Plasma-derived lipids accumulate in the subintimal space in the initial phase of plaque formation (20). Lipid pools do not only initiate an inflammatory reaction, but also increase biomechanical stress (37). Liquid cholesterol in these pools crystalizes, leading to volume expansion (38). This volume increase destabilizes the atherosclerotic plaque. Moreover, cholesterol crystals can injure the arterial wall and disrupt the plaque (39).

Cholesterol crystals are found in ∼39% of de novo culprit lesions of patients with either ACS or chronic coronary syndrome (CCS), and correlate with high-risk morphological features of culprit lesions (40). Cholesterol crystals are more often observed in culprit lesions in ST-elevation ACS patients compared to non-ST-elevation ACS patients, as is an increase in macrophage accumulation, spotty calcifications, mean lipid arc, thin-cap fibroatheromas (TCFAs) and thrombus (41). This supports the hypothesis that cholesterol crystals increase plaque vulnerability and trigger plaque rupture.

The necrotic core results from cell death and the inability to clear this debris. Macrophages play a major role in this process (42). Hypoxia, lipids and oxidative stress have shown to induce apoptosis in different cell types, including leukocytes and vascular smooth muscle cells (43). These signals trigger DNA fragmentation and expression of cell-surface markers that attract phagocytes (44). As macrophages are also the main phagocytes within the atherosclerotic plaque, effective clearance depends on neighboring cells. This process is called efferocytosis (45). There is no inflammatory reaction associated with apoptosis or efferocytosis, as cellular constituents are phagocytosed instead of released in the surroundings (44, 46). Within lipid-laden foam cells, intracellular cholesterol crystallizes and induces apoptosis (39). Efferocytosis becomes insufficient, which leads to a pool of dead macrophage forming a necrotic core.

A TCFA containing a large necrotic core, infiltrated by a high amount of macrophages is often displayed as “a classical example” of a lesion prone for rupture (47–49). The fibrous cap is defined as a distinctive layer of connective tissue overlying the necrotic core. It consists of smooth muscle cells in an extracellular matrix of collagen, proteoglycans and elastin. The media and adjacent adventitia may be infiltrated by varying degrees of lymphocytes, macrophages and foam cells (50, 51). Historically, fibrous caps with a minimum thickness of <65 µm are considered to be thin, as histopathological analysis showed a cap thickness of <64 µm in 95% of arteries with ruptured plaque (49).

Neovascularization and intraplaque hemorrhage are common phenomenons within atherosclerotic plaques (52, 53). Neovessels are already established in the early phase of the atherosclerotic process and mainly originate from angiogenesis out of the vasa vasorum (53). Such neovessels are thin-walled, more fragile and may function as entrance for erythrocytes, lipids and inflammatory cells (35). As example, neovessel density is more prominent at sites infiltrated by macrophages and lymphocytes (53). Neovessels exhibit inadequate endothelial integrity, making them susceptible for microvascular leakage, which is thought to induce intraplaque hemorrhage (54). Accumulation of erythrocytes is associated with lesion instability and necrotic core expansion (52). Therefore, intraplaque hemorrhage may promote inflammation.

Atherosclerotic calcification is initiated within areas of inflammation (55). Serial in vivo imaging in apoE^−/− mice showed that inflammation precedes osteogenic activity and that the initially formed crystals colocalize with macrophages (56). Proposed mechanisms of calcification include the nucleation of necrotic debris into calcium phosphate crystals, reduced activity of inhibitors of vascular calcification and transdifferentiation of intraplaque vascular smooth muscle cells and circulating hematopoietic stem cells into an osteo-, and chondrogenesis phenotype (55, 57). In turn, calcium phosphate crystals have shown to induce a pro-inflammatory response by macrophages (58). This suggests a positive feedback loop between inflammation and calcification. The calcium phosphate crystals congregate into microcalcifications. These microcalcifications, if present in the fibrous cap, may cause microfractures that could destabilize the atherosclerotic plaque (59). Therefore, presence of microcalcifications may refer to a more vulnerable phase in the progression of atherosclerotic plaque within regions of inflammation. Whereas increasing density is thought to reflect a stabilizing process (55).

Invasive coronary angiography has established itself as a reference standard for the assessment of coronary artery disease (60). It provides a two-dimensional representation of the coronary lumina, by injecting contrast media and performing different radiographic projections, with minimal information on the vessel wall. Coronary angiography is able to identify, albeit suboptimally, the presence of calcification and thrombus. Calcified lesions can be recognized as apparent radiopacities before contrast injection (61). Thrombus can be determined by contrast filling defects and intraluminal lucencies on the “luminogram” (62). The application of other intracoronary diagnostic tools offers the opportunity to look beyond luminal dimensions to identify previously indiscernible lesions.

Thermography was introduced in the early 2000s as an alternative tool to detect coronary artery wall inflammation (63). It was suggested that temperature heterogeneity could identify high-risk lesions. “Hot plaque”, was supposed to reflect the higher metabolic rate of inflammatory cells. However, intracoronary thermography could not meet its expectations (63). In vivo experiments showed that intracoronary thermistors could not detect subtle changes in temperature during substantial influence of pressure, cardiac motion and coronary blood flow (64). Thereafter, it fell in oblivion. Notwithstanding, contemporary imaging techniques do have the ability to target inflammation.

Novel high-definition IVUS may reach an axial resolution of approximately 40–60 µm using high-frequency ultrasound signals (60 MHz) (65). IVUS can differentiate between various plaque components, since calcified plaques are brighter with acoustic shadowing, while lipid-rich plaques appear less echo dense. Furthermore, IVUS can evaluate serial changes in coronary atheroma, to measure for example the effect of statins or PCSK9 inhibitors on atheroma volume (66–69). Using spectral analysis of the “backscattered” ultrasound signals, IVUS offers the opportunity to estimate plaque composition, so-called virtual histology IVUS (VH-IVUS). VH-IVUS has been used to differentiate fibrous plaque, fibrofatty plaque, necrotic core and calcium (Figure 3), which correlates with histologic samples (70). Moreover, serial changes in plaque morphology and pharmacological effects can be identified with VH-IVUS. For example, the “Integrated Biomarker and Imaging Study 2” (IBIS-2) trial showed that darapladib, a “lipoprotein-associated phospholipase A2 inhibitor”, prevented necrotic core expansion after 12 months (71). However, not all studies could confirm the accuracy of VH-IVUS. Necrotic core determined by VH-IVUS did not correlate with histology within a porcine model (72). Therefore, concerns about the validity remain.

IVUS is less suited to detect microcytic or molecular factors of inflammation, such as macrophage accumulation or cholesterol crystals, owing to its relatively low resolution (73). However, high-risk plaque features detected by IVUS seem to be positively correlated with circulating inflammatory biomarkers, reflecting higher inflammatory activity (74–77). Within the “European Collaborative Project on Inflammation and Vascular Wall Remodeling in Atherosclerosis—Intravascular Ultrasound” (ATHEROREMO-IVUS) study, higher plaque burden and TCFA lesions were associated with higher levels of circulating tumor necrosis factor alpha and lower levels of circulating interleukin-10 (74). Furthermore, lesions with a plaque burden of ≥70% or TCFA were independently associated with higher rate of major adverse cardiac events within the same study (78).

In addition to tissue characterization, mechanical stress can affect coronary arteries (79). It is feasible that plaque rupture occurs at a location subject to higher mechanical stress. Mechanical strain refers to the tensile stress caused by the pulsatile intravascular blood pressure, whereas wall shear stress results from the tangential component of shearing deformation from blood flow (79). Mechanical strain can be assessed by using the displacement of radiofrequent IVUS signals at two different intracoronary pressures. This technique is called palpography (80). Patients presenting with ACS have more high strain spots than patients with CCS (80). Furthermore, the number of high strain spots seemed positively correlated with levels of hsCRP (80). Another in vivo study using Yucatan minipigs showed that regions with high strain levels were associated with presence of macrophage (81). However, within “The Providing Regional Observations to Study Predictors of Events in the Coronary Tree” (PROSPECT) trial, no difference was found in strain values between thin- and thick-cap fibroatheroma. They could not confirm the correlation between high strain spots and hsCRP in humans (82). Therefore, the diagnostic value of palpography remains uncertain.

Although smaller in magnitude than mechanical strain, wall shear stress is receiving increasing attention because of its biomechanical relevance. Low wall shear stress acts as a pro-inflammatory and pro-atherogenic stimulus on endothelial cells (79). A three-dimensional reconstruction of the coronary artery lumen is required, which also can be obtained using coronary angiography in combination with IVUS or optical coherence tomography (OCT). Thereafter, wall shear stress maps can be constructed using coronary geometries and computational fluid dynamics calculations (79). Unfortunately, results from in vivo studies on the correlation of wall shear stress and plaque progression remain scarce and conflicting. Both low and high wall shear stress have been associated with atherosclerosis and inflammation (79). Therefore, more clinical studies are needed to explore its use.

IVUS can be combined with near-infrared spectroscopy (NIRS), which projects near-infrared light on the coronary wall, after which the reflected light is analyzed. Since cholesterol has unique features in the wavelength region, NIRS can be applied to characterize lipid-rich plaque, expressed by the lipid core burden index (LCBI) (83, 84). This index is calculated as the number of pixels with a probability of lipid core plaque > 0.6 divided by the total number of pixels and multiplied by 1,000. The MaxLCBI_4mm is often used to detect the presence of a large lipid pool, which is the maximum LCBI value for any 4-mm segment (85). The PROSPECT II study showed that highly lipidic lesions with a MaxLCBI_4mm ≥ 324.7 were independently associated with future cardiac events (86). Moreover, risk of non-culprit major adverse cardiovascular events (MACE) increased significantly for each 100-unit increase in MaxLCBI_4mm in the “Lipid Rich Plaque” (LRP) study (87). NIRS confirms the crucial role lipids fulfil in the development of cardiovascular events and NIRS can be used to assess response to medical therapy, primarily lipid-lowering therapies (88). However, NIRS is unable to identify crystallization of cholesterol and lacks the ability to differentiate between inflamed or non-inflamed lesions. Furthermore, no association between LCBI and inflammatory biomarkers have been found so far (89).

A novel OCT-NIRS catheter is being developed to provide simultaneous microstructural and compositional imaging (Figure 4) (90). The superior resolution and characteristics of OCT could overcome some limitations inherent to IVUS imaging, as discussed in the next paragraph. A first-in-human study using OCT-NIRS is ongoing (NCT05241665).

OCT can provide microstructural images up to a maximal axial resolution of 10 µm at the cost of penetration depth, compared with IVUS (1–2 mm vs. 10 mm). The high resolution not only enables more detailed visualization of calcified nodules, thrombi, TCFAs, plaque erosions and ruptures, but also cholesterol crystal accumulation, macrophages and microvessels (Figure 5) (91, 92).

Macrophages scatter light efficiently, which creates signal-rich regions called bright spots with a cast shadow behind (93). (Figure 5A) While OCT has not been proven to distinguish between active and inactive macrophages (94), bright spots do have a strong correlation with inflammation measured by hsCRP (95). Bright spot density is significantly higher in lipid plaques compared to fibrous plaques and plaques with TCFA show a trend toward higher bright spot density (95). Moreover, bright spot density is also significantly higher in plaques with rupture than those without (95). These findings imply the ability of OCT to provide an overall estimate of macrophage accumulation. Furthermore, presence of OCT-defined macrophage accumulation is associated with adverse clinical outcome. In the “Relationship Between OCT Coronary Plaque Morphology and Clinical Outcome” (CLIMA) study, presence of macrophage accumulation in native left anterior descending coronary artery was associated with more clinical events, especially in the copresence of other high-risk plaque features including a thin fibrous cap and a large lipid arc (96).

Cholesterol crystals appear as thin, linear structures with high signal intensity, often localized nearby lipid-rich plaque (Figure 5C). It is suggested that needle-shaped cholesterol crystals could perforate the fibrous cap, causing plaque instability (93).

Intraplaque neovessels can be identified as well delineated signal-poor voids, which can be followed in consecutive frames (Figure 5D) (93). An ex vivo OCT study showed that coronary atherosclerotic plaques with neovessels were accompanied with greater luminal narrowing (97). Subsequent neovessel rupture could induce intraplaque hemorrhage. However, studies about neovessels and neovessel rupture on OCT imaging remain scarce.

OCT is the only imaging modality with sufficient spatial resolution to adequately measure fibrous cap thickness (Figure 5B). Numerous prospective and retrospective studies have demonstrated an association between OCT-identified TCFA and clinical outcome, whether or not in combination with other features of plaque instability (96, 98). The fibrous cap thickness cutoff to define TCFA differs between studies, as it has been suggested that the 65 um cutoff obtained in histopathological studies should be enlarged to adjust for potential tissue shrinkage during histopathological tissue processing (93). Nevertheless, Jiang et al. found a similar optimal cutoff of 66.7 um to distinguish lesions at higher risk of causing events (98). In this study, 883 patients were included, all 3 main epicardial vessels were scanned and follow-up lasted up to 4 years. OCT can differentiate whether ACS arises from rupture of the fibrous cap or endothelial injury with an intact fibrous cap, i.e. plaque erosion. These distinct patterns might have different underlying pathobiologies. The presence of OCT-identified culprit plaque rupture is associated with lower levels of T-cells but higher levels of effector molecules involved in the innate immune response compared to ACS with intact fibrous caps (99). This may indicate that the adaptive immune system plays an important role in inducing endothelial erosion. In proteomics analysis, patients with ruptured plaques also had a higher inflammatory response and more MACE during 2 years of follow-up (100).

OCT allows to evaluate change in high-risk plaque characteristics (101). The “High-Resolution Assessment of Coronary Plaques in a Global Evolocumab Randomized Study” (HUYGENS) showed that intensive lipid-lowering therapy with high-dose statins and evolocumab increased minimum fibrous cap thickness and decreased the macrophage index on serial OCT. The combination of statin and evolocumab resulted in more favorable changes than statin therapy alone (102).

Near-infrared fluorescence (NIRF) imaging is an emerging technique, allowing the visualization of molecular processes within the atherosclerotic plaque. It uses imaging agents which bind to specific targets, including protease activity, LDL, fibrin deposition and microcalcifications (106–108). These imaging agents consist of an fluorochrome, conjugated to an antibody, molecule or peptide. After injection, a NIRF-catheter is advanced within the coronary artery. A continuous wave laser diode emits excitational light within a 650 to 950 nm window (NIR spectral region) to stimulate the fluorophores. The subsequent fluorescence emission is collected and filtered within the NIRF-catheter (109).

Protease-activatable fluorophores have been developed to visualize enzymatic activity. At baseline, fluorophores emittance is quenched, but increases significantly after cleavage (110). Enzymatically active cathepsins, detected by NIRF, seem to colocalize with cathepsins and macrophages on immunohistochemistry in animal and human atheromata (107, 111, 112). Matrix metalloproteinase-specific fluorophores have shown similar results on NIRF imaging (113, 114). Indocyanine green (ICG) is an imaging agent, which can directly visualize macrophages. After injection, ICG is internalized by macrophages and foam cells by binding to albumin or LDL (108). In a recent study, the ICG NIRF signal, measured in freshly isolated carotid plaques, was highest in the most stenotic area. Subsequent histopathological analyses established that ICG targeted endothelial abnormalities, such as disrupted fibrous caps and areas of neovascularization. ICG concentrated on zones of plaque lipids, macrophages and intraplaque hemorrhage (115). Furthermore, fluorophores have been developed to target fibrin deposition, activated factor XIIIa or thrombin activity to assess thrombosis. Validation of these fluorophores mostly rely on non-invasive NIRF imaging techniques (116–118). However, intravascular NIRF has been able to detect fibrin deposition overlying stent struts in rabbits (119).

Hybridization of intravascular molecular and structural imaging could potentially allow further study of the pathophysiological mechanisms of arterial plaques (120, 121). Both NIRF-OCT and NIRF-IVUS are being developed. The dual-of modality OCT and NIRF, can detect fluorescence from naturally occurring molecules, called near-infrared autofluorescence (NIRAF). NIRAF is elevated in advanced necrotic core-containing lesions and is associated with a high-risk morphological plaque phenotype (Figure 6) (122, 123). Interestingly, NIRAF elevation is specific to plaques with macrophage accumulations, as shown by OCT (122). However, the converse is not true, since many areas with elevated macrophage accumulation on OCT were NIRAF negative. This could be explained by low sensitivity/high specificity of NIRAF to macrophage accumulation, or by the concept of different macrophage phenotypes. The underlying molecular and chemical mechanisms that produce NIRAF are not yet fully understood.