sectional images, it was considered a lipid-rich plaque. For each plaque, the largest lipid arc and lipid length were measured. TCFA was defined as a lipid-rich plaque with a fibrous cap thickness measuring ≤65 μm. Plaque rupture was defined as an intimal interruption and cavity formation in the plaque. Calcification was defined as a well-delineated, signal-poor region with sharp borders. Thrombus was defined as an irregular high- or low-backscattering mass protruding into the lumen. Thrombus type was further classified as either red or white. Red thrombus was defined as a high-backscattering protrusion with signal-free shadowing in the OCT images. White thrombus was defined as a signal-rich, low-backscattering projection. Macrophage accumulation was defined as bright spots with high OCT backscattering signal variances. Intimal vasculature was defined as a black hole or a tubular structure within a plaque. Plaque rupture, calcification, thrombus, macrophage, and intimal vasculature were assessed only for whether they were present or absent. The interobserver and intraobserver agreement measured as a κ statistic for each plaque characterization detected by OCT were also reported in our previous study (10).
Continuous data are presented as mean ± SD or median and interquartile range for non-normally distributed data. Comparisons of variables among the 3 groups were performed using one-way analysis of variance or the Kruskal-Wallis test when the variance was heterogeneous. Categorical variables were compared using the chi-square test. Independent predictors of intrawall HIS or intraluminal HIS were identified by entering all variables associated with a probability value <0.10 in the univariate analysis into a logistic regression analysis. The odds ratio and its 95% confidence interval for significant independent variables in the multivariate analysis were calculated. All calculations were performed using SPSS version 22.0 (SPSS, Chicago, Illinois) and p values <0.05 were considered significant.
Localization of HISs and clinical characteristics, angiographic, and OCT findings
Of 100 lesions from 100 patients, 37 patients (37%) had intrawall HIS, whereas intraluminal HIS was observed in 24 patients (24%). The remaining 39 patients (39%) had non-HIS lesions. Table 1 shows the baseline clinical characteristics and angiographic findings in patients with intrawall or intraluminal HIS, and non-HIS. The PMR was significantly higher in both types of HIS lesion than in non-HIS lesions. Moreover, the PMR of intraluminal HIS was significantly higher than that of intrawall HIS. There were no significant differences in age, sex, or presence of risk factors, except for hypertension, among the 3 groups. The distribution of the culprit vessel was significantly different. Both types of HIS were observed more frequently shown in the right coronary artery and less frequently in the left circumflex artery than non-HIS. The percent diameter stenosis was significantly greater in the lesions with intraluminal HISs than in non-HIS lesions.
The relationship between intrawall or intraluminal HIS and non-HIS on T1WI and plaque morphology assessed by OCT is shown in Table 2. The frequency of lipid-rich plaque, TCFA, plaque rupture, calcification, thrombus, macrophage, and intimal vasculature differed significantly among the 3 groups. In both types of HIS lesion, the frequency of lipid-rich plaque, TCFA, and macrophage accumulation was higher than in non-HIS lesions. The frequency of plaque rupture, thrombus, and intimal vasculature was highest in intraluminal HIS lesions. In contrast, the frequency of calcification was highest in non-HIS lesions.
Factors related to the localization of HISs
Multivariate logistic regression analyses were performed to identify independent factors associated with total HISs, and intrawall or intraluminal HISs, compared with non-HISs (Table 3). Total HISs were associated with thrombus and the absence of calcification. Intrawall HISs were associated with macrophage accumulation (p = 0.019) and the absence of calcification (p = 0.014). In contrast, thrombus (p = 0.003) and intimal vasculature (p = 0.045) were independent factors associated with intraluminal HISs.
A representative case of an intrawall HIS lesion on T1WI compared with plaque morphology on OCT is shown in Figure 2. Whole-heart coronary MR angiography and CAG show a severe coronary stenosis in the mid right coronary artery. The area corresponding to the vessel wall in the stenotic lesion shows HIS on T1WI. Macrophage accumulation was detected by OCT, and calcification was not present in the culprit lesion. Figure 3 shows a case of an intraluminal HIS lesion. In this case, the HIS lesion occupying the lumen in the left anterior descending coronary artery contained a large thrombus and intimal vasculature.
Localization of HISs and clinical severity of angina pectoris
Figure 4 shows the correlation between the localization of HISs and the clinical severity of coronary syndromes. Among patients with intrawall HIS, intraluminal HIS, and non-HIS, the distribution of the clinical severity of angina according to Braunwald classification was significantly different (p < 0.03) (Figure 4A). The 59% of patients with non-HISs were clinically diagnosed with stable angina pectoris. However, in patients with intrawall HISs the frequency of UAP was greater. Moreover, 50% of patients with intraluminal HISs experienced rest angina, such as Braunwald class II or III.
From the opposite point of view, according to the clinical severity, the frequency of the HIS classification was again significantly different (Figure 4B). In patients with UAP with Braunwald class II or III angina, the frequencies of intrawall and intraluminal HIS were nearly equal. In patients with UAP with class I, the frequencies of intrawall HIS and non-HIS were nearly equal. Conversely, in patients with stable angina pectoris, non-HIS lesions were the most frequent (55%).
To the best of our knowledge, this is the first study to show a relationship between the localization of coronary HISs on T1WI, plaque morphology assessed by OCT, and the clinical severity of angina pectoris.
Kawasaki et al. (8), using intravascular ultrasound and multislice computed tomography, reported that coronary HIS on T1WI was associated with a high frequency of low attenuation and positive remodeling. Jansen et al. (9) demonstrated that 10 of 18 patients who suffered acute myocardial infarction within 72 h after the initial onset of symptoms were found to have intracoronary thrombus as detected by invasive CAG, and that HIS on T1WI correctly identified intracoronary thrombus. Recently, although only in a small number of patients, we also demonstrated a direct association between coronary HISs on T1WI and the presence of intracoronary thrombus as detected through OCT (10). In the present study population, multivariate analysis revealed that thrombus and intimal vasculature were independent factors associated with intraluminal HISs. The presence of intimal vasculature in ruptured and hemorrhagic plaques may predispose to thrombus formation. Findings from previous reports (9,10) and our present study indicate that the presence of early intracoronary thrombus containing methemoglobin can produce T1 shortening, which might be related to the intraluminal HIS formation on T1WI.
In contrast, the relationship between intrawall HISs and plaque morphology remains unclear. Histopathologic studies using carotid endarterectomy specimens demonstrated that intrawall HIS on T1WI was a marker of carotid complicated plaque with methemoglobin in intraplaque hemorrhage (5–7). Several lines of evidence have suggested that an increase in the amount of lipid core, mechanical stresses, overproduction of oxygen free radicals by macrophages, and lack of calcification could lead to breakdown of microvessels and intraplaque hemorrhage production (3,4). Our present multivariate analysis demonstrated that intrawall HIS was related not only to the presence of macrophages, but also to a lack of calcification. Moreover, we showed that in intrawall HIS lesions, the presence of lipid-rich plaques (intrawall HIS, 76%; non-HIS, 33%) and intimal vasculature (intrawall HIS, 46%; non-HIS, 15%) was more frequent than in non-HIS lesions, although these differences were not statistically significant by multivariate analysis.
Taking previous data together with our present findings, one can speculate that coronary intrawall HISs on T1WI may indicate intraplaque hemorrhage associated with inflammation. Some observations indicate that hemorrhage components appear as signal-poor OCT regions that must be distinguished from lipid necrotic pools (15). On the current OCT imaging, the discrimination of hemorrhage and lipid component and the identification of intimal vasculature, which were present within hemorrhage or had a smaller size, with a diameter of <50 μm, are major issues that require additional validation studies using histopathologic materials from coronary rather than carotid arteries.
Questions arise concerning what appears as HIS on T1WI, or why there are differences in PMR values between intrawall and intraluminal HISs. At this stage, it is fair to state that this remains speculative. Methemoglobin from hemoglobin breakdown after erythrocyte extravasation under the various intraplaque environments has been proposed to account for the high signal intensity of intraplaque hemorrhage. Therefore, intraplaque hemorrhage includes components of different ages and volumes of methemoglobin (4,7). In contrast, when intraluminal thrombus is formed, intact erythrocytes are trapped within a mesh of platelets and fibrin. Moreover, intraluminal thrombus formation develops from plaque rupture based on the presence of vulnerable complex plaques, associated with a necrotic core and intraplaque hemorrhage. However, the signal intensity of thrombus on T1WI varies according to thrombus age. Habs et al. (16) reported that multicontrast MR imaging with T1WI is useful for determining the age of arterial hematoma in patients with spontaneous cervical artery dissection and classify cases into four stages: 1) acute (≤3 days); 2) early subacute (>3 days); 3) late subacute (>7 days); and 4) chronic (>14 days). In their study, acute hematoma produced hypointensity on T1WI. By contrast, early or late subacute hematoma produced hyperintensity on T1WI. Finally, the signal intensity in the chronic stage was hypointense on T1WI. This might explain why some patients have intraluminal thrombus on OCT but no HIS on T1WI. Thus, the proportion, age, and volume of erythrocytes, including methemoglobin, may determine the PMR values. Noguchi et al. (11) demonstrated that a PMR cutoff value of 1.4 was best for identifying vulnerable coronary plaques associated with future cardiac events. Moreover, their stratified analysis using PMR values of 1.0 and 1.4 revealed that the incidence of cardiac events was well differentiated: 25.8% for PMR ≥1.4, 8.4% for PMR 1.0 to 1.4, and 1.1% for PMR <1.0. Thus, HISs with a higher PMR are likely to represent vulnerable plaques that develop into cardiac events (11).
Previous angiographic studies have revealed that the incidence of angiographic evidence of complex lesions and/or thrombosis rose progressively with higher UAP classes (17). However, at the same time, these studies showed that, even in Braunwald class I, one-third of these patients had a complex lesion. The present study demonstrated that in patients with UAP with Braunwald class II or III the frequencies of intrawall HIS and intraluminal HIS were nearly equal, whereas intrawall HIS was the most frequent in class I. Our observation is of considerable interest, because it seems that the presence and localization of HISs in a coronary artery is related to the clinical severity of angina pectoris, as well as plaque vulnerability. Moreover, these observations could be of help in the identification of lesions or patients at high risk, which could lead to more appropriate risk-stratification and treatment. T1WI is a promising tool for the in vivo detection of vulnerable plaques associated with thrombus or intraplaque hemorrhage. Intraplaque hemorrhage has been associated with more rapid growth of the lipid core and accelerated enlargement of the plaque size, resulting in luminal narrowing (6,7). If it was possible to identify patients with intraplaque hemorrhage in the coronary arteries, that could prove invaluable in optimizing treatment strategies for atherosclerosis. Furthermore, it has already been reported that HISs on T1WI are significantly associated with coronary events (11). Therefore, this refinement in the judgment and interpretation of HISs provides incremental value to the MR evaluation of coronary atherosclerotic plaques.
First, because an inversion-recovery gradient-echo sequence is used for the T1WI, issues with spatial resolution and partial volume effect could provide artifacts that look like HIS. It is sometimes difficult to discriminate the localization of HIS, although the κ value for this finding between investigators was sufficiently high. Moreover, in cases where the lumen was not, or even if partly, identified, they were classified into intraluminal HISs. This method of analysis might have masked important findings and potentially led to inaccurate results. Second, in the present study, areas with PMR >1.0 were defined as HIS. Using this definition, the median PMR of intrawall HISs was 1.48, and this prevalence was 37%, relatively high. Further studies are needed to establish the appropriate PMR cutoff value for identifying vulnerable plaque and predicting future cardiac events.
Third, because no comparison with histopathologic data was performed in this study, the precise characterization of both types of HISs remains unknown. In particular, macrophage accumulation was not quantified or rigorously validated. Therefore, the results should be interpreted with caution. Fourth, MR imaging visualizes the entire wall, albeit with a limited spatial resolution. Meanwhile, OCT cannot visualize as deeply within some lipid-containing plaques because of the poor axial penetration. Nevertheless, OCT is acknowledged as one of the most reliable tools for assessment of coronary plaque characterization. Moreover, because the most important morphological determinants of plaque vulnerability are superficial, the region of greatest interest was still within the imaging range of the current OCT system. We consider, therefore, that the quality of both coronary MR and OCT data obtained by using this approach is sufficiently high enough to validate our conclusion. Last, patients eligible for an early invasive strategy or without an OCT examination before percutaneous coronary intervention because of crossing failure, were excluded from this study. Therefore, our results were limited by selection bias and may not apply to such patients.
This study shows that intrawall and intraluminal HISs on T1WI in patients with angina are related to the different types of vulnerable plaque morphology assessed through OCT and the clinical severity. This novel concept regarding the localization of coronary HISs could impact on treatment strategies for atherothrombotic disease and may be useful for understanding the pathophysiological mechanisms of atherothrombotic plaque formation.
COMPETENCY IN MEDICAL KNOWLEDGE: The localization of HISs, such as coronary intrawall or intraluminal, on T1WI with the noncontrast MR technique is related to the different types of vulnerable plaque morphology assessed through OCT and the clinical severity of angina pectoris. This refinement in the judgment and interpretation of HISs provides incremental value to the MR evaluation of coronary atherosclerotic plaques.
TRANSLATIONAL OUTLOOK: The proportion, age, and volume of erythrocytes, including methemoglobin, may determine the PMR values. Further studies are needed to establish the appropriate PMR cutoff value for identifying vulnerable plaque and staging of intraluminal thrombus or intraplaque hemorrhages in vivo.
This work was supported by JSPS KAKENHI (23591057). All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- coronary angiography
- high-intensity signal
- magnetic resonance
- optical coherence tomography
- the ratio between the signal intensities of coronary plaque and cardiac muscle
- thin-cap fibroatheroma
- T1-weighted imaging
- unstable angina pectoris
- Received March 10, 2015.
- Revision received June 3, 2015.
- Accepted June 26, 2015.
- American College of Cardiology Foundation
- Virmani R.,
- Kolodgie F.D.,
- Burke A.P.,
- Farb A.,
- Schwartz S.M.
- Michel J.B.,
- Virmani R.,
- Arbustini E.,
- Pasterkamp G.
- Moody A.R.,
- Murphy R.E.,
- Morgan P.S.,
- et al.
- Takaya N.,
- Yuan C.,
- Chu B.,
- et al.
- Sun J.,
- Underhill H.R.,
- Hippe D.S.,
- Xue Y.,
- Yuan C.,
- Hatsukami T.S.
- Kawasaki T.,
- Koga S.,
- Koga N.,
- et al.
- Jansen C.H.,
- Perera D.,
- Makowski M.R.,
- et al.
- Ehara S.,
- Hasegawa T.,
- Nakata S.,
- et al.
- Noguchi T.,
- Kawasaki T.,
- Tanaka A.,
- et al.
- Yabushita H.,
- Bouma B.E.,
- Houser S.L.,
- et al.
- Tearney G.J.,
- Yabushita H.,
- Houser S.L.,
- et al.
- Prati F.,
- Regar E.,
- Mintz G.S.,
- et al.
- Rupprecht H.J.,
- Sohn H.Y.,
- Kearney P.,
- Bickel C.,
- Nafe B.,
- Meyer J.