Atherosclerosis is a chronic inflammatory disease in which changes in wall shear stress (WSS) and LDL infiltration are pivotal factors. WSS refers to the tangential force exerted by blood on the vessel wall as it flows parallel to the direction of blood flow (6). Under physiological conditions, WSS promotes endothelial cell (EC) growth and inhibits the expression of inflammatory factors and inflammation-related pathways (7, 8). The accumulation of lipids and foam cells in the subintima disrupts the fluidity of the vessel wall, resulting in turbulent flow and subsequent alterations in WSS (9). These alterations in WSS have detrimental effects on ECs, including impaired function and compromised integrity of the arterial intima (10, 11).

In the initial phases of subintimal lipid deposition and accumulation of atheromatous substances, compensatory vascular mechanisms prevent luminal narrowing. However, when the growth of atherosclerotic plaques exceeds the compensatory capacity of the vessels, it leads to alterations in hemodynamics within the plaque and vascular structure (12). Protrusion of the plaque into blood vessels results in narrowing of the lumen diameter, disturbing blood flow downstream of the plaque. Consequently, an area characterized by a low WSS and high oscillatory vascular stress is established, promoting the persistent growth of atherosclerotic plaques (13, 14). Conversely, regions with higher WSS may exhibit a correlation with plaque regression. Lan et al. demonstrated a significant association between the total area of the high WSS region encompassing the stenotic lesion and the area of the high WSS region proximal to the lesion with regression of symptomatic intracranial atherosclerotic plaques (15).

Within the subintima, LDL infiltrates and undergoes gradual oxidation as it lacks sufficient protection from antioxidants (16). Oxidized low-density lipoprotein (ox-LDL), formed by LDL in the presence of oxides such as reactive oxygen species, is a strong ligand for macrophage scavenger receptors (cluster of differentiation (CD) 36, SR-AI/II, and SR-BI), contributing to their uptake into macrophages (17). Furthermore, ox-LDL promotes smooth muscle cell migration from the tunica media (via platelet-driven growth factor and basic fibroblast growth factor) to sites of lipid deposition and abnormal proliferation (via insulin-like growth factor 1 and epidermal growth factor), which involves the secretion of extracellular matrix proteins (17, 18). Macrophage-like VSMCs also play a crucial role in facilitating the phagocytic clearance of ox-LDL and contribute to the formation of foam cell-like VSMCs that strongly associate with the lipid core (19). This combination of factors results in the progressive formation and development of a necrotic core of the plaque (Figure 1).

The dynamic equilibrium between plaque regression and progression is influenced by various biological processes, including inflammation, cell death, extracellular matrix degradation, and connective tissue repair responses. Plaque regression usually results from reduced plaque lipid content, LDL levels, and inflammatory status (20, 21). When lipid-lowering therapy reduced LDL levels to less than 70 mg/dL, investigators observed a reduction in the plaque lipid core area and significant plaque regression (22, 23). The classic lipid-lower statins not only inhibit LDL synthesis but may also promote the polarization of M2-like macrophages (24). M2-like macrophages accumulate at the sites of injury and release mediators that suppress inflammation in plaques, clear apoptotic cells, and promote tissue repair (25). It’s worth noting that Faisel et al. discovered that plaque regression in the carotid artery was commonly observed in larger plaques with fewer fibrotic characteristics, which indicate that plaque remission may not necessarily correspond to reduced cardiovascular risk (26).

The risk of different plaque events progressively increases as lipids and atheromatous material accumulate in the subintimal region. Disease progression is significantly influenced by plaque erosion, rupture, thrombosis, neovascularization, and IPH.

Neovascularization is characterized by immature blood vessels with sparse or even absent VSMCs and a lack of tight junctions between ECs, resulting in leakage of blood components from the vessels (27, 28). Hypoxic conditions within the plaque stimulate the release of vascular endothelial growth factor, promoting neovascularization, elevating microvessel density, and disrupting vascular organization (29). Moreover, in advanced atherosclerotic lesions, matrix metalloproteinase (MMP)-9 and plasma myeloperoxidase play a crucial role in promoting oxidation and inflammation, which consequently result in the degradation of the fibrous cap and compromises the stability of plaques. These factors ultimately facilitate the formation of IPH and infiltration of blood cells, which further promote plaque rupture. Zhao et al. reported that IPH was independently associated with a significant increase in lipid core volume (percentage difference in relative lipid core volume change: 50.3%/year, 95% CI: 19.4, 89.2, p < 0.001) and plaques with IPH had a greater decrease in lumen area than plaques without IPH (mean: −0.4 ± 0.9 versus 0.3 ± 1.4 mm²/year, p = 0.033) (30). This suggests that plaques with IPH undergo a transition into vulnerable plaques. After correcting for traditional risk factors, plaque load, and volume of the lipid-rich necrotic core, IPH volume remained significantly associated with the risk of acute ischemic stroke on the ipsilateral side (31). Bos et al. also observed a higher risk of stroke in patients with carotid IPH (32).

Plaque erosion, although milder than plaque rupture, is also associated with plaque stability. It disrupts the structural integrity of the fibrous cap of the plaque, leading to the exposure of its contents to the bloodstream. Consequently, plaque erosion often coincides with thrombotic-like plaque rupture. Upon analyzing OCT images of 115 patients with acute coronary syndrome, Kato et al. found that plaque erosion may already be present in early atherosclerosis (33). Patients with plaque erosion showed less positive vascular remodeling and neovascularization, indicating that plaque erosion is an early warning sign of plaque rupture [(34); Figure 2, left panel].

Disruption of calcium and phosphorus homeostasis occurs in foam and smooth muscle cells within microdomains after apoptosis. In addition, calcification-associated extracellular matrix vesicles aggregate to form foci of microcalcification (35). Moreover, differentiation of VSMCs into calcified cells is intricately linked to inflammatory and oxidative responses. These observations suggest that plaque calcification is a complex outcome arising from multifactorial interactions. The universal consensus is that the degree of calcification reflects the risk of developing atherosclerotic plaques. Dense calcification is generally indicative of stable plaques with a lower risk of adverse vascular events (36). Conversely, multiple small lesions featuring low-density calcification, particularly in proximity to the lipid pool and fibrous cap, signify fewer stable lesions and heightened risk of vascular events. Significant stenosis is observed in advanced atherosclerotic vessels and has been used as a predictor of myocardial infarction (37).

Atherosclerotic risk factors have a broad effect on the arterial system (38). However, plaque formation in the major large vessels is specifically influenced by factors such as hemodynamics (39), morphology (40), and variation in biochemical parameters (41). Atherosclerotic plaques mainly occur in large-and medium-sized arteries, including the carotid, coronary, femoral, and the circle of Willis arteries. Intracranial atherosclerotic plaques are eccentrically distributed in the basilar, anterior, middle, and posterior cerebral arteries and their branches, all of which have a diameter > 3 mm (42). In an observational study, Xu et al. divided the sagittal location of the middle cerebral artery (MCA) into four quadrants based on high-resolution MRI (HR-MRI): superior, inferior, dorsal, and ventral, and found that MCA plaques were more frequently located in the walls of the ventral and inferior sections (43). Sun et al. further discovered that MCA plaques were most common in the proximal M1 segment, whereas MCA plaques causing infarction were mainly distributed in the ventral and superior wall (44). Their investigation also revealed the prevalence of plaques within the distal segment of the basilar artery (BA) (44). In a subsequent analysis involving 86 patients with plaques in the BA, Zheng et al. categorized all cross-sections exhibiting eccentric plaques based on their orientation in the anterior, posterior, or lateral (left or right) center of the vessel. The results of this analysis revealed a higher likelihood of plaque distribution in the posterior wall among patients with pontine infarction [(45); Figure 3]. Patients with intracranial atherosclerosis (ICAS) have more severe strokes and longer hospitalization than those without (46).

Interestingly, ICAS exhibits a higher prevalence in the Eastern population and manifests at an earlier age than in the Western population. This difference may be partly associated with the metabolic syndrome (47). Metabolic syndrome is a chronic noninfectious syndrome characterized by a cluster of vascular risk factors, including insulin resistance, hypertension, abdominal obesity, impaired glucose metabolism, and dyslipidemia (48). Hypertension is more prevalent among Asian populations. Several gene polymorphisms associated with high salt sensitivity have been implicated as causative factors, including genes encoding α-adducin, angiotensinogen, and aldosterone synthase (49, 50). Lifestyle and diet also contribute to these differences. In Asian populations, heightened alcohol consumption and smoking are contributory factors that accelerate the progression of atherosclerosis. Excessive alcohol consumption in individuals with aldehyde dehydrogenase deficiency is associated with elevated blood pressure (49). Conversely, Western diets are closely associated with hypercholesterolemia, which is more strongly correlated with extracranial atherosclerosis than with intracranial atherosclerosis (50). Furthermore, the distribution of body fat and adiponectin levels in Eastern populations differs from those observed in Western countries (47, 51). These factors may be associated with the higher incidence of ICAS in Asian populations.

Circumferential extracranial atherosclerotic plaques rarely form in the carotid artery or its branches. Instead, these plaques tend to exhibit an eccentric distribution, primarily located at the carotid bifurcation and the external and posterior walls of the internal carotid artery (1). In coronary arteries, plaques predominantly develop in the proximal and mid left anterior descending artery, followed by proximal right coronary artery. Additionally, the proximal anterior descending artery exhibits higher plaque calcification compared to the other arteries (52, 53). Femoral artery plaques are primarily localized in the posterior wall of the common and deep femoral arteries (Figure 3). These plaques display lower levels of lipid deposition and inflammation and demonstrate a higher tendency for osteogenesis (54).

Atherosclerotic plaques may be distributed across multiple vessels. In a cross-sectional study of 3,067 adults (aged between 50 and 75 years) in southeastern China, Pan et al. observed that atherosclerotic plaques were predominantly distributed in the aortic and iliac arteries. The presence of plaque in either the aortic or iliofemoral arteries indicated an 85.3% likelihood of the plaque being present in other vascular territories (55). A similar trend was identified by Lambert et al. in people with a low to moderate risk of CVD, in which they divided the arterial tree into 31 segments and analyzed the narrowest part of each segment. They found that 27% of the participants had atherosclerosis in multiple vessels and that the atherosclerosisis relatively evenly distributed throughout the cardiovascular system, such as the abdominal aorta, iliac arteries, subclavian arteries, and femoral arteries (56).

Radiographic risk factors for atherosclerosis include arterial occlusion, the degree of stenosis, and thrombotic lesions. To effectively assess the risk of CVD, researchers have persistently explored innovative examination techniques that provide comprehensive insights into plaque characteristics and vascular hemodynamics (Table 1).

Although the current diagnostic standards and guidelines for coronary artery atherosclerosis are mainly related to geometric parameters derived from two-dimensional images of the coronary arteries, the characteristics of 3D reconstructed plaques have also received increasing attention.

Based on various types of imaging examination, researchers have used computer simulation technology to determine the characteristics of plaques in vitro. Through 3D reconstruction of plaques using carotid ultrasound, Spence et al. quantified the area and volume of plaques and described the relationship between plaque texture algorithms and plaque calcification (96). Another study by the same team confirmed the accuracy of plaque volume measurement (97). The scope of interest within 2D ultrasound is constrained by the movement of the ultrasound probe, thereby posing challenges in achieving clear visualization of the target site. Measurements pertaining to the volume of tissue or lesions are inclined to be less precise, often contingent upon the proficiency of the operator. While 2D ultrasound generates a Doppler report encompassing various hemodynamic measurements, it falls short of delivering a comprehensive depiction of the arterial anatomy along with precise localization details. Conversely, 3D ultrasound offers an intuitive and efficient means to visualize tissues or plaques in three dimensions, alleviating operator fatigue and enhancing diagnostic precision (98, 99).

Currently, 3D imaging of atherosclerotic plaques provides a comprehensive depiction of the arterial lumen, external vessel wall, and calcified plaque (volume, surface area, and maximum length). This approach significantly minimizes the need for invasive methods of examination and their associated impact on patients (100). Guo et al. used 3D HR-MRI to investigate vascular remodeling and plaque morphology in patients with severe vertebrobasilar stenosis (101). In addition, another study compared an HR-MRI-based 3D carotid plaque radiomics model constructed using the radiomic features of 3D T1-SPACE sequences with their contrast-enhanced counterparts using a conventional imaging model. The results revealed that an HR-MRI-based 3D carotid radiomics model exhibited improved accuracy in detecting vulnerable carotid plaques (102). Furthermore, Becher et al. employed Adipo-Clear and immunolabeling in conjunction with light-sheet microscopy to conduct a 3-D reconstruction of mouse cerebral arteries. In contrast to traditional inspection, this approach enables the acquisition of valuable information regarding the plaque shape, total volume, cellular volume, and cell-free volume (103). In intracranial atherosclerosis, the utilization of 2D black-blood techniques for visualizing angulated lesions or intricate intracranial arteries might hinder the attainment of cross-sectional images that are perpendicular to the arterial longitudinal axis. Conversely, the implementation of through-multiplanar 3D reconstruction empowers researchers to examine intracranial plaques from various orientations, boasting a heightened spatial resolution. Moreover, this approach facilitates the capture of cross-sectional images precisely at the location of maximal luminal stenosis, as well as the proximal and distal reference sites. Notably, the 3D technique effectively circumvents tilt artifacts, a proficiency absents in 2D scans that, in turn, could lead to an overestimation of both true wall thickness and vessel area (101, 104).

OCT-based reconstruction methodology and computational fluid dynamics (CFD) simulations are utilized for the computation of local hemodynamic quantities. Migliori et al. utilized time-averaged WSS (TA-WSS) to depict the blood flow characteristics around plaques before and after stent implantation surgery. This technique provides a remarkable and reliable visual image of blood flow around the plaque before and after stenting. An increase in the lumen cross-sectional area downstream of the lesion resulted in significant recirculation of the stream. Post-percutaneous coronary intervention surgery, a smooth lumen surface and well-apposed stent struts facilitated unobstructed blood flow without recirculation (105). This suggests that the combination of 3-D reconstruction and CFD simulation provides a visually compelling assessment of the therapeutic efficacy of stent implantation surgery.

The utilization of 3D reconstruction techniques to forecast plaque progression hinges on the formulation of the model itself. The precision of this endeavor is greatly influenced by the approach to construction and the methodology of correction. Errors that arise possess the potential to reverberate significantly, exerting a profound impact on the ultimate outcomes. While the propagated error in shear stress calculations appears negligible, it assumes a substantial magnitude when it comes to several variables within the plaque growth model. These inaccuracies in prognosticating plaque development could be attributed to several factors: (1) divergences in the reconstructed vessel geometry; (2) alterations in pressure gradients between the epithelial and endothelial boundaries; and (3) disparities between the assumptions regarding the initial concentration of VSMC within the arterial wall and the physiological reality (106).

Integrating various examination techniques and computer simulations, researchers have developed a variety of ‘virtual reality’ techniques to detect atherosclerotic plaques and blood vessels (107).

In 3D models, researchers estimate almost any hemodynamic parameter, including WSS, flow velocity around plaques, and carotid stenosis (107–109). Owing to the complexity of plaque geometry and composition, constructing a 3D plaque model is time consuming. Some researchers have proposed a 1-D/3-D hybrid model that aims to balance computational efficiency without compromising the accuracy of hemodynamic indicators (110). It should also be noted that because of the assumed rigid structural properties of the vessel wall, the WSS obtained using 3D models was higher than the actual value (111). Subsequently, a fluid–structure interaction (FSI) model was proposed to address the above shortcomings.

Using the FSI model, researchers have simulated the effects of atherosclerosis on the vascular and blood flow states, WSS, LDL permeability, vulnerable plaques, and surrounding features (112). Pakraven et al. employed the FSI model to investigate the status of coronary artery ECs and discovered that areas prone to atherosclerosis exhibited at least one of the following three attributes: low time-averaged WSS, high WSS angles, and high longitudinal strain (113). Using the FSI model based on IVUS and OCT, Guo et al. acquired more accurate WSS, stress, and strain. The integration of plaque morphology from OCT and IVUS along with mechanical risk factors from the FSI model yielded the highest sensitivity and specificity for predicting plaque progression (114). The complementary nature of these techniques has substantially enhanced their accuracy in predicting plaque progression and assessing cardiovascular risk.

The attenuation values on CT corresponds to the different plaque components. Some studies suggested attenuation values not exceeding 50 Hounsfield units (HU) for lipid components, whereas others suggested values below 30 or 60 HU (141–143). Although an attenuation threshold over 130 HU is commonly used to indicate calcification on two-dimensional CT, there are discrepancies in 3D reconstruction studies, where some suggest thresholds of over 300 or 400 HU for calcification (144). In 2015, Puchner et al. suggested a threshold of >180 HU for calcified plaques (145); in contrast, in 2018, Kigka et al. indicated that attenuation values over 400 HU should be considered for calcified plaques (146). Limited by the resolution of CT and polymorphisms in plaque contents, quantitative differentiation of plaque components by attenuation values is difficult. Thus, the evaluation threshold for calcified and non-calcified plaques (mixed plaques) varies among clinical studies. According to the prevailing consensus, high-density calcification is generally defined as an HU density exceeding 350, while HU values ranging from 131 to 350 indicate fibrous plaques, and a range of 31 to 130 HU corresponds to fibrofatty plaques. Additionally, HU values ranging from-30 to 30 are associated with the presence of a necrotic core (147, 148). Notably, low-density plaques (HU < 30) have been identified as the most robust predictors of myocardial infarction in patients presenting with stable chest pain (149). Coronary artery calcium (CAC) scores obtained by CT are associated with myocardial infarction risk and are independent risk factors for cardiovascular events. Liu et al. examined 19 patients with symptomatic stenosis caused by left coronary plaques and found a positive correlation between the number and volume of coronary plaques and calcification (150). Yoon et al. also showed a positive correlation between the plaque calcification score and degree of stenosis in 66 patients (151). Notably, CAC scores have been employed in guidelines as a surrogate measure for estimating the 10-year risk of cardiovascular events and for making lipid-lowering therapy decisions. Patients with CAC scores >400 demonstrated a significantly higher likelihood of developing CVD (147, 149, 152).

The thickness of the thin fibrous cap is defined as less than 65 μm, and this criterion is most commonly used to assess plaque stability (153, 154). Owing to artifacts such as edge blur and halo effects, fibrous caps are difficult to image using CT. Recently, a study demonstrated the effectiveness of spectral photon-counting CT in quantifying fibrous cap thickness, plaque area, and lipid-rich necrotic core area, showing good consistency with histopathological measurements (attenuation values in voxels distinguish plaque regions; the thickness of the regions is measured by multiplying the number and size of voxels) (63). This new technology offers promising possibilities for precise quantitative analysis of lipid cores and fibrous caps in plaques. In a preclinical study, spectral photon-counting CT identified monocyte accumulation in the arterial wall by recognizing a tracer reflecting the progression of atherosclerosis (155). The center frequency of IVUS imaging should be higher than 70 MHz to identify thin fibrous cap, which is a challenge for current IVUS imaging systems (156). In contrast, HR-MRI demonstrates the capacity to visualize the fibrous cap effectively. Instances of fibrous cap rupture are identifiable by the presence of partially obscured and irregular surfaces within T₁-weighted or T₂-weighted images (157).

The ability of CT to predict IPH remains unclear. Some studies have shown no significant difference in attenuation values between plaques with and without IPH on CT (158). However, Saba et al. found a statistical correlation between IPH and low HU values (HU values <25 after contrast medium administration, indicating the presence of IPH with a sensitivity and specificity of 93.22 and 92.73%, respectively) (159). It is worth stating that the difference may be explained by the presence of a lipid-rich necrotic core (61). HR-MRI effectively addresses this gap by demonstrating excellent utility. In the context of extracranial atherosclerosis, HR-MRI’s capability to identify IPH exhibits commendable intersubject reproducibility and reliability. Notably, HR-MRI displays a robust sensitivity range (81–90%) coupled with a high specificity range (74–90%) in the discrimination of IPH, fibrous cap, and lipid cores (160). These capabilities are further bolstered when employing contrast-enhanced HR-MRI. In observational studies focusing on basilar and internal carotid artery plaques indicate that IPH manifests as a signal intensity surpassing 150% of adjacent gray matter’s signal intensity in T₁-weighted images (80, 157).

In the past decade, mathematical modeling and simulation have emerged as valuable noninvasive tools in the field of cardiovascular disease, aiding both basic scientific research and clinical decision-making processes. Specifically, CFD plays a crucial role in identifying hemodynamic alterations. CFD models were meticulously calibrated to ensure accurate representation by leveraging diverse examination data and various parameters.

CFD simulations utilize clinical imaging data, such as results of ultrasound, OCT, CT, and MRI, to derive patient-specific estimates of crucial hemodynamic parameters, including the flow rate, pressure, fractional flow reserve, and WSS (161–163). After microcirculatory disturbances in the artery, there is a decrease in blood flow and translesional pressure drop, accompanied by an increase in the fractional flow reserve without significantly affecting the non-culprit branches (161). Nonetheless, severe stenosis in the large branches results in elevated distal microcirculatory resistance, significantly increasing the flow velocity and instantaneous wave-free ratio in the cognate branches (164). Furthermore, CFD has been employed to evaluate the impact of stent implantation on peripheral flow status. The WSS and LDL filtration rates exhibit notable variations depending on stent geometry and are closely associated with the final treatment outcome and occurrence of stent restenosis, which were previously challenging to assess in vivo (136). In static CFD simulation, the removal of side branches with a radius of less than 50% of the parent vessel has a negligible effect on the accuracy of the fractional flow measurement. Similarly, in transient CFD simulation, the impact remains minor. This observation contributes to the simplification of CFD model construction and adjusts for the effects of geometric variations of the surrounding arteries (165). CFD techniques have been extensively applied in atherosclerosis studies owing to their practicality and effectiveness. This has led to the derivation of numerous significant parameters or metrics, enabling a more visual assessment of CVD.

The WSS is the frictional force exerted by flowing blood on the vessel wall (6). TA-WSS and wall shear stress gradient (WSSG) are vital indicators reflecting the local hemodynamic state. Compared with static simulations, TA-WSS can reflect the average changes in WSS across cardiac cycles (105). Studies have shown that regions with high TA-WSS and low oscillatory shear index have larger necrotic core sizes, larger macrophage areas, and thinner fibrous caps in atherosclerotic plaques (166). Moreover, local TA-WSS is significantly higher in patients with ischemic stroke or transient ischemic attack than in patients with asymptomatic carotid artery stenosis, indicating that TA-WSS may help stratify the risk of vulnerable carotid plaques (167).

The WSSG quantifies the magnitude of changes in the WSS along blood vessels. Under normal physiological conditions, a minimal gradient of shear stress is exerted by blood flow on the inner vessel wall (i.e., low WSSG) (168). Changes in the flow state occur when the blood vessel branches or narrows, resulting in elevated WSS and subsequent a high WSSG (169). An increase in WSSG levels is closely related to EC damage, heightened vascular permeability, and inflammatory infiltration, thus having a role in the formation and progression of atherosclerosis. In addition, WSSG is widely used in predicting arterial aneurysms. A high WSS influences the formation of MCA aneurysms, and positive WSSG promote this progression (170, 171).

Unlike the WSS, the axial plaque stress (APS) is calculated by isolating the axial component of the force exerted on the lumen or the plaque. The APS provides insights into the stress experienced by the plaque along the axis of the blood vessel and is affected by factors such as plaque shape, core stiffness, length of the lipid core, severity of the plaque, and vascular remodeling. Notably, in some studies the APS was significantly higher than the WSS (172–174). By constructing a CFD model of coronary artery blood flow, researchers discovered a linear correlation between upstream APS variation and the severity of vascular stenosis (173, 174). Alegre-Martinez et al. further elucidated that APS affects the risk of atherosclerotic plaque rupture and is related to the predilection for specific sites of plaque rupture (172). Another study reported that APS is an independent predictor of high-risk plaques and an independent risk factor for acute coronary syndrome (175).

Translesional pressure gradients are important indicators of arterial hemodynamics. Changes in blood flow in front of and behind the atherosclerotic plaques are linked to the risk of plaque rupture. Studies use the translesional pressure ratio (PR = Pressure_{post-stenotic}/Pressure_pre-stenotic) as a surrogate indicator to describe translesional pressure gradient, with PR ≤ median defined as low PR, indicating a larger translesional pressure gradient (176, 177). Through the construction of CFD models, Leng et al. found that a low PR was an independent risk factor for recurrent stroke in patients with sICAS (177). Another study showed that lower systolic blood pressure may be associated with an increased risk of stroke recurrence in patients with larger translesional pressure gradients (176). More studies are warranted to clarify the relationship between translesional pressure gradient and atherosclerosis-related diseases to guide better prevention and interventions.

In 2007, Cancel et al. demonstrated in vitro that under convective conditions, leaky junctions of ECs were the main pathway for LDL uptake by arteries (>90%) (178). Low WSS reduces the expression of microfilament bundles and disrupts the barrier function of ECs, causing leaky junctions between ECs and subsequently resulting in LDL infiltration into the subendothelium (11). In 2018, a study showed that the concentration of the LDL boundary layer increased when recirculation occurred near the wall and that hypertension intensified this effect by promoting the entrapment of a higher number of LDL particles (179). This may provide a foundation for LDL infiltration into the subintima of the arteries. Roustaei et al. further explained that the reason for the increased LDL infiltration rate in hypertension is the effect of reduced WSS on the number of leaky junctions and the impact of FSI on the widening of endothelial pores (180). This revealed the profound impact of hemodynamics on lipid trans-wall transport and EC function, verifying an important basis for the development of atherosclerosis (Figure 2, right panel).

Nanoparticles serve as transport media and carriers for diverse targeting substances. Researchers have used specially labeled nanoparticles to precisely track atherosclerotic lesions and increase the signal intensity of different imaging modalities (181). For example, anti-CD68 receptor-targeted Fe-doped hollow silica nanoparticles and iron oxide nanoparticles have been used to identify neovascularization and macrophages within plaques to evaluate plaque status (182, 183). Using nanoparticles as tracers, Hossain et al. analyzed blood flow and vascular deposition of circulating nanoparticles that recognize vascular cell adhesion molecule 1, E-selectin, and intercellular adhesion molecule 1 in the femoral artery of patients with peripheral artery disease (184). Furthermore, nanoparticles have the potential to provide precise treatment for atherosclerosis. Synthetic HDL-mediated targeted delivery of liver X-receptor agonists promotes cholesterol efflux from macrophages (185). In addition, CGS 25966 and CGS 27023A are N-sulfonamidoacetyl amino esters that non-selectively inhibit MMP-1, MMP-2, MMP-3, and MMP-9 by chelating zinc ions at the active site of the enzyme, rendering them potential therapeutic targets (186).

Abnormalities in inflammation and lipid metabolism predict the risk of atherosclerosis. Plaque inflammation and its related metabolic markers are valuable indicators of plaque stability.

Growth differentiation factor 15 (GDF-15) is a member of the transforming growth factor β superfamily, which is closely associated with lipid metabolism and inflammation. During the initial phase of atherosclerosis, researchers have observed smaller atherosclerotic lesions in GDF-15^−/− mice; however, this difference disappeared after 12 weeks (187). In a recent cross-sectional study, Shima et al. discovered that high-sensitivity C-reactive protein and GDF-15 levels were significantly higher in patients with coronary artery disease (p = 0.091 and p < 0.001, respectively) (188). Heduschke et al. pointed out that recombinant GDF-15 promotes macrophage autophagy activity and GDF-15^−/− mice show reduced macrophage autophagy activity in plaques, which may be related to plaque regression and stability (189). Ackermann et al. further demonstrated that silencing GDF-15 in human macrophages inhibits oxidized low-density lipoprotein-induced lipid accumulation (190). Overall, the role of GDF-15 in the influence of inflammation on atherosclerosis appears to be multifaceted, suggesting that GDF-15 may serve as a predictor of plaque stability (191).

Neopterin, an oxidation product of 7,8-dihydroneopterin, is produced by activated macrophages stimulated by interferon-γ released from T lymphocytes (192, 193). Co-cultivation of ox-LDL with macrophages promotes the conversion of 7,8-dihydroneopterin to neopterin, highlighting its involvement in facilitating the clearance of ox-LDL (193). Shirai et al. showed that neopterin inhibits foam cell formation, migration, and proliferation of VSMCs. Moreover, neopterin suppresses the phosphorylation of nuclear factor kappa-B transcription factor in human arterial macrophages and increases the expression of peroxisome proliferator-activated receptor γ (194). Clinical studies have shown that patients with coronary heart disease have significantly higher serum neopterin levels. In one study neopterin concentration was positively correlated with the degree of coronary artery stenosis (195). Sugioka et al. also observed that neopterin levels were significantly higher in patients with complex carotid atherosclerotic plaques than in those with non-complex plaques (196). The observed elevation in neopterin protein expression within atherosclerotic plaques is likely attributable to the endogenous upregulation of neopterin proteins, which serve as a defensive response against the progression of atherosclerosis (194).

Other novel biological markers, such as galectin-3, may serve as surrogates to reflect plaque inflammation and calcification (197, 198), additionally pregnancy-associated plasma protein-A has been associated with vulnerable plaque features in patients with coronary artery disease (199), and the count of CD16⁺ monocytes has shown a correlation with preclinical CVD (200).

Atherosclerosis is a multifactorial disease influenced by genetic, environmental, and pathophysiological factors. Multiple genes and ncRNAs regulate lipid metabolism, inflammation, and endothelial and smooth muscle cell function. Gene expression analysis can reveal the dynamic state of a disease and provide insight into its potential causes. Meng et al. compared gene expression differences between healthy individuals and patients with atherosclerosis and found that TPM2 was significantly downregulated in atherosclerotic samples (201). In another study, ITGAX, CCR1, IL1RN, CXCL10, CD163, and MMP-9 were found to be significantly upregulated in atherosclerotic samples (202). Several genes participate in the various processes associated with atherosclerosis. For instance, SVEP1 induces the proliferation of vascular smooth muscle cells (VSMCs), elevates integrin levels, and triggers plaque inflammation, thereby contributing to atherosclerosis development irrespective of blood lipid levels (203). In contrast, JCAD governs the Hippo/YAP/TAZ pathway and mediates endothelial dysfunction, thus fostering atherosclerosis (204). These genes have the potential to serve as valuable risk indicators and treatment targets in individuals with atherosclerosis.

miRNAs are a class of endogenous small single-stranded non-coding RNAs approximately 22 nucleotides in length that regulate post-transcriptional gene expression by degrading target mRNAs or blocking their translation (205). miRNAs play important roles in regulating pathological processes such as cell adhesion, proliferation, lipid uptake, efflux, and the production of inflammatory mediators (206). The upregulation of miR-9 suppresses the formation of vulnerable atherosclerotic plaques through negative regulation of the p38MAPK pathway via OLR1 and enhances vascular remodeling in mice with acute coronary syndrome (207). Moreover, miR-9 has similar functions of inhibiting SDC2 and the FAK/ERK signaling pathway, reducing the plaque area in aortic atherosclerosis (208). Furthermore, the upregulation of miR-9 decreased the levels of tumor necrosis factor-α, IL-6, and IL-1β (207, 208). MiR-181a-5p and miR-181a-3p collectively reduce the expression of pro-inflammatory cytokines, decrease macrophage, leukocyte, and lymphocyte infiltration, and block nuclear factor kappa B activation and vascular inflammation by targeting TAB2 and NEMO (209). In contrast, overexpression of miRNA-155 promotes the activation of the nucleotide-binding oligomerization domain-like receptor protein 3 inflammasome induced by ox-LDL, exacerbating atherosclerosis in ApoE^−/− mice (210). Clinical studies have shown that compared with carotid atherosclerosis patients with stable plaques, miR-223 and miR-126 are downregulated in patients with unstable plaques (211). Besides, MiR-21 is highly positively correlated with the maximum lipid core area, number of lesion vessels, number of macrophages, and number of vulnerable plaques in patients with acute coronary syndrome and negatively correlated with fibrous cap thickness (212). These findings indicate the potential use of miRNAs as predictive indicators of plaque stability.

In addition to miRNAs, other non-coding RNAs, such as long non-coding RNA and circular RNA, have garnered growing attention owing to their involvement in atherosclerosis (205). Furthermore, single mutations, small insertions/deletions, and copy number variants in genes related to lipid metabolism are also associated with atherosclerosis (213). The regulation of genes related to lipid metabolism, vascular inflammation, and EC function is influenced not only by genetic factors, but also by changes in DNA methylation caused by environmental factors. These epigenetic changes may contribute to atherosclerosis (214).