Identification and treatment of the vulnerable coronary plaque

Acute coronary syndrome mostly arises from rupture or erosion of a vulnerable plaque. Vulnerable plaques typically appear as lipidrich plaques with a thin cap, called thin-cap fibroatheromas. Various intracoronary imaging techniques can be used to detect vulnerable plaques, such as intravascular ultrasound (IVUS), optical coherence tomography (OCT) and near-infrared spectroscopy (NIRS), each visualizing different high-risk plaque characteristics. IVUS and its post-processing techniques, such as virtual histology IVUS, can primarily be used to identify calcified and soft plaques, while OCT is also able to quantitatively measure the cap thickness. The addition of NIRS allows the exact measurement of lipid content in the plaque. Non-invasive imaging techniques to identify vulnerable plaques, such as computed tomography, are less often used but are evolving and may be of additional diagnostic use, especially when prophylactic treatments for vulnerable plaques are further established. Pharmacological treatment with lipid-lowering or anti-inflammatorymedication leads to plaque stabilization and reduction of cardiovascular events. Moreover, the implantation of a stent or scaffold for the local treatment of vulnerable plaques has been found to be safe and to stabilize high-risk plaque features. The use of drug-coated balloons to treat vulnerable plaques is the subject of ongoing research. Future studies should focus on non-invasive imaging techniques to adequately identify vulnerable plaques and further randomized clinical studies are necessary to find the most appropriate treatment strategy for


Introduction
Since the introduction of percutaneous coronary intervention (PCI) in the 1970s, the mortality rates of chronic coronary syndrome (CCS) and acute coronary syndrome (ACS) have steadily decreased [1,2]. However, ischemic heart disease is still a major cause of death and disability and it accounts for approximately one-third of all deaths in people older than 35 years [3]. The American Heart Association (AHA) estimated that ischemic heart disease, and specifically ACS, accounts for around $150 billion health costs in the United States each year [4]. Even patients with prior PCI for ACS who are on guideline-directed secondary preventive therapy, have a significant residual risk for repeat coronary events. In the first year, this risk is mainly driven by target-lesion related events (e.g., stent-related), while in the years succeeding the residual risk primarily arises from lesions in other coronary segments, which were often not considered rupture-prone during the initial PCI [5]. Therefore, early identification of these lesions responsible for (recurrent) coronary events could be of clinical significance and preventive treatment of these lesions might reduce disease burden and related health costs.
In 1989, Muller et al. [6] introduced the concept "vulnerable plaque" for-often hemodynamically insignificant-coronary plaques that were at increased risk for rupture. It was observed that in most cases of ACS, the stenosis grade of the culprit was mild (i.e., <60% diameter stenosis), and most obstruction was caused by the formation of a thrombus [7,8]. A decade later, in 2003, Naghavi et al. [9,10] provided the first consensus documents on vulnerable plaques. Since then, numerous studies have focused on vulnerable plaques to better understand the concept and the prevalence. The challenge for future studies now lies in finding the optimal detection modality and treatment option for vulnerable plaques [11]. In this review we provide a contemporary overview of (i) the morphology, (ii) the identification and (iii) possible treatment strategies of vulnerable plaques.

The vulnerable plaque
The term "vulnerable plaque" is typically used for plaques that are prone to cause a coronary event, either by rupture or erosion causing acute thrombosis, or by rapid plaque progression leading to significant stenosis and subsequent flow limitation [9]. The highest disease and mortality burden in cardiovascular disease is caused by ACS, which in the majority of cases results from (sub-)total occlusion of the coronary artery due to thrombotic obstruc- tion [12]. The formation of an intraluminal thrombus is usually precipitated by plaque rupture, which causes the release of thrombogenic factors leading to platelet aggregation and thrombin formation [13]. It is therefore of clinical importance to detect these rupture-prone plaques in an early stage. Accordingly, several studies retrospectively reviewed previous coronary angiography images of patients that presented with myocardial infarction (MI) to understand the morphology of the ruptured plaques causing the event [14][15][16]. It was found that those lesions causing a repeat event were often non-obstructive with <50% diameter stenosis during the initial coronary angiography. Therefore, research then focused on the plaque morphology to identify plaque features that make these non-obstructive plaques vulnerable for rupture. One of these specific characteristics was observed by Libby in 1995, who found that those vulnerable plaques often had a large lipid-rich core, and can therefore be referred to as atheroma or lipid-rich plaques [17]. Lipid accumulation in the coronary artery wall arises when minimal endothelial injury occurs at locations with low endothelial shear stress, for instance at bending points or near side branches [18]. In case of plaque rupture or erosion, the lipid-rich plaque releases its content into the circulation, activating tissue factor and the coagulation cascade and, in parallel, attracting circulating platelets [19]. The interaction between the released plaque components, platelets receptors and coagulation factors subsequently leads to further platelet aggregation and thrombus formation [19]. Therefore, the quality of the cap overlying the lipid core affects the risk for plaque rupture. Plaque rupture results from a disruption of the interplay between cap strength and local plaque stress. The local plaque stress is a consequence of the high fibrous cap stiffness and high endothelial shear stress [20,21]. A post-mortem study by Burke et al. [22] in a total of 113 men showed that most ruptured plaques had only a thin fibrous cap, with a mean thickness of 23 µm and 95% of ruptured plaques with a thickness of <64 µm. The specific histopathological description of this rupture-prone plaque is therefore called a thin-cap fibroatheroma (TCFA) [23,24], see Fig. 1.
The AHA proposed a scheme to classify advanced atherosclerotic plaques according to plaque type [25,26]. This classification scheme ranges from the initial fatty streak, as found in children and adolescents (Type I) to more advanced lesions with lipid-rich, confluent and necrotic cores (Type IV), lesions with calcifications or large fibrous  [26] and Virmani schemes [24]. * Total occlusion as resulting from prior thrombi. caps (Type V) or complicated lesions (Type VI). However, the classification implies that plaque evolution occurs following a linear pattern, which is often not the case. Therefore, Virmani et al. [24] suggested an alteration classification where the more advanced (Type IV-VI) lesions are described according to a more histopathological approach. According to the authors, a TCFA corresponds with a type IV atherosclerotic plaque according to the AHA classification schema [24,26]. Fig. 2 displays a schematic overview of the different plaque types.
Naghavi et al. [9,10] wrote a consensus document to uniformly define vulnerable plaques. The authors advice to use the term "vulnerable plaque" for "all thrombosisprone plaques and plaques with a high probability of undergoing rapid progression, thus becoming culprit plaques". Five major criteria for a vulnerable plaque have been identified: (1) the presence of a thin cap with large lipid core; (2) the presence of active inflammation (i.e., infiltration of macrophages); (3) endothelial denudation with superficial platelet aggregation (i.e., erosion); (4) a fissured plaque (mostly indicating recent rupture) and (5) stenosis >90%. Minor criteria include the presence of superficial calcified nodules, glistening yellow plaque on angioscopy, in-traplaque hemorrhage, endothelial dysfunction or outward remodeling [9]. In this review we refer to the term "vulnerable plaque" for the lipid-rich rupture-prone plaques.
The first prospective natural history study to correlate plaque characteristics to subsequent cardiovascular events was the PROSPECT study, conducted between 2004 and 2006 [27]. This study included almost 700 patients treated with PCI for ACS who underwent additional 3-vessel angiography and intravascular ultrasonography (IVUS). Major adverse cardiovascular events (MACE), consisting of cardiac death, cardiac arrest, MI or rehospitalization for unstable or progressive angina, were adjudicated to be either related to the initial culprit lesion, or to an additional nonculprit lesion. The PROSPECT investigators found that approximately 20% of patients experienced follow-up events, equally attributable to culprit and non-culprit lesions, in the 3 years following successful PCI. The majority of nonculprit events was a result of angiographic mild lesions with a mean diameter stenosis of 32% and one-third of the lesions with a diameter stenosis of <30%. Analysis of the 3vessel IVUS images (both gray-scale and virtual-histology) revealed an average of 2 TCFAs in half of the studied patients. This was, together with a plaque burden >70%, the most important predictor for long-term non-culprit events. Interestingly, almost none of the non-culprit events arose from non-TCFA plaques.
These findings were confirmed in the recently published PROSPECT II study [28]. This was a prospective natural history study using 3-vessel IVUS with coregistration of near-infrared spectroscopy (NIRS) in almost 900 patients with recent myocardial infarction. Patients were clinically followed-up for 4 years and again a large plaque burden of >70% was predictive for non-culprit MACE (adjusted odds ratio of 3.49 on a patient level and 12.94 on a plaque level), as well as a high lipid burden as measured with NIRS (adjusted odds ratio of 2.27 on a patient level and 7.83 on a plaque level). On the contrary, plaques lacking these high-risk features were found to have a favorable long-term prognosis.
The identification of lesions causing non-culprit events might become even more important, as the risk of culprit events by in-stent restenosis and (very) late stent thrombosis has declined considerably since the advent of newer generation drug-eluting stents (DES).

Invasive imaging techniques to detect vulnerable plaques
Assessment of plain angiography images may inform the operator about lesion specifics such as the presence of calcifications, thrombosis, or the extent of luminal obstruction as expressed in percentage of diameter stenosis. However, morphological characteristics of plaques cannot be distinguished without the use of intracoronary imaging. Several techniques exist to characterize plaques. Here we describe the most commonly used intracoronary imaging techniques: IVUS, optical coherence tomography (OCT) and NIRS, which are also summarized in Table 1 (Ref. ).

Intravascular ultrasound
IVUS is an invasive imaging technique that uses ultrasound to visualize the inside of the coronary artery walls [55]. The technique is built on the principles of ultrasonography, where high-energy sound waves are radiated into the tissue, and the reflection is returned to the transducer and converted into images. Echogenic components with acoustic shadowing indicate the presence of calcium, while echolucent components may indicate "soft plaque" [30,31]. IVUS has a high tissue penetration (4-8 mm), enabling the evaluation of the entire coronary artery structure (including the external elastic membrane), but it traditionally had a limited resolution (axial 100-150 µm and lateral 200 µm), although modern IVUS systems such as the high definition (HD) IVUS system (ACIST medical systems, Eden Prarie, MN, USA), OptiCross IVUS (Boston Scientific, Marlborough, MA, USA) and the Makoto intravascular imaging system (Nipro, Bridgewater Township, NJ, USA) boast ax-ial resolutions as low as 40,22, and 20 µm, respectively [53,56].
Conventional grayscale IVUS can be used for the assessment of luminal dimensions and plaque morphology and to evaluate stent deployment. With grayscale IVUS, the plaque burden can be measured using the following formula: (total vessel area−lumen area) total vessel area × 100% [46]. Plaque burden corresponds with the atheroma area and was found to be the strongest predictor for MACE in the PROSPECT study [27]. Vulnerable plaques associated with ACS present with greater plaque burden than stable plaques [42]. In order to overcome the limitation of grayscale IVUS that lipid and fibrous content are hard to distinguish, post-processing IVUS-based methods were introduced that use the backscattered radiofrequency signal to enhance tissue characterization [57]. Examples are virtual histology IVUS (VH-IVUS), Integrated Backscatter IVUS (IB-IVUS), iMAP-IVUS (Boston Scientific, Marlborough, MA, USA) and Automated Differential Echogenicity (ADE). These techniques provide more insight into the plaque composition and vulnerability and could for instance visualize the presence of a necrotic core [56,58]. The reported sensitivity and specificity of (VH-)IVUS to identify TCFAs, based on pathology studies, varies between 64-92% and 78-93% respectively [59,60].

Optical coherence tomography
OCT uses coherent near infrared light to generate images by measuring the intensity of light returning from the vessel wall. Due to its high resolution (10-20 µm), OCT has become a frequently used imaging modality for vessel wall characterization and plaque quantification [61]. Lipid tissues (e.g., lipid-rich plaques) appear as signal-low regions with diffuse borders, while fibrous tissues present as homogeneous signal-rich regions with low attenuation [32]. As a result, OCT can visualize important features of (potential) vulnerable plaques, such as the thin fibrous cap and underlying lipid core [29], Fig. 3. Other hallmarks of plaques such as cholesterol crystals, calcifications, neovascularization and macrophage infiltration can also be characterized by OCT [38,44]. An important limitation of vulnerable plaque detection and characterization via OCT is the poor tissue penetration of this modality, e.g., when lipid deposits are covered by a calcified cap. Pathological validation studies have shown a high sensitivity (90-94%) and specificity (90-92%) of OCT to detect lipid-rich plaques [62]. Others have suggested that OCT is less suitable to distinguish TCFAs from non-TCFAs (positive predictive value [PPV] 41%) and overestimates the incidence of TCFAs [60].

Near-infrared spectroscopy
NIRS is a relatively new catheter-based imaging modality developed to detect lipid-rich plaques [36], Fig. 4.
The NIRS system provides a spatial map of lipid distribution in the coronary artery wall, a chemogram, and produces a quantification of the amount of lipid, called the lipid core burden index (LCBI). To combine the chemogram with morphological imaging of the coronary vessel and plaque structure, NIRS is usually combined in an integrated catheter with either IVUS (IVUS-NIRS) or, less commonly, with OCT (OCT-NIRS) [64][65][66].
In 2013, Madder et al. [34] demonstrated that a maximum LCBI per 4 mm segment (maxLCBI 4mm ) of more than 400 adequately distinguished culprit segments causing ST-segment elevation MI from control segments consisting of lipid-free autopsy specimen, which provided a sensitivity of 85% and specificity of 98%. This threshold of maxLCBI 4mm >400 was also confirmed for non-ST-segment elevation ACS (NSTE-ACS), providing a sensitivity of 64% and specificity of 94% in patients with non-ST-segment elevation MI and 39% and 90% respectively in patients with unstable angina [35].
Accordingly, a prospective study by Madder et al. [67] in 121 patients confirmed that the presence of lipid-rich plaques with maxLCBI 4mm >400 caused a 10-fold higher risk for MACE than plaques with maxLCBI 4mm <400. Later, the LRP study, published in 2019, studied this on a large scale in a total of 1563 patients with suspected coronary artery disease [33]. All patients underwent 3-vessel IVUS-NIRS and patients with maxLCBI 4mm >250 plus a random subset of patients with maxLCBI 4mm <250 underwent clinical follow-up for 24 months. On a patientlevel, it was found that the presence of maxLCBI 4mm >400 Fig. 4. Lipid-rich plaque on intravascular ultrasound with near-infrared spectroscopy. (A) demonstrates the chemogram as obtained with NIRS pullback. This is a chemical map of the coronary artery, with the most distal part of the pullback on the right and the proximal part on the left. All pixels that demonstrate a probability of >60% to contain lipid, based on autopsy studies, will shift from red to yellow. The LCBI is then calculated as the amount of yellow pixels in the chemogram, or per region of interest (e.g., per 4 mm). The maxLCBI4mm in this patient is 543, i.e., more than 400, demonstrating the presence of a lipid-rich plaque at~55 mm of the pullback. In (B), the IVUS cross-section of the left anterior descending artery at the location of the lipidrich plaque demonstrates a structural overview of the corresponding part within the coronary artery. An echolucent (i.e., "dark") area is present, indicating the presence of lipid. The plaque burden, calculated as (total vessel area -lumen area) / total vessel area × 100% is 74% in this patient.
was associated with a 2-fold increased risk for non-culprit MACE, and plaque-level analysis found that segments with maxLCBI 4mm >400 were associated with a 3-fold in- The non-calcified plaque appears echolucent on IVUS (indicated by the asterisk) and the yellow NIRS signal (the surrounding circle) indicates that the plaque is lipid-rich.
creased risk for non-culprit MACE. In the PROSPECT II study, the cut-off value for lipid-rich plaques was based on the highest percentile of maxLCBI 4mm in the total cohort, which corresponded with 324.7 [28]. This threshold yielded an adjusted odds ratio of 2.3 on patient-level for future non-culprit MACE. It was suggested by the authors that the definition for vulnerable plaques should consist of a plaque burden of ≥70% and (optionally) a minimal luminal area of ≤4.0 mm 2 , both measured with IVUS, together with a maxLCBI 4mm of >325. Lowering the maxLCBI 4mm threshold from 400 to 325 could increase the specificity to detect vulnerable plaques, but comes with reduced sensitivity.

Non-invasive imaging techniques to detect lipid-rich plaques
Apart from these aforementioned intracoronary imaging techniques, several non-invasive imaging techniques exist that could identify vulnerable plaque characteristics.

Coronary angiography computed tomography
A widely available non-invasive imaging technique to visualize coronary artery disease (CAD) is computed tomography coronary angiography (CTCA). CTCA has a very high diagnostic accuracy to detect coronary stenosis in patients without previously known CAD with a sensitivity of 95-99% and a specificity of 64-83% [68,69]. Moreover, CTCA may be used to assess plaque morphology to visualize non-stenotic coronary plaques [48,49]. The plaque density as assessed on CTCA with Hounsfield Units (HU) has been found to correspond well with echogenicity on IVUS, enabling CTCA to differentiate between the different plaque types [70], Fig. 5. Plaques that are scored as soft plaques with necrotic cores based on echogenicity of IVUS typically appear with low attenuation on CTCA (HU <30), while calcified plaques appear with high HU (HU around 500). Apart from the density of coronary plaques, also the attenuation pattern may be of additional information to further classify high-risk plaques. Studies have found that the addition of analysis of the so-called "napkin-ring sign", i.e., low CT attenuation in the center of the plaque but with a rim area of higher attenuation, could increase the diagnostic accuracy of CTCA to identify high-risk plaques [71,72]. It is believed that the presence of this napkin-ring sign corresponds with the OCT features of TCFA [73].
Other lesion features apart from plaque morphology that can be visualized with CTCA include vessel remodeling. IVUS-based studies found that high-risk plaques that cause ACS often show positive, i.e., outward, remodeling of the coronary artery at the lesion site [50]. This IVUS fea-ture can also be visualized with CTCA and the finding has been associated with ACS [51,52]. Additionally, perivascular adipose tissue can be visualized using CTCA, which can be used as a biomarker for coronary inflammation and thus identify inflamed, vulnerable atherosclerotic plaques [74].

Cardiovascular magnetic resonance imaging
Cardiovascular magnetic resonance (CMR) imaging is often used in cardiovascular medicine to evaluate structural heart disease. CMR is not frequently used in coronary plaque analysis, but it has been used to classify atherosclerotic plaques in the carotid arteries [75]. Therefore, studies have explored the utilization of CMR as a diagnostic tool to identify vulnerable coronary plaques. Károlyi et al. [37] analyzed 28 plaques of donor hearts with CMR with proven CAD. The authors classified lipid-rich plaques as those that appear hypointense on the T2-weighted CMR images, calcified plaques as those with hypointense areas on T1-weighted images, but not on the ultrashort echo time images, and fibrotic plaques as isointense regions on all sequences. This classification system was adapted from that of the carotid artery classification. These CMR classified plaques were then correlated with histopathological findings and it was demonstrated that CMR could adequately distinguish calcified plaques (sensitivity 100% and specificity 90%) and lipid-rich plaques (sensitivity 90% and specificity 75%), suggesting that CMR could be an additional diagnostic tool for coronary plaque analysis.
On a larger scale, Noguchi et al. [76] examined a total of almost 600 patients that had suspected or known CAD who underwent CMR (without contrast) to investigate whether T1-weighted imaging could identify high-risk coronary plaques that cause future events. The plaque-tomyocardium signal intensity was identified as the most important predictor for future coronary events. This signal intensity ratio itself is incrementally associated with highrisk plaque morphology [77]. Noguchi et al. [76] used a cut-off value of ≥1.4 to identify high-intensity plaques and found that those plaques had a 4-fold higher risk for coronary events than plaques with lower signal intensity.

Nuclear imaging
Nuclear imaging using radioactive tracers containing a 18 F-fluoride isotope, such as 18 F-fluorodeoxyglucose ( 18 F-FDG), or a Technetium-99m isotope can be used to visualize myocardial perfusion and metabolic activity [78,79]. To investigate the inflammation and lipid infiltration that are associated with vulnerable plaque morphology, nuclear molecular imaging may be of additional use. Several pathological processes that occur in vulnerable plaques can be targeted for imaging, such as inflammation and (micro)calcifications [39,40]. In a prospective clinical trial, Joshi et al. [80] performed 18 F-fluoride positron emission tomography (PET) imaging on 40 patients with MI and 40 patients with stable CAD and corresponded the findings with IVUS and histology. High uptake of 18 Ffluoride was found in plaques that had high-risk features on IVUS, such as positive remodeling and necrotic core. A limitation of nuclear imaging of the coronary arteries is that is often obscured by the tracer uptake of the myocardium, which in the study of Joshi et al. [80] was mainly seen with 18 F-FDG but not with 18 F-sodium fluoride ( 18 F-NaF). 18 F-NaF has the additional advantage that microcalcifications can be visualized, that-unlike the macrocalcifications as seen on CTCA-are associated with coronary plaque progression and vulnerable plaque formation [45]. 18 F-NaF PET imaging has therefore been the subject of ongoing research to guide treatment strategies, such as the PREFFIR study (ClinicalTrials.gov, identifier NCT02278211) and the EVOLVE study (ClinicalTrials.gov, identifier NCT03689946).
Another nuclear imaging modality is single positron emission computed tomography (SPECT). There has been a report of the use of SPECT to identify vulnerable plaques with TCFA-like features, mostly focusing on the detection of neovascularization, but evidence for a diagnostic value of SPECT in the vulnerable plaque analysis is limited [81].

Systemic treatment 4.1.1 Lipid lowering drugs
Current guidelines recommend lipid-lowering treatment in patients with established coronary artery disease irrespective of low-density lipoprotein (LDL) cholesterol [82]. The abovementioned technological advances in coronary imaging, including the ability to quantify plaque burden, have enabled evaluation of the recommended lipidlowering treatment. The ASTEROID investigators demonstrated that two years of statin treatment did not only reduced LDL cholesterol, but also led to regression of plaque burden measured by quantitative coronary angiography [83]. Multiple studies using serial IVUS imaging have later reaffirmed that statin therapy can slow disease progression and even promotes plaque regression in a dose-response manner [84][85][86].
Novel imaging modalities have also sparked interest in the effects of lipid-lowering treatment on plaque composition, looking beyond mere plaque burden or size to the role of macrophage infiltration (inflammation) and fibrous cap thickness. In the early 2000's, Crisby et al. [87] demonstrated that statin therapy effects plaque composition. Carotid plaques obtained during endarterectomy from statin-treated patients showed a lower lipid burden and less inflammatory material compared to carotid plaques from non-statin-treated patients, suggesting a plaque stabilizing effect of statin treatment. Other studies have later reiterated that the anti-inflammatory properties of statin therapy can positively impact plaque stability [88,89]. Multiple studies have also demonstrated that statin treatment can thicken and harden (by calcification) the fibrous cap of lipid-rich plaques, making these plaques less prone for rupture [90][91][92][93][94].
In patients who do not reach their treatment goal with the maximum tolerated dose of statin, a cholesterol absorption inhibitor (ezetimibe) or a proprotein convertase subtilisin kexin type 9 (PCSK9) inhibitor on top of statin treatment should be considered [82]. Clinical trials have shown the positive effect of ezetimibe on cardiovascular outcomes in certain subgroups, but data on the effect of ezetimibe on lipid-rich plaques is limited [95,96]. PCSK9 inhibitors have also demonstrated to effectively decrease LDL cholesterol levels and prevent adverse cardiovascular events in clinical trial and real-world populations [97][98][99]. In statintreated patients, additional use of PCKS9 inhibitors induced plaque regression measured by IVUS-imaging in a greater percentage of patients compared to placebo [100]. Until recently, it remained unclear if administration of PCSK9 inhibitors also effects plaque composition in statin-treated patients [101]. During the ESC Congress 2021 the first results of the HUYGENS study (ClinicalTrials.gov, identifier NCT03570697) were presented. HUYGENS assessed whether intensified lipid lowering treatment with PCSK9 inhibitors on top of the maximally tolerated statins effected the high-risk features of lipid-rich plaques in 161 patients with non-ST-segment elevation ACS [102]. In total, 135 patients completed repeated OCT imaging after 12 months of treatment with either a PCSK9 inhibitor (evolocumab) or placebo. More intensive lipid-lowering using a PCKS9 inhibitor resulted in an increase of the minimum fibrous cap thickness and a decrease in the maximum lipid arc. At 12 months, only one in eight patients treated with a PCSK9 inhibitor in addition to the maximally tolerated statin dose had a fibrous cap thickness of <65 µm, a feature associated with a high risk of plaque rupture. Again, the degree of lipid-lowering was proportional to the observed beneficiary effects.

Anti-inflammatory drugs
Accumulation of cholesterol within the vessel wall induces inflammation by activating the innate immune response and is associated with plaque instability [103]. Therefore, systemic treatment with anti-inflammatory agents might be beneficiary in patients with coronary artery disease. In recent years especially the use of colchicine, an anti-inflammatory agent previously only used for conditions such as pericarditis and gout, has received considerable interest [104,105]. Observational studies had already demonstrated that the use of colchicine was associated with a lower prevalence of MI in patients receiving colchicine for inflammatory conditions [106][107][108]. The open-label LoDoCo trial was the first randomized controlled trial evaluating the effect of low-dose colchicine (0.5 mg once daily) in addition to standard secondary prevention therapies on a composite endpoint of ACS, out-of-hospital cardiac arrest and ischemic stroke in patients with CCS (n = 532) [104]. Colchicine use drastically reduced the primary outcome (hazard ratio 0.33; 95% CI 0.18-0.59; p < 0.001) and was generally well tolerated, although 11% of patients assigned to colchicine withdrew within 30 days due to perceived intestinal intolerance [104]. Subsequently, the double-blind LoDoCo2 trial was performed to confirm whether lowdose colchicine use prevented cardiovascular events in patients with chronic coronary disease. A total of 5522 patients underwent randomization and were followed for a median duration of 28.6 months. The composite endpoint of cardiovascular death, spontaneous MI, ischemic stroke or ischemia-driven coronary revascularization occurred in 6.8% of colchicine-treated patients and in 9.6% of placebotreated patients (hazard ratio 0.69, 95% CI 0.57-0.83; p < 0.001). Hence, the use of low-dose colchicine may now be considered in secondary prevention of cardiovascular disease, particularly if other risk factors are insufficiently controlled or if recurrent events occur under optimal therapy according to the 2021 ESC Guidelines on cardiovascular disease prevention in clinical practice [109]. In 2017, the CANTOS investigators demonstrated that the anti-inflammatory drug canakinumab, a monoclonal antibody targeting interleukin-1β, led to a significantly lower rate of recurrent cardiovascular events at a dose of 150 mg every 3 months, but was also associated with a higher incidence of fatal infection compared to placebo [110]. Other anti-inflammatory agents, such as lipoprotein-associated phospholipase A2 inhibitors or low-dose methotrexate, are currently not recommended for secondary prevention due to uncertain efficacy or their unfavorable risk profile [111][112][113].

Local treatment
Despite these aforementioned effects of systemic treatment on stabilization of lipid-rich plaques, it appears that solely a strict systemic lipid-lowering regime and even adding an anti-inflammatory agent might not be enough. Patients on intensive statin therapy, according to treatment allocation in the PROVE IT TIMI 22 trial, had a residual risk for cardiovascular events of roughly 20% during 2 years follow-up [114]. This stresses the need for an additional, more local treatment of vulnerable plaques.
The recently published PROSPECT ABSORB trial explored local treatment of vulnerable plaques by stenting with a bioresorbable vascular scaffold (BVS) [64]. The rationale of this study was based on a previous study where OCT images were assessed after BVS or bare-metal stent implantation of TCFAs and a potential beneficial effect on plaque stabilization was suggested since neo-intimal tissue development was observed after short-and mid-term follow-up [115]. The idea was that this would lead to thickening of the fibrous cap overlying the lipid pool and a decrease of wall shear stress, resulting in a reduction of plaque rupture or erosion. BVS was considered to be able to ob- ↑ Fibrous cap thickness [115] ↑ MLA [117] ↑ Fibrous cap thickness [122,123] ↓ Plaque burden [83][84][85][86] ↑ MLA [116,117] ↓ Plaque burden [121][122][123] ↓ Inflammation [88,89] ↓ Inflammation [121] Clinical endpoints ↓ MACE [95][96][97][98][99]126]  tain similar neointimal response and shear stress reduction as DES [116,117]. It was preferred over DES since it was observed that the neointimal response caused by BVS led to smaller compromise of the coronary lumen than with baremetal stent, plus it would overcome the placement of a permanent stent [115]. The PROSPECT II study, as discussed above, enrolled almost 900 patients with ACS to perform 3-vessel IVUS-NIRS after treatment of every flow-limiting stenosis [28]. All patients that had non-obstructive plaques with a plaque burden greater than 65% on IVUS, the study definition for vulnerable plaque, in the PROSPECT II study were included in the PROSPECT ABSORB trial and randomized to medical therapy or BVS treatment of the identified vulnerable plaque [64]. Randomized patients underwent follow-up angiography after 25 months with repeat IVUS-NIRS and it was found that the minimum luminal area was greater in patients treated with BVS (7 mm 2 versus 3 mm 2 , p < 0.0001) and that the lipid burden, as measured with NIRS, was significantly reduced (lipid burden 6% versus 27%, p < 0.0001; maxLCBI 4mm 62 versus 269, p < 0.0001). Thus, revascularization with BVS seemed effective in the reduction of the lipid burden. But since the study was not powered for clinical endpoints, an effect on target lesion failure could not be confirmed. The prematurely-halted PECTUS study had a similar trial design to the PROSPECT ABSORB study, although the chosen intracoronary imaging modality was OCT [118]. The study was halted when the ABSORB stent was removed from the market. Although no definite conclusions could be drawn from the 34 patients that had been randomized between BVS or optimal medical therapy by the time of termination of the study, no differences in the primary endpoint of target-lesion failure were observed between the randomization groups. The PREVENT study (Clinicaltrials.gov, identifier NCT02316886) is currently ongoing and aims to include 1600 participants with both ACS and CCS to investigate treatment of vulnerable plaques with either BVS or DES versus optimal medical therapy alone.
A major concern of local treatment of a lipid-rich plaque is the risk for distal embolization of the lipid pool and potential periprocedural MI [119,120]. However, no procedural MI occurred in the entire PROSPECT ABSORB trial, and the primary safety endpoint of target-lesion failure (consisting of cardiac death, target-vessel related MI or target-lesion revascularization) was equal between randomization groups.
Still, implantation of a stent or scaffold in non-flowlimiting, thus non-ischemic, plaques could introduce potential harms of a new, stent-related "disease" such as stent thrombosis or in-stent restenosis that would otherwise not occur. Therefore, the DEBuT-LRP study (ClinicalTrials.gov, identifier NCT04765956) is currently investigating the safety and efficacy of treatment of lipid-rich plaques with a drug-coated balloon (DCB). An animal study demonstrated that in cholesterol-fed rabbits with balloon-injuryinduced lesions in the aorta, angioplasty with a paclitaxelcoated balloon led to a higher reduction of inflammation and plaque burden than plain balloon angioplasty [121]. Human studies confirmed that DCB treatment led to plaque burden reduction and even an increase of the fibrous cap thickness [122,123]. Thus DCBs are a proven alternative to stenting for several indications, but could now also play a role in the treatment of lipid-rich plaques [124]. The DEBuT-LRP study aims to enroll 40 patients with NSTE-ACS that will subsequently undergo 3-vessel IVUS-NIRS and a lipid-rich plaque (i.e., maxLCBI 4mm >400) will be treated with DCB. DCB-treated patients will undergo follow-up angiography after 9 months with repeat IVUS-NIRS for the primary endpoint of maxLCBI 4mm reduction of the treated LRP. If proven effective, the DCB may be a safe alternative to stenting in the treatment of a lipidrich plaque. Table 2 (Ref. [83][84][85][86]88,99,104,105,110,[115][116][117][121][122][123]125,126]) summarizes the effects of different treatment strategies on vulnerable plaque features.

Conclusions
ACS mostly arises from rupture or erosion of a vulnerable plaque. Vulnerable plaques typically appear as lipidrich plaques with a thin cap, called TCFA. These vulnerable plaques can be detected using intracoronary imaging and possibly also with non-invasive imaging techniques. Currently, the implantation of a stent or scaffold for the treat-ment of vulnerable plaques has been found to be safe and to stabilize high-risk plaque features. Future studies should focus on optimizing imaging techniques and evaluating the effectiveness of vulnerable plaque stabilization on clinical endpoints.

Author contributions
AvV and NMRvdS wrote the manuscript with support from JPSH and BEPMC. All authors read and approved the final manuscript.

Ethics approval and consent to participate
Not applicable.