Atherosclerotic plaque components can be categorized by utilizing various optical properties of different tissues (Figure 9.2). Fibrous plaques are identified as homogenous, highly backscattering, low attenuation lesions. The emitted light is absorbed more intensely by lipids which leads to a high level of posterior signal attenuation; plaques appear as low backscatter regions with poor delineation of the external contour. Calcifications are also regions of low backscatter (signal poor), but they let light filter rather than absorb it, so that they maintain well‐delineated external borders. However, differentiation of signal‐poor areas is not always straightforward and calcium can be misinterpreted as lipids, particularly if the calcification is situated deep in the vessel wall.
Figure 9.2 Plaque characterization with optical coherence tomography. (a) Normal coronary anatomy organized in three layers. (b) Lipid‐rich plaque; note possible macrophage accumulations around 3 o’clock. (c) Lipid‐rich plaque in a saphenous vein graft. (d) Fibrous plaque. (e) Calcification extending deep into the vessel wall between 4 and 11 o’clock. Outer border of the calcification cannot be delineated. (f) Calcification between 12 and 4 o’clock. (g) Thrombus protruding into the lumen. (h) Neovascularization (arrows).
An early ex vivo study established a sensitivity and specificity ranging 71–79% and 97–98% for fibrous plaques, 95–96% and 97% for fibro‐calcific plaques, and 90–94% and 90–92% for lipid‐rich plaques with low inter‐observer and intra‐observer variability [18]. Other studies showed less impressive results, with only 45% of lipid‐laden atheromas identified, with higher but still suboptimal identification success in fibro‐calcific and fibrous plaques (68% and 83%, respectively) [19]. Misinterpretation in this study was mainly caused by low OCT signal penetration, which precluded the detection of lipid pools or calcium behind thick fibrous caps and by misclassification of calcium deposits for lipid pools and vice versa [19]. Furthermore, artifacts such as superficial shadowing and tangential signal dropout can produce images with signal‐poor regions covered by a thin signal rich layer mimicking thin‐cap fibroatheromas (TCFA). Rather than relying only on subjective visual interpretations, algorithms based on the optical attenuation coefficient to classify plaques quantitatively have been proposed; however, as yet, these algorithms are not sufficiently robust to be used in the clinical setting [20].
OCT imaging can also demonstrate thrombi as protrusions or floating masses. Red and white thrombi can be identified via the differences in attenuation intensity, with red thrombi showing high attenuation and complete wall shadowing and white thrombi appearing as low attenuation intraluminal masses or layers [21].
Vulnerable plaque assessment
Imaging in acute coronary syndromes (ACS) includes ruptured plaques and histomorphologic features that can be detected by OCT (superficial lipids, fibrous cap thickness as well the presence of macrophages and neovascularization).
A semi‐quantitative definition of the presence of superficial lipids in ≥2 quadrants is used to describe lipid‐rich plaques in OCT studies [21]. However, low penetration depth prevents accurate evaluation of the lipid core thickness. Pathologic studies reported that ruptured plaques harbor a thin fibrous cap being <65 μm in 95% of ruptured plaques [22]. Using OCT, the fibrous cap can be detected as a high signal homogeneous band covering the lipid‐rich core. OCT was shown to provide accurate measurement of fibrous cap thickness (FCT) ex vivo [23]; previous studies demonstrated thinner fibrous caps in patients with ACS [21,24]. A study that aimed to evaluate the relationship between FCT and plaque rupture in vivo found that in 95% of ruptured plaques, the thinnest FCT was <80 μm and so the investigators proposed this value as an alternative in vivo threshold [25].
The progression of atherosclerosis and plaque vulnerability is critically affected by macrophages, identified as high signal regions appearing either distinct or confluent punctate visually. With dedicated software, OCT‐derived indices can be used to identify macrophages [26]. Nonetheless, macrophages should only be considered in the presence of a fibroatheroma, because there have not yet been any studies to confirm macrophages on normal vessel walls or intimal hyperplasia. In addition, high speckle from microcalcifications or cholesterol crystals can also appear similar to macrophages [27].
Plaque neovascularization is considered as a feature of vulnerable plaques. These microvessels are inherently fragile and leaky, giving rise to local extravasation of plasma proteins and erythrocytes [28]. OCT reveals these vessels as small black holes in the atherosclerotic plaque [29]. The presence of these microchannels is associated with vulnerable features such as thin fibrous cap and positive remodeling [30]. In a larger study, microchennels characterized culprit lesions of patients with ACS, and were not present in non‐culprit lesions of patients with ACS or in stable patients [31]. Another study found no difference in the prevalence of microchannels in ACS and non‐ACS patients; however, the closest distance from the lumen to the microchannel was shorter in ACS subjects than in non‐ACS [32].
OCT imaging over time can provide insights on the efficacy of the therapeutic strategies for plaque stabilization. In an initial study, patients on preceding statin therapy were found to have a reduced incidence of ruptured plaques and a trend toward thicker fibrous caps [33]. The influence of statins on fibrous caps was further investigated in a study of 40 patients with previous myocardial infarction. FCT was found to increase in both the statin and control group over time, but more so in the statin group [34]; this was confirmed ensewhere [35]. Atorvastatin therapy at 20 mg/day provided a greater increase in FCT than 5 mg at 18‐month follow‐up [36]. Despite comparable reduction in total cholesterol and low density lipoprotein cholesterol levels with statin therapy in ACS patients, non‐culprit lesions without neovascularization showed greater increase in fibrous cap thickness than lesions with neovascularization at a 6–12 month follow‐up [37]. These important insights into the operative mechanism of statins reveal subtle qualitative rather than gross quantitative arterial wall changes, explaining the reduction of clinical endpoints achieved despite the unchanged angiographic lumen dimensions and minimal volumetric plaque changes by IVUS [38–40]. The dramatic decrease of cholesterol levels obtainable with PCSK9 inhibitors [41,42] showed mild IVUS volumentric changes and inconsistent morphologic changes with virtual histology [43] leading to ongoing OCT serial studies (ClinicalTrials.gov Identifier: HUYGENS clinical trial NCT03570697, PACMAN‐AMI clinical trial NCT03067844).
Acute coronary syndromes: identification of the culprit plaque and distinction rupture/erosion
The unique ability of OCT to identify small plaque rupture and to distinguish rupture and erosion offers a major added value in the characterization of patients with ACS before treatment (Figure 9.3). Johnson et al. described plaque rupture as the presence of a disrupted discontinuous fibrous cap overlying a lipid core. It is typically associated with a vessel wall cavity, result of the embolization of its necrotic core. Plaque rupture can be observed in a variety of clinical scenarios and often thrombi are found on rupture edges. Plaque rupture, TCFA, and red thrombus are more frequent in patients with ST‐elevation myocardial infarction (STEMI). Additionally, the size of the ruptured cavities was greater and an aperture opposite to the flow direction was more frequent in patients with STEMI [44]. OCT also showed differences in the morphology of ruptured plaques in non‐STEMI and asymptomatic coronary artery disease (CAD) [45]. Pathologic studies suggest that ruptures do not necessarily lead to ACS, and ruptures can heal and cause plaque progression. Using OCT, healed plaques can be identified as multiple layers of different optical densities overlying a large
