2015 Volume 79 Issue 10 Pages 2231-2237
Background: The ability of iMap-intravascular ultrasound (IVUS) tissue characterization to detect thin-cap fibroatheroma (TCFA) identified on optical coherence tomography (OCT) has not yet been fully elucidated.
Methods and Results: We evaluated 86 coronary lesions from 73 patients with stable angina pectoris using iMap-IVUS and OCT. We defined OCT-derived TCFA (OCT-TCFA) as lipid-rich plaque with a <65-μm-thick fibrous cap. The external elastic membrane (EEM) cross-sectional area (CSA), lumen CSA, plaque plus media (P+M) CSA, plaque burden and remodeling index were measured on gray-scale IVUS. Plaque components categorized on iMap-IVUS as fibrotic, lipidic, necrotic or calcified are presented as absolute area and proportion (%) of total plaque area. OCT-TCFA (22 lesions) had significantly greater EEM CSA, P+M CSA, plaque burden and remodeling index than non-TCFA (64 lesions). Significantly larger %necrotic area, absolute lipidic and necrotic areas and smaller %fibrotic areas were found in OCT-TCFA than in non-TCFA. On multivariate analysis, absolute necrotic area was an independent predictor of OCT-TCFA. The area under the ROC curve for absolute necrotic area required to identify OCT-TCFA was 0.86. The sensitivity, specificity, positive and negative predictive values of absolute necrotic area ≥7.3 mm2 for identifying OCT-TCFA were 77%, 88%, 68% and 92%, respectively.
Conclusions: Coronary lesions with greater iMap-IVUS absolute necrotic area were closely associated with OCT-TCFA. (Circ J 2015; 79: 2231–2237)
Acute coronary syndrome (ACS) mostly arises due to rupture of vulnerable plaque followed by luminal thrombosis.1 Thin-cap fibroatheroma (TCFA) is a primary type of rupture-prone vulnerable plaque that includes a large necrotic core, occupying approximately one-quarter of the plaque area, covered by a thin fibrous cap, usually <65 μm.2 The detection of TCFA might be relevant to prognosis and optimal treatment choice, but simple and reliable methods of diagnosis have not yet been established.
Editorial p 2112
Gray-scale intravascular ultrasound (IVUS) is a popular intracoronary imaging modality that is used to observe coronary plaques and measure plaque burden in the clinical setting, but its resolution of 100–150 μm is insufficient for accurate assessments of plaque characteristics.3,4 During the past decade, tissue characterization using radiofrequency (RF) signal-based IVUS (RF-IVUS) systems have been developed to provide objective and quantitative information on plaque components.5,6 iMap-IVUS (Boston Scientific, Natick, MA, USA) is the most recent RF-IVUS system to incorporate a 40-MHz rotating single element catheter. iMap-IVUS can classify coronary plaque into fibrotic, lipidic, necrotic or calcified components using spectral RF analysis with a classification algorithm generated from histological findings ex vivo.7 Validation studies have proven that iMap-IVUS can accurately characterize tissue both ex vivo and in vivo.8 We therefore postulated that tissue characterization using iMap-IVUS would reflect TCFA components. We selected intravascular optical coherence tomography (OCT) as the reference standard because its 10–20-μm resolution allows accurate measurement of fibrous cap thickness, and it has been proposed as an optimal modality for detecting TCFA in vivo.9,10 Thus, the aim of the present study was to determine whether tissue characterization using iMap-IVUS could reflect OCT-determined TCFA in vivo.
We compared the characteristics of coronary plaques in vivo between iMap-IVUS and OCT between April 2010 and January 2014. Ninety-three patients with stable angina pectoris were recruited from patients scheduled to undergo elective percutaneous coronary intervention (PCI) for native de novo coronary lesions at Nagasaki University Hospital. All lesions selected for imaging were the target of planned PCI. Target lesions were defined as being responsible for ischemia identified on electrocardiography, left ventricular wall motion abnormalities, angiographic lesion morphology, stress myocardial perfusion imaging and fractional flow reserve. Patients with a left main lesion, or lesions that were ostial or severely calcified (>180°), chronic total occlusion, or poor-quality OCT or IVUS were excluded. We thus analyzed 86 target lesions from 73 patients with stable angina pectoris defined as a constant frequency, duration or intensity of angina symptoms for 6 weeks before admission. This study complied with the Declaration of Helsinki with regard to human investigation, and the Ethics Committee of Nagasaki University Hospital approved the protocol. All of the included patients provided written, informed consent to participate in the study before enrollment.
AngiographyWe assessed traditional lesion types using the American Heart Association/American College of Cardiology classification,11 and angiographically mapped culprit vessels and lesion locations (proximal, mid, or distal). Minimal lumen diameter, reference vessel diameter, diameter stenosis and lesion length were measured on quantitative coronary angiograms using the CASS II system (Pie Medical Imaging, Maastricht, the Netherlands).
Imaging ProcedureIVUS IVUS was carried out after administration of intracoronary isosorbide dinitrate (1 mg) using a 40-MHz, Atlantis SR Pro 2/OptiCross IVUS catheter (Boston Scientific). IVUS of target lesions was carried out using automated pullback at 0.5 mm/s. Images were digitally stored for subsequent off-line analysis.
OCT After IVUS, OCT of 38 (44%) and 48 (56%) target lesions was acquired using M2 OCT time-domain (LightLab Imaging, Westford, MA, USA) or C7 OCT frequency-domain (St. Jude Medical, St Paul, MN, USA) imaging systems, respectively. In brief, the M2 system uses a 3F Helios occlusion balloon (LightLab Imaging) that is inflated proximal to the lesion at 0.4–0.6 atm during image acquisition. A 0.016-inch imaging wire is automatically pulled back from a distal to a proximal position at a rate of 1.0 mm/s, and saline is continuously infused from the tip of the occlusion balloon to remove blood from the field of view. The C7 system requires the advance of a 2.7-F Dragonfly OCT imaging catheter (LightLab Imaging) distal to the lesion, with automated pullback at a rate of 20 mm/s starting as soon as injected contrast media has cleared any blood. The acquired OCT images were digitally stored for subsequent offline analysis.
Imaging AnalysisGray-Scale IVUS Before OCT, quantitative gray-scale IVUS of all target lesions was carried out every 1.0 mm independently of the cardiac cycle using the validated software, echoPlaqueTM 3.0 (INDEC Medical Systems, Santa Clara, CA, USA). Lumen cross-sectional area (CSA), external elastic membrane (EEM) CSA and plaque plus media (P+M) CSA were measured according to the American College of Cardiology Clinical Expert Consensus Document.12 P+M CSA was calculated as EEM CSA minus lumen CSA. Plaque burden was calculated as P+M CSA/EEM CSA. These IVUS parameters were measured at the minimum lumen CSA sites and reference sites. The minimum lumen CSA site was determined as the location of the smallest amount of lumen in all analyzed cross-sections. The reference sites consisted of cross sections that appeared the most normal within 10 mm proximal and distal to the lesion. The remodeling index was calculated as the lesion EEM CSA divided by the mean reference EEM CSA. Positive remodeling was defined as remodeling index >1.0.13 Echo-attenuated plaque was defined as atherosclerotic plaque with ultrasound signal attenuation without very high intensity echo reflectors that involved 90° of the vessel circumference.14
iMap-IVUS A single cross-section of the minimum lumen CSA site measured on gray-scale IVUS was analyzed on iMap-IVUS using QIvus 2.1 (Medis Medical Imaging Systems, Leiden, the Netherlands) as described.7 Briefly, vessel and lumen borders were traced using automatic edge detection and manually corrected when necessary. Plaque components were then classified as fibrotic (light-green), lipidic (yellow), necrotic (pink) or calcified (light-blue), and are described as absolute areas (mm2) and proportions (%) of total plaque area. Guidewire artifacts were masked from all images. One investigator who was blinded to the OCT data initially analyzed the iMap-IVUS images. The same investigator and another subsequently reanalyzed the same images to assess the intraobserver and interobserver reproducibility of the measurements. The intraobserver (r=0.98, 0.96, 0.98 and 0.91) and interobserver (r=0.96, 0.97, 0.98 and 0.97) variability for fibrotic, lepidic, necrotic and calcified tissues, respectively, were acceptable.
OCT The OCT images were quantitatively and qualitatively analyzed using an offline review system (St. Jude Medical) at a single cross-section of the location corresponding to the minimum lumen CSA site determined on gray-scale IVUS. Identical cross-sections between IVUS and OCT were determined based on the distance from landmarks, such as branches, calcifications, and previously implanted stents. One investigator who was blinded to the iMap-IVUS data analyzed the OCT images. The same investigator and another analyzed the same images once again to assess the intraobserver and interobserver reproducibility of OCT analysis. Plaque with high backscattering and relatively homogeneous OCT signals was defined as fibrous. Signal-poor lesions with poorly delineated borders indicated a lipid core, and signal-rich homogeneous lesions overlying a lipid core indicated a fibrous cap. A lipid core that subtended ≥90° with a protractor located in the center of a lumen cross-section was considered as lipid-rich plaque. The thinnest part of a fibrous cap was measured 3 times, and the average was defined as fibrous cap thickness. We defined OCT-derived TCFA (OCT-TCFA) as lipid-rich plaque with the thinnest part of the fibrous cap <65μm.15 Fibrous and lipid-rich plaque with thick fibrous caps were classified as non-TCFA. Calcium deposition was defined as a signal-poor or heterogeneous region with a sharply delineated border.16 Intraobserver and interobserver agreement regarding OCT-TCFA detection was within an acceptable range (κ=0.84 and 0.78, respectively). Figure 1 shows representative OCT and iMap-IVUS images.
Representative optical coherence tomography (OCT), gray-scale and iMap-intravascular ultrasound (IVUS). (A) Fibrous plaque (non-thin-cap fibroatheroma [non-TCFA]). (B) Lipid-rich plaque (non-TCFA). Fibrous cap (arrows) is 193 μm thick. (C) TCFA. Fibrous cap (arrows) is 60 μm thick. Plaque components on iMap-IVUS are classified as fibrotic (light-green), lipidic (yellow), necrotic (pink) or calcified (light-blue). *Guidewire artifacts.
Statistical analysis was performed using IBM SPSS Statistics version 20.0 (SPSS, Chicago, IL, USA) and MedCalc version 13.2 (MedCalc Software, Ostend, Belgium). Continuous data are expressed as mean±SD and were compared using unpaired t-test for normally distributed values, or Mann-Whitney U-test. Categorical data are presented as n (%) and were compared using the chi-squared test or Fisher’s exact test. Relationships between OCT and IVUS data were assessed using Pearson’s correlation coefficient or Spearman’s rank correlation coefficient. Significant IVUS parameters indicating the presence of OCT-TCFA were identified on multivariate logistic regression analysis. Variables with P<0.05 on univariate analysis were entered into the multivariate analysis. Cut-offs for IVUS parameters with regard to identification of OCT-TCFA were determined from the Youden index based on receiver operating characteristic (ROC) curves. Sensitivity, specificity, positive and negative predictive values, and accuracy were calculated at a cut-off. Interobserver and intraobserver agreement for TCFA identification was evaluated using Cohen’s κ test of concordance. P<0.05 was considered statistically significant.
Characteristics of the patients and angiographic lesions of target vessels are summarized in Table 1. A total of 86 coronary target plaques were selected from 73 patients.
| Clinical characteristics | |
| Male | 56 (77) |
| Age (years) | 69±10 |
| Hypertension | 52 (71) |
| Hypercholesterolemia | 41 (56) |
| Diabetes mellitus | 35 (48) |
| Current smoker | 17 (23) |
| Body mass index (kg/m2) | 24.0±3.3 |
| Lesion characteristics | |
| No. target lesions | 86 |
| Target vessel | |
| LAD/LCX/RCA | 39 (45)/18 (21)/29 (34) |
| Target lesion location | |
| Proximal/Mid/Distal | 38 (44)/30 (35)/18 (21) |
| Reference diameter (mm) | 2.81±0.56 |
| Minimal lumen diameter (mm) | 0.90±0.33 |
| %Diameter stenosis | 68±12 |
| Lesion length (mm) | 15.9±9.1 |
Data given as mean±SD or n (%). LAD, left anterior descending artery; LCX, left circumflex artery; RCA, right coronary artery.
Comparison of OCT, gray-scale IVUS and iMap-IVUS for TCFA and non-TCFA lesion is given in Table 2. Based on OCT analysis, 22 (26%) and 64 (74%) target plaques were classified as OCT-TCFA and non-TCFA, respectively. Sixty-four non-TCFA plaques included 12 (14%) that were fibrous and 52 (61%) that were lipid rich with a thick fibrous cap. On gray-scale IVUS, EEM CSA, P+M CSA, plaque burden, and remodeling index were significantly greater, and the prevalence of positive remodeling and echo-attenuated plaque were significantly higher in OCT-TCFA than in non-TCFA. On iMap-IVUS, significantly larger %necrotic areas, absolute lipidic and necrotic areas, and a significantly smaller %fibrotic area were seen in OCT-TCFA compared with non-TCFA. Other iMap-IVUS parameters were similar between the 2 groups.
| Non-TCFA (n=64) | TCFA (n=22) | P-value | |
|---|---|---|---|
| OCT | |||
| Fibrous plaque | 12 (14) | 0 | – |
| Lipid-rich plaque with thick fibrous cap | 52 (61) | 0 | – |
| Calcium deposition | 22 (34) | 7 (32) | 0.83 |
| Fibrous cap thickness (μm) | 192.0±137.5 | 54.2±6.8 | <0.001 |
| Lipid arc (°) | 166±66 | 204±75 | 0.11 |
| Gray-scale IVUS | |||
| EEM CSA (mm2) | 12.5±5.0 | 18.8±6.2 | <0.001 |
| Lumen CSA (mm2) | 2.5±0.8 | 2.8±0.8 | 0.086 |
| P+M CSA (mm2) | 10.0±4.9 | 16.0±5.9 | <0.001 |
| Plaque burden (%) | 78.1±7.5 | 83.7±6.1 | 0.002 |
| Remodeling index | 1.00±0.19 | 1.10±0.24 | 0.035 |
| Positive remodeling | 18 (29) | 14 (64) | 0.003 |
| Echo-attenuated plaque | 13 (20) | 13 (59) | 0.001 |
| iMap-IVUS | |||
| % Fibrotic area | 48.8±17.8 | 33.0±15.4 | <0.001 |
| % Lipidic area | 10.2±3.8 | 10.8±3.5 | 0.42 |
| % Necrotic area | 35.9±16.1 | 53.0±14.3 | <0.001 |
| % Calcified area | 1.7±2.0 | 1.1±0.8 | 0.64 |
| Absolute fibrotic area (mm2) | 4.37±2.03 | 4.93±2.14 | 0.29 |
| Absolute lipidic area (mm2) | 1.02±0.71 | 1.66±0.76 | 0.001 |
| Absolute necrotic area (mm2) | 3.75±2.80 | 8.56±3.35 | <0.001 |
| Absolute calcified area (mm2) | 0.15±0.16 | 0.19±0.16 | 0.20 |
Data given as mean±SD or n (%). CSA, cross-sectional area; EEM, external elastic membrane; IVUS, intravascular ultrasound; OCT, optical coherence tomography; P+M, plaque plus media; TCFA, thin-cap fibroatheroma.
Gray-scale IVUS parameters negatively correlated with fibrous cap thickness (Table 3), except for remodeling index. On iMap-IVUS, %necrotic area, absolute lipidic and necrotic area negatively correlated, whereas %fibrotic area positively correlated, with fibrous cap thickness.
| Fibrous cap thickness (μm) | ||
|---|---|---|
| r | P-value | |
| Gray-scale IVUS | ||
| EEM CSA (mm2) | −0.49 | <0.001 |
| P+M CSA (mm2) | −0.49 | <0.001 |
| Plaque burden (%) | −0.34 | 0.001 |
| Remodeling index | −0.17 | 0.12 |
| iMap-IVUS | ||
| % Fibrotic area | 0.62 | <0.001 |
| % Lipidic area | −0.17 | 0.12 |
| % Necrotic area | −0.62 | <0.001 |
| % Calcified area | 0.18 | 0.092 |
| Absolute fibrotic area (mm2) | 0.04 | 0.72 |
| Absolute lipidic area (mm2) | −0.45 | <0.001 |
| Absolute necrotic area (mm2) | −0.64 | <0.001 |
| Absolute calcified area (mm2) | −0.02 | 0.88 |
Abbreviations as in Table 2.
Logistic regression analysis identified significant IVUS parameters associated with OCT-TCFA. The final multivariate model included the parameters EEM CSA, P+M CSA, plaque burden, positive remodeling, echo-attenuated plaque, %fibrotic, %necrotic area and absolute lipidic and necrotic areas that had P<0.05 in the univariate model. On multivariate analysis, absolute necrotic area was an independent parameter associated with OCT-TCFA (Table 4).
| OR (95% CI) | P-value | |
|---|---|---|
| Univariate | ||
| EEM CSA | 1.22 (1.09–1.36) | <0.001 |
| P+M CSA | 1.23 (1.10–1.37) | <0.001 |
| Plaque burden | 1.14 (1.04–1.24) | 0.004 |
| Positive remodeling | 4.38 (1.57–12.21) | 0.005 |
| Echo-attenuated plaque | 5.67 (1.99–16.22) | 0.001 |
| % Fibrotic area | 0.93 (0.89–0.97) | 0.002 |
| % Necrotic area | 1.09 (1.04–1.14) | <0.001 |
| Absolute lipidic area | 2.86 (1.42–5.75) | 0.003 |
| Absolute necrotic area | 1.57 (1.28–1.92) | <0.001 |
| Multivariate | ||
| Absolute necrotic area | 1.56 (1.28–1.91) | <0.001 |
Abbreviations as in Table 2.
Figure 2 shows the ROC curve analysis of the absolute necrotic area for classifying OCT-TCFA. The area under the curve of the absolute necrotic area was 0.86. The optimal cut-off for the absolute necrotic area to identify OCT-TCFA was ≥7.3 mm2. The sensitivity, specificity, positive, negative predictive values, and accuracy for this cut-off to detect OCT-TCFA are summarized in Table 5.
Receiver operating characteristic (ROC) curves for identification of optical coherence tomography-derived thin-cap fibroatheroma (OCT-TCFA) using absolute necrotic area. AUC, area under the curve.
| No. true positives | 17 |
| No. true negatives | 56 |
| No. false positives | 8 |
| No. false negatives | 5 |
| Sensitivity (%) | 77 |
| Specificity (%) | 88 |
| Positive predictive value (%) | 68 |
| Negative predictive value (%) | 92 |
| Accuracy (%) | 85 |
Abbreviations as in Table 2.
The main findings of the present study are as follows. Greater iMap-IVUS necrotic and lipidic components and a smaller iMap-IVUS fibrotic component were more frequently associated with OCT-TCFA. Larger absolute necrotic area was the sole iMap-IVUS parameter independently associated with OCT-TCFA. The optimal cut-off of absolute necrotic area to distinguish OCT-TCFA from non-TCFA was ≥7.3 mm2.
Various imaging modalities have been used to detect TCFA in vivo. Although gray-scale IVUS has served as the gold standard of current invasive imaging modalities for detecting coronary plaque, it might not accurately assess plaque characteristics because of its low resolution. The RF-IVUS system can differentiate various plaque phenotypes and might thus be able to identify vulnerable plaque. Virtual histology (VH)-IVUS (Volcano, Rancho Cordova, CA, USA) is the first commercially available RF-IVUS system to include a 20-MHz transducer catheter. The reported accuracy of VH-IVUS for detecting fibrous tissue, fibro-fatty components, dense calcium and necrotic cores ex vivo is 90.4%, 92.8%, 90.9% and 89.5%, respectively.5 In contrast, iMap-IVUS is a novel 40-MHz RF-IVUS system that uses spectral RF analysis to classify coronary plaque into fibrotic, lipidic, necrotic and calcified components. In a validation ex vivo study, the accuracy of this system for detecting necrotic, lipidic, fibrotic and calcified components was 97%, 98%, 95%, and 98%, respectively.7 An evaluation of inter- and intraobserver, intracatheter and intercatheter iMap-IVUS measurements has found the reproducibility acceptable.17 Compared with electronic non-sheath based VH-IVUS imaging catheters, the iMap-IVUS catheter is a sheath-based mechanical imaging catheter that enables more precise pullback as well as measurement of the size and composition of atherosclerotic plaques.17 A recent study using iMap-IVUS found that coronary plaques with positive remodeling contained greater lipidic and necrotic components compared with those with negative remodeling.18 Other iMap-IVUS studies have shown that culprit plaques have more lipidic and necrotic components and a smaller fibrotic component in patients with, than without ACS.19,20 This suggests that iMap-IVUS can provide useful information for identifying vulnerable plaque in vivo.
The present results closely agree with those of a previous VH-IVUS study that demonstrated that greater plaque burden, remodeling index, necrotic component and a smaller fibrous component were associated with OCT-TCFA.21 The accuracy of diagnosing OCT-TCFA, however, differs between iMap-IVUS and VH-IVUS. In the present study, the sensitivity, specificity, positive and negative predictive values of absolute necrotic area ≥7.3 mm2 to identify OCT-TCFA were 77%, 88%, 68% and 92%, respectively. In contrast, Kubo et al showed that the sensitivity, specificity, positive and negative predictive values of VH-IVUS to identify OCT-TCFA were 89%, 86%, 59% and 97%, respectively.22 Thus, the sensitivity of iMap-IVUS to identify OCT-TCFA is apparently lower than that of VH-IVUS. Several factors could explain the differences in diagnostic accuracy between iMap-IVUS and VH-IVUS. Despite having better resolution, far-field images provided by the 40-MHz catheter in iMap-IVUS are not always as clear as those obtained with 20-MHz VH-IVUS due to amplified attenuation.23 Such attenuated areas are often considered necrotic on iMap-IVUS. Second, iMap-IVUS has limitations for tissue characterization for acoustic shadows comprising dense calcium, because accurate RF data are difficult to obtain from such areas. For example, acoustic shadows behind dense calcium are frequently classified as necrotic tissue on iMap-IVUS. Therefore, patients with severely calcified lesions (>180°) were excluded from the present study. Third, shadows arising from guidewire artifacts are expressed as necrotic tissue on iMap-IVUS.23 Although the software helps to mask such shadows and exclude them from analysis, information is always lost depending on the location of the guidewire. For instance, if the guidewire shadow is projected onto coronary plaque, a large part of that plaque will be excluded from analysis, compared with projection of the shadow onto a region of the vessel wall without plaque.17 Thus, plaque areas described as “necrotic components” on iMap-IVUS might be overestimated compared with the true amount of necrotic core determined on histopathology. Indeed, %necrotic area of TCFA in the present study was larger than that identified in a postmortem histological study (53.0±14.3% vs. 25.2±15.7%).24 Such overestimation of necrotic areas might lead to the relatively low sensitivity of iMap-IVUS in detecting OCT-TCFA. In contrast, the negative predictive value of iMap-IVUS for OCT-TCFA was relatively high (92%). Thus, iMap-IVUS may be able to be used to exclude OCT-TCFA.
In the present study, iMap-IVUS analysis was done for a single frame at the minimum lumen CSA site. A previous study showed that VH-IVUS analysis needs at least 3 consecutive frames to detect TCFA.25 Others have reported the usefulness of volumetric VH-IVUS analysis using more many frames per entire target lesion.26 In contrast, the methodology of frame analysis to identify TCFA in iMap-IVUS analysis has not been established, because iMap-IVUS and pathological findings have not been fully validated. In addition, volume analysis is difficult to apply in the clinical setting because it is time-consuming, labor intensive and requires specialist training. For these reasons, iMap-IVUS analysis was done only for a single frame. Further investigation is needed to establish an appropriate methodology for frame analysis using iMap-IVUS.
We used OCT-TCFA as a surrogate for histological TCFA. Often conceptually regarded as vulnerable plaque, TCFA plaques are characterized by large necrotic core encased by thin fibrous cap,2 a feature that is visualized on OCT as a signal-rich layer over a signal-poor area.9 Kume et al noted a close correlation between fibrous cap thickness on OCT and histology.27 Uemura et al showed that OCT-TCFA is a potential predictor of coronary plaque progression.15 Accurate image interpretation on OCT-TCFA, however, requires specific skill, because situations such as macrophage accumulation in plaque can mimic a thin fibrous cap. Macrophages are visualized on OCT as punctate, highly backscattered structures with significant signal attenuation; and thin bands of accumulated macrophages close to the luminal surface can mimic TCFA.28 Another situation is tangential signal dropout. If the imaging beam strikes the tissue at a glancing angle, a signal-poor area with diffuse borders covered by a thin signal-rich layer can be generated that resembles TCFA.28 Although the range of intraobserver and interobserver agreement for a diagnosis of OCT-TCFA was acceptable in the present study, some TCFA might have been over-diagnosed. Thus, plaques diagnosed as OCT-TCFA in the present study do not precisely correspond to histopathological TCFA. Therefore, the present results require confirmation through direct comparisons between iMap-IVUS, OCT and histopathology.
Clinical ImplicationsiMap-IVUS has the capability to exclude vulnerable plaque, therefore iMap-IVUS may be able to rule out high-risk plaque in the prediction of future coronary events. Recently, plaque characteristics including necrotic cores as assessed on VH-IVUS have been reported to be independently predictive for recurrent coronary events in patients with coronary artery disease.29,30 The use of VH-IVUS for the prediction of events has been limited in daily clinical practice. The relatively small number of occurrences of cardiovascular events did not allow us to evaluate whether adding imaging by such an invasive modality to a prognostic model with conventional risk factors would result in improved risk prediction.
Study LimitationsSome limitations of the present study require acknowledgement. This observational study was carried out at a single-center with a small patient cohort and we assessed only target lesions undergoing planned PCI in a single vessel. All coronary arteries cannot be easily assessed on IVUS and OCT because the insertion of imaging catheters into 3 vessels is very time-consuming and the risk of complications is increased. Only patients with stable angina pectoris were included because iMap-IVUS cannot discriminate thrombus (a frequent component of culprit lesions associated with ACS) from fibrotic tissue. Thus, the present results might have been affected by selection bias and cannot be generalized to all patients with coronary artery disease. Time domain-OCT systems were used more frequently for patients with OCT-TCFA than with non-TCFA lesions (59% vs. 39%, P=0.14). The lower axial resolution of time, compared with frequency domain-OCT, might influence evaluation of OCT-TCFA. Finally, corresponding images of time domain- or frequency domain-OCT, and iMap-IVUS might not necessarily be acquired during the same phase of the cardiac cycle, because pullback speed differed among the 3 techniques. Furthermore, the resolution of OCT and IVUS differs. Although considerable care was taken to ensure precise comparative site assessments, we could not deny the possibility of errors in cross-sectional matching among the 3 modalities.
On iMap-IVUS, greater necrotic and lipidic areas and a smaller fibrotic area were predominant components of OCT-TCFA plaque. Larger absolute necrotic area in coronary plaque was closely associated with OCT-TCFA.
This study was partly supported by a JSPS Grant-in-Aid for Young Scientists (B) (24790767) (to S.K.).
