Introduction
Outcomes after percutaneous coronary intervention (PCI) have improved significantly over the past few decades, in large part because of newer stent platforms, better implantation techniques, and improved adjunctive medical therapy. However, the risk of recurrent cardiac events after successful PCI remains high, and the culprit site for these events often differs from that of the index procedure. Traditional clinical risk models, such as the Framingham risk score, are useful for predicting overall risk of atherosclerosis and coronary events. However, clinical risk indicators do not provide anatomic information, and angiographic parameters have poor predictive accuracy in identifying the specific high-risk lesions responsible for future coronary events. Early angiographic studies suggested that the culprit lesions for these recurrent events were often only mildly stenotic at the time of index angiography, whereas pathological studies of patients with fatal cardiac events found that most acute occlusive thromboses occurred at sites with a large plaque burden and severe luminal narrowing. Therefore, a significant amount of research has been devoted to exploring the risk factors for individual plaque progression and thrombosis. In this chapter, we explore the concept of plaque vulnerability as introduced by pathological studies (see Chapter 3 ), and the contribution of intravascular imaging, in particular, optical coherence tomography (OCT), toward extending our understanding of vulnerable plaques and facilitating their identification in vivo. The use of cardiac computed tomography and magnetic resonance imaging to characterize plaque composition in large arteries is discussed in Chapter 9 and Chapter 33 . Molecular imaging applications are addressed in Chapter 32 .
Definition of Vulnerable Plaque
Vulnerable plaques have been defined as those at high risk for evolving into culprit lesions, including both plaques that are prone to provoking thrombosis and plaques at risk for rapid progression. The framework for studying vulnerable plaque features was established by data from autopsy studies that examined the culprit lesions of patients with sudden coronary death (see also Chapter 3 ). These studies revealed three main patterns observed in thrombotic culprit lesions: plaque rupture, plaque erosion, and calcified nodules. Ruptured plaques are the most common, accounting for approximately 60% of cases, and are characterized by fibrous cap disruption with overlying thrombus that is in continuity with an underlying necrotic core. In contrast, eroded plaques are found in approximately 35% of cases and are characterized pathologically by coronary thrombus with an intact fibrous cap. The endothelial lining is commonly absent, exposing the intima, which is primarily composed of smooth muscle cells and proteoglycan. Calcified nodules include approximately 5% of cases and are identified pathologically by fibrous cap disruption and thrombus overlying a fractured calcified plate.
Pathological studies are limited to the retrospective identification of features prevalent in culprit plaques, which are assumed to contribute to plaque vulnerability. However, the introduction and development of intravascular imaging modalities has enabled the study of these same features in vivo and prospectively, and multiple studies have sought to validate and extend the vulnerable plaque hypotheses generated by pathological studies. Although each of the intravascular imaging modalities has unique advantages and disadvantages, they can be broadly grouped into modalities that provide primarily anatomic or compositional information, such as OCT, intravascular ultrasound (IVUS), near-infrared spectroscopy (NIRS), and intravascular magnetic resonance imaging (IV-MRI), and modalities that provide functional or biomechanical information, such as thermography, elastography and/or palpography, and near-infrared fluorescence (NIRF) imaging. In addition, some modalities that provide primarily anatomic information may also indirectly facilitate functional assessment; for example, the identification of macrophages may serve as a marker of local inflammation. However, none of these imaging modalities independently provides physiologic information on coronary hemodynamics, which is better evaluated with techniques such as fractional flow reserve (see Chapter 17 ).
Intravascular Assessment of Coronary Plaque Anatomy or Composition
Optical Coherence Tomography
OCT is a high-resolution intravascular imaging modality that enables the detailed characterization of plaque morphology in vivo. In a manner analogous to the use of sound waves in ultrasonography, OCT technology measures the magnitude and time delay of backscattered light waves to produce images resembling an “optical biopsy,” with 15- to 20-μm resolution. This represents a 10-fold improvement in resolution compared with IVUS and has allowed for the in vivo differentiation of fibrous, lipid, and calcific plaques ( Figure 10-1 ). In addition, OCT can be used in the detection of intraluminal thrombus and the identification of plaque rupture, plaque erosion, and calcified nodules ( Figure 10-2 ). The OCT definitions for these thrombosis mechanisms are based on the pathological definitions, with slight modifications to account for the possibility that luminal thrombus may dissolve or embolize distally, precluding its identification on in vivo imaging. Plaque rupture is identified on OCT imaging as lipid plaque with fibrous cap discontinuity and cavity formation inside the plaque. Because the coronary endothelial lining remains below the resolution of OCT imaging, OCT-defined plaque erosion is confirmed by the presence of attached thrombus overlying an intact and visualized plaque. Probable plaque erosion is identified by attached thrombus in the absence of underlying plaque or neighboring superficial lipid or calcium, or if there is culprit site luminal irregularity without an attached thrombus. Calcified nodules appear on OCT imaging as sites with fibrous cap disruption and underlying plaque characterized by protruding calcification, superficial calcium, and significant calcium adjacent to the lesion.
The high resolution of OCT imaging has enabled the identification of many features suggested by pathological studies to be associated with plaque vulnerability. Because of the high prevalence of plaque rupture as the underlying mechanism for coronary thrombus formation, the features and risk factors associated with plaque vulnerability for rupture are the most commonly studied, and therefore, are the best understood. These features include large extracellular lipid pools, thin fibrous caps, small calcifications, macrophage accumulation, microchannels, and cholesterol crystals ( Figure 10-3 ). However, these characteristics may differ from the features and risk factors associated with plaque erosion and calcified nodules.
Extracellular Lipid Pools
Autopsy studies have identified large lipid cores as a common feature of culprit plaques in patients with sudden coronary death. On OCT imaging, lipid appears as a homogeneous area with low-signal intensity (dark) and high-signal attenuation (significant shadowing) in contrast to fibrous tissue that appears as a homogeneous area with high-signal intensity (bright) and low-signal attenuation (no significant shadowing). Lipid-rich plaques are defined as those with a lipid arc of more than 90 degrees on cross-sectional imaging (see Figure 10-3A ). OCT studies have shown that lipid-rich plaques are more prevalent in the culprit lesions of patients with acute unstable presentations such as ST-elevation myocardial infarction (STEMI) or non–ST-elevation MI (NSTEMI) compared with those presenting with stable angina pectoris (SAP). Furthermore, lipid pools are associated with plaque progression. In a study of 53 patients with 69 nonculprit plaques (<50% luminal stenosis on angiography), OCT-identified lipid pools were significantly more prevalent in lesions that progressed on angiographic follow-up performed between 6 and 9 months later compared with lesions that did not progress (100% vs. 61%; P = .02).
Thin-Cap Fibroatheroma
Thin-cap fibroatheroma (TCFA) are postulated to represent the vulnerable precursor lesion for plaque rupture because of their morphological similarity (see also Chapter 3 ). On pathological examination, most ruptured coronary plaques have fibrous caps measuring less than 65 μm, and therefore, this threshold has been used to define thick- versus thin-cap fibroatheroma. The sharp contrast in appearance between lipid and fibrous tissue on OCT imaging, coupled with its high resolution, make it an ideal intravascular imaging modality for measuring fibrous cap thickness, thereby identifying TCFA (see Figure 10-3B ). Multiple OCT studies have shown a higher prevalence of TCFA at culprit sites in patients with acute or unstable clinical presentations compared with those with SAP. Initially, nonculprit plaques with TCFA are more likely to show progression on angiographic follow-up than those without TCFA. Moreover, statin therapy has been shown to increase fibrous cap thickness, suggesting that one of the mechanisms underlying the clinical benefit of statins is the stabilization of vulnerable TCFA plaques ( Figure 10-4 ).
Calcifications
Coronary artery calcium score, as assessed using cardiac computed tomography, has been shown to correlate with total atherosclerotic burden and risk for future events. However, biomechanical models and pathological studies suggest that the pattern of vascular calcification may be a more important determinant of local plaque vulnerability than the total burden of calcium. On OCT imaging, calcium deposits appear as heterogeneous areas of high- and low-signal intensity with low-signal attenuation and a sharp demarcating border. Spotty calcium deposits are defined as small calcifications with an arc ≤90 degrees on cross-sectional imaging (see Figure 10-3C ). In contrast to ultrasound signals, which are highly attenuated by calcium, the light waves used in OCT imaging are able to penetrate calcium, thereby allowing for more detailed characterization of calcium deposits and better visualization of structures deep to those deposits than is possible with IVUS. In a study of 189 patients with coronary artery disease who underwent OCT imaging of culprit lesions, the number of spotty calcium deposits was significantly greater in patients presenting with acute MI and unstable angina (UA) compared with those presenting with SAP. In addition, these calcium deposits were more superficial in location in the MI and UA groups than in the SAP groups. Although all imaged lesions in this study were culprit lesions, plaque rupture as an underlying mechanism correlated positively with the number of spotty calcium deposits and inversely with the number of large calcium deposits. Taken together, these results support the mechanistic hypothesis that small calcifications can increase plaque vulnerability for rupture.
Macrophages/Bright Spots
Inflammation plays an important role in the development of atherosclerosis. Pathological studies suggest that macrophages can also promote plaque vulnerability, because lipid-laden macrophages have been shown to constitutively produce extracellular matrix–degrading enzymes (see Chapter 3 ). Macrophage accumulations have been identified on OCT imaging as a linear series of signal-rich (bright) spots with high-signal attenuation (see Figure 10-3D ). The original technique used to objectively identify macrophages on OCT images, termed “normalized standard deviation,” has recently been called into question because of methodological problems. In response, a new algorithm has been developed to more objectively identify bright spots within an OCT image, correcting for differences in signal intensity caused by tissue depth, distance from the catheter, and signal-to-noise ratio. In a pathological validation study, OCT bright spots identified using this algorithm were poorly specific for macrophages, which were present in only 57% of bright spot–positive regions. Instead, OCT bright spots were correlated with the interface of plaque components with different optical indexes of refraction, including not only macrophages, but also cholesterol clefts in necrotic cores and the interfaces between old and new fibrous tissue, calcium and lipid or fibrous tissue, fibrous cap and lipid pools, or neovascularization and the media. In addition, not all macrophage accumulations identified histologically were visible on OCT imaging as bright spots, which may relate to differences in the backscattering properties of the different types of macrophages present (for example, M1 vs. M2 macrophages). Therefore, OCT bright spot density may be more reflective of plaque complexity than specific macrophage accumulation.
Microchannels/Neovascularization
Mechanistically, the development of microchannels (neovascularization) within a plaque is believed to facilitate inflammatory cell infiltration and necrotic core formation. These factors, along with a greater risk of intraplaque hemorrhage, may contribute to more rapid plaque progression. On OCT imaging, microchannels are identified as small black holes or tubes measuring 50 to 300 μm in diameter and spanning at least three consecutive cross-sectional frames on pull-back imaging (see Figure 10-3E ). In one study (n = 53), OCT-identified microchannels were more prevalent in nonculprit plaques that showed progression on follow-up angiography than those without progression (77% vs. 14%; P <.01). In another OCT study, among culprit lesions in patients with UA, plaques with microchannels had significantly thinner fibrous caps, greater lipid arcs, longer lipid lengths, and a greater prevalence of TCFA than those without microchannels. Plaques with microchannels have been shown to respond less favorably to statin therapy in terms of fibrous cap thickening compared with plaques without microchannels, despite comparable reductions in serum cholesterol levels. Moreover, OCT-identified microchannels have also been associated with coronary endothelial dysfunction. In a study of patients (n = 40) with early coronary artery disease, endothelial function testing with acetylcholine revealed significantly worse function in segments with microchannels compared with segments without microchannels. In addition, the segments with both OCT-identified microchannels and macrophage accumulation had even more severe endothelial dysfunction, compared with segments with only microchannels or only macrophage accumulation. The authors speculated that there was an incremental effect of inflammation and intimal neovascularization on atherogenesis, in which intimal neovascularization increased vascular wall blood flow and facilitated the penetration of inflammatory cells into the developing plaque, and activated macrophages to promote further angiogenesis and additional macrophage recruitment.
Cholesterol Crystals
Cholesterol crystal content has been shown in pathological studies to be an independent predictor of thrombus formation and clinical events. Volume expansion during cholesterol crystallization is hypothesized to cause disruption and perforation of neighboring fibrous tissue, thereby contributing to plaque vulnerability for rupture. On OCT imaging, cholesterol crystals appear as thin, linear, signal-rich structures with low-signal attenuation (see Figure 10-3F ). Although few OCT studies have directly investigated the potential role of cholesterol crystals in plaque vulnerability, the presence of OCT-identified cholesterol crystals has been associated with a greater prevalence of other vulnerable plaque features, including lipid-rich plaque, spotty calcifications, and microchannels.
Limitations of Optical Coherence Tomography
One of the limitations of OCT imaging is the need to image through a blood-free field, which is achieved by flushing the vessel with Ringer lactate, saline, or contrast during pull-back imaging. Although improvements in image acquisition speed have eliminated the need for proximal balloon occlusion during this process, it nonetheless remains difficult to evaluate ostial lesions using OCT because of image artifacts created by residual blood in the ostial vessel lumen. In addition, the tissue penetration depth of OCT imaging is limited to approximately 3 mm, compared with the 8- to 10-mm depth of IVUS imaging. Penetration depth is important for the evaluation of plaque burden and arterial remodeling, because these parameters require measurement of the vessel area delineated by the external elastic membrane (see Chapter 3 ). Plaque area is calculated as the difference between vessel and lumen areas, and plaque burden is calculated as the proportion of vessel area composed of plaque. In addition, the direction and degree of arterial remodeling is determined by comparing the area bounded by the external elastic membrane at the culprit site with that of reference vessel segments. Because of the limited penetration of the OCT signal and its attenuation by lipid, OCT has not previously been used to assess plaque burden or remodeling. However, a recent study has explored the feasibility of measuring vessel area using OCT imaging of eccentric lipid-rich plaques by using the portion of the vessel circumference that is visible to extrapolate the remaining vessel contour that is not clearly identified because of excessive depth or shadowing by overlying lipid.
Intravascular Ultrasonography
IVUS is an older, and therefore, a more extensively studied and widely available catheter-based imaging modality than OCT. Although IVUS produces images with significantly lower resolution compared with OCT, it does not require clearance of blood from the lumen and can therefore be used to evaluate ostial coronary segments. In addition, IVUS remains the most robust intravascular imaging modality for the assessment of plaque burden and arterial remodeling.
On gray-scale IVUS, echolucent plaque regions have been correlated with the presence of a lipid-rich core, and the combination of echolucency and high-signal attenuation in the absence of bright calcium has been defined as “attenuated plaque.” In addition to positive remodeling, the presence of attenuated plaque is associated with increased plaque vulnerability. Moreover, calcifications are readily identified on IVUS imaging as bright, signal-rich structures with high-signal attenuation, and the presence of small “spotty calcium” deposits have been more frequently identified in patients with unstable coronary presentations compared with those with SAP.
The spatial resolution of gray-scale IVUS imaging is 100 to 200 μm, which is insufficient for the evaluation of TCFA, macrophages, microchannels, or cholesterol crystals. To facilitate more detailed plaque assessment, algorithms have been developed to perform spectral analysis of the raw IVUS radiofrequency backscatter signal, transforming this information into a color-coded representation of plaque composition. The most widely used method is virtual histology (VH)-IVUS, which depicts fibrous tissue as green, fibrofatty tissue as light green, the necrotic core as red, and dense calcium as white ( Figure 10-5 ). VH-IVUS–identified TCFA have been defined as necrotic core-rich (≥10%) plaques, without evident overlying fibrous tissue, and with plaque volume ≥40% seen on at least three consecutive frames.
