The Vulnerable Atherosclerotic Plaque: Scope of the Literature FREE

The concept of the vulnerable plaque has been proposed as a paradigm to improve the prevention of cardiovascular disease by identifying atherosclerotic plaques at higher risk for causing future cardiovascular events.

Proposed definition criteria of vulnerable plaque, based on examining plaques that have already caused a clinical event, include active inflammation, a thin cap with a large lipid core, endothelial denudation, fissured cap, severe stenosis, or combinations of these findings.

Studies on biomarkers of plaque vulnerability are cross-sectional and thus cannot provide information on the predictive value of examined associations.

Imaging studies are almost all limited by cross-sectional or retrospective design, diverse imaging methods or imaging features studied, and lack of reliable documentation on whether identified lesions were responsible for future cardiovascular events in the few prospective studies.

Studies on therapies lack a standard definition of plaque vulnerability and are limited by the obligatory surrogate nature of the outcomes studied.

Substantial conceptual and methodological challenges need to be addressed to realize the potential promise of the vulnerable plaque concept.

Atherosclerosis is a chronic condition with acute cardiovascular manifestations. Most commonly, the acute manifestations of atherosclerosis are triggered by a local arterial occlusion with a thrombus overlying a preexisting atherosclerotic plaque. Despite major advances in the prevention and treatment of cardiovascular disease, it remains the leading cause of morbidity and mortality worldwide, accounting for 35% of all deaths in the United States and 30% of all deaths globally in 2005 (1–2). A particular challenge to combating the epidemic of cardiovascular disease is the sudden and often unpredictable nature of its acute manifestations. For many patients, the first sign of atherosclerosis is an acute myocardial infarction, sudden cardiac death, or a disabling stroke. This has fueled considerable research aimed at refining existing algorithms for risk stratification and developing new methods to identify at-risk persons before the occurrence of a cardiovascular event so that primary preventive measures can be initiated. Furthermore, among patients who have survived a cardiovascular event, the risk for a subsequent event remains relatively high—approaching 1 in 4 despite aggressive treatment (3). Such recurrence rates highlight the need for novel approaches to secondary prevention of cardiovascular disease and to the treatment of index events.

Over the past 2 decades, the concept of “vulnerable plaque” has gained attention as a paradigm to improve risk stratification and potentially lead to newer invasive and noninvasive therapeutic options to prevent and treat atherothrombotic cardiovascular disease. Narrative reviews have previously described the concept and premise of the vulnerable plaque, as well as related technologies in development (4–5). To better understand the current evidence for the concept of vulnerable plaque, we sought to systematically explore recently published literature guided by specific aims: 1) to provide an overview of recent definitions and thinking on the concept of vulnerable plaque; 2) to describe the current understanding of the natural history of vulnerable plaques; 3) to provide an evidence map of diagnostic and therapeutic methods currently investigated for the detection and treatment of vulnerable plaques; and 4) to identify challenges and future research directions for this evolving concept. Given the still poorly defined nature of the topic, we decided to describe the research community's definitions and exploration of vulnerable plaque, not systematically review the results of individual clinical studies or make clinical recommendations.

We based our article on a technical brief produced by the Tufts Medical Center Evidence-based Practice Center (Boston, Massachusetts) for the Agency for Healthcare Research and Quality (6), which expanded on our 2004 report (7). To assess the volume and type of evidence available on vulnerable plaque, describe how the research field is evolving, and identify topics that may require further research, we created an evidence map primarily on the basis of study abstracts that outline the tests that have been evaluated, the populations in which they have been evaluated, and the types of studies that have been used. Vulnerable plaque is not an established medical diagnosis and is still an evolving concept. Therefore, any reference to vulnerable plaque in this report concerning its natural history, diagnostic methods, and treatments is, by necessity, inferential. It refers to conditions that might be classified under the current concept of vulnerable plaque.

We searched MEDLINE to identify studies published in English on vulnerable plaque. Because vulnerable plaque is an emerging concept, no specific Medical Subject Headings are available. Therefore, we used only text words. The search terms included vulnerable plaque, unstable plaque, atheromatous plaque, ulcerative plaque, and related words. The search was limited to studies conducted in humans. Because we focused on the current thinking about the vulnerable plaque concept, we limited the search from 2003 to April 2010. Four independent reviewers examined the abstracts. A cardiologist re-evaluated all potentially relevant abstracts. Furthermore, in order to appreciate the ongoing research effort on vulnerable plaque, we did a systematic search of the ClinicalTrials.gov registry to identify observational and interventional studies on vulnerable plaque. Protocols of retrieved entries were reviewed for use of predictors or outcomes relevant to the concept of vulnerable plaque, and those studies were tabulated.

We aimed to identify abstracts of studies that fit a broad concept of vulnerable plaque. These included studies with the stated or implied aim of examining plaque features thought to be related to the susceptibility of an atherosclerotic plaque to cause a clinical event (for example, myocardial infarction or stroke). These features included histopathologic (for example, thin-cap fibroatheroma, macrophage infiltration, and presence of lipid core or intraplaque hemorrhage) or imaging (for example, plaque echolucency and density, presence of ulceration, or intraplaque hemorrhage) findings. We included primary studies that were conducted in living humans, were on carotid or coronary artery tissue removed during procedures on living humans, or were autopsy studies. We also collected narrative reviews, commentaries, and editorials on the concept of vulnerable plaque.

From eligible primary studies, the same 4 reviewers extracted information from the available abstracts: study design (cross-sectional or longitudinal follow-up or prospective or retrospective collection of data), sample size, population (with or without history of carotid or coronary artery disease, living patients, tissue samples, or autopsy samples), predictor of interest (imaging characteristics of the plaque, histopathology, biomarkers, and others), information on treatments, outcomes (clinical, imaging, or histopathologic), and whether the study measured direct effects of a particular diagnostic tool on physician decision making or patient outcomes.

The Agency for Healthcare Research and Quality participated in the formulation of the review objectives but did not participate in conducting the review or in the preparation, review, or approval of the manuscript for publication.

The MEDLINE search yielded 1466 titles, of which 463 abstracts (both primary and review articles) qualified for inclusion (Appendix Table 1). The number of review articles on vulnerable plaque (n = 221) published in the past 7 years is almost equal to the number of primary studies (n = 242) on the topic (Appendix Table 2). Most primary studies were of cross-sectional design (n = 216 [89%]), comparing different potential features of vulnerable plaque (for example, plaque characteristics on pre-endarterectomy imaging with tissue histology from postsurgical specimens); the remaining 26 studies were longitudinal studies (24 prospective) that evaluated the natural history of plaques or the effect of treatment on plaque features thought to be related to the propensity of a plaque to cause a clinical event (for example, plaque characterization by imaging methods and subsequent risk for cardiovascular events). Since 2003, about 64 articles on vulnerable plaque have been published annually, with a slight increase in the annual number of publications over time.

Of the 242 primary studies, 114 (47%) were conducted in patients with coronary artery disease and 130 (54%) in patients with carotid artery disease; 5 studies evaluated both coronary and carotid artery disease (Appendix Table 2). Only 3 studies were conducted in participants without a history of cardiovascular disease (healthy persons or persons with cardiovascular risk factors but no clinical disease). The primary studies enrolled populations grouped into 3 categories: living persons (143 studies, 59%); tissue samples obtained from living patients during carotid endarterectomy or coronary atherectomy (79 studies, 33%); and autopsy specimens of the coronary or carotid arteries (20 studies, 8%). The predictors of interest (that is, factors examined for their association with plaque vulnerability) included biomarkers (36 studies, 15%); histopathologic findings (60 studies, 25%); imaging features (120 studies, 49%); therapeutic approaches (17 studies, 7%); and combinations of these and other predictors (9 studies, 4%).

The term vulnerable plaque was first used 20 years ago in the context of studying triggers of acute cardiovascular disease (8). It was proposed that acute thrombosis resulting in total arterial occlusion is preceded by the development of the vulnerable atherosclerotic plaque (8). Plaque vulnerability was defined as the susceptibility of a plaque to rupture, thus causing a clinical cardiovascular event. The introduction of this concept paralleled an increase in appreciation of the limitations of imaging arterial lumens and quantifying risk based merely on the severity of arterial stenoses. In several prospective and retrospective serial angiographic studies, the culprit lesion in nearly two thirds of patients with acute coronary events was shown to have less than 70% (often <50%) diameter narrowing on coronary angiography weeks or months before the index event (9–12). In addition, the site of myocardial ischemia found during a stress test (an indication of a hemodynamically important stenosis in the coronary artery supplying that territory) was not found to accurately predict the site of future myocardial infarction (13).

In retrospective autopsy studies, 3 histologic features were more commonly observed in plaques thought to be responsible for most acute coronary events compared with stable plaques: a larger lipid core (>40% of total lesion area), a thinner fibrous cap (<65 µ), and more inflammatory cells (about 26% macrophage infiltration of fibrous cap compared with 3% in stable plaques) (14–16). Such observations fueled research into invasive and noninvasive imaging tests to detect these histologic features and, in doing so, identify plaques that presumably are more likely to rupture. Since its introduction, the term vulnerable plaque has been used interchangeably in reference to the concept of propensity to result in an acute cardiovascular event or to denote a plaque with the histologic hallmarks of culprit lesions from autopsy studies. A more inclusive definition was proposed in 2003 to include not only susceptibility to rupture but, more broadly, susceptibility to thrombose or rapidly progress to a culprit lesion (17). This broadening of the definition was based on observations that rupture of plaques, although common in culprit lesions, is not universal. Nearly one third of such lesions exhibit erosion or nodular calcification without rupture of the fibrous cap (18).

On the basis of retrospective studies on culprit plaques (that is, plaques that have caused an acute event), investigators in the field proposed several criteria to define a vulnerable plaque—a plaque that is at high risk for causing an acute event. The major criteria included active inflammation; a thin cap (<100 µ) with a large lipid core (>40% of the plaque's total volume); endothelial denudation with superficial platelet aggregation; fissured cap, which may indicate a recent rupture; or severe stenosis, which would make the plaque more prone to shear stress or may be a marker of other less stenotic but vulnerable plaques (17) (Figure). According to this proposal, the presence of at least 1 of these major criteria may indicate a higher risk for plaque complication. The minor criteria for plaque vulnerability included the presence of superficial calcified nodules; yellow color, which may indicate a larger lipid core; intraplaque hemorrhage; endothelial dysfunction (impaired endothelial vasodilator function); and expansive (positive) remodeling, which refers to compensatory outward enlargement of the vessel wall without luminal compromise (17). Notably though, the predictive utility of these criteria has not been prospectively validated. In addition to the local features characterizing plaque vulnerability, evidence suggests that systemic factors may play a role in plaque instability, including the presence of a systemic inflammatory state (17). This provides the rationale to studying serum biomarkers that may identify patients with high-risk lesions (vulnerable blood), which, along with vulnerable myocardium, form the triad of vulnerability that defines the vulnerable patient (17).

After the introduction of the criteria for the definition of vulnerable plaque, 60 studies published in the past 7 years focused on histopathologic findings of culprit plaques from the coronary or carotid vessel walls (Appendix Table 3). Twenty-eight studies evaluated coronary artery samples, derived from living persons during coronary atherectomy or from autopsy specimens. Among the hallmark histopathologic features of vulnerability, macrophage infiltration of the plaque was the most commonly examined but only in 4 studies. Most of the studies analyzed the tissue expression of molecules or cells proposed to be involved in the pathophysiologic processes of the disease (such as C-reactive protein, matrix metalloproteinases, and lipoprotein-associated phospholipase A2) as potential pathology markers. In 19 studies, the histopathology features were examined in association with clinical outcomes, most commonly acute coronary syndrome (ACS), whereas in the remaining studies, experimental pathology markers (that is, inflammatory molecules expressed in the plaque tissue) were compared with reference histopathologic features, such as cap thickness, large lipid core, and inflammatory cells. Thirty-two studies evaluated carotid artery samples, almost exclusively taken from living persons (n = 31), given the easier availability of carotid artery samples after carotid endarterectomy, compared with coronary tree samples. Macrophage infiltration was again the most commonly evaluated hallmark feature of vulnerability but only in 3 studies; other studies evaluated the tissue expression of molecules, such as matrix metalloproteinases and vascular endothelial growth factor. Study outcomes included either clinical disease (that is, symptomatic carotid disease) in 18 studies or other histopathologic features in the remaining 14 studies.

In this recent literature sample of histopathology studies, most studies did not evaluate the histopathologic features of the proposed vulnerable plaque definition. Instead, novel pathology markers with potential clinical or research utility were sought, mainly by examining the expression of molecules or cells within the plaque tissue. These studies may augur the future directions of active investigations.

Because of the lack of a standard definition for vulnerable plaque, we found no studies that explicitly evaluated the natural history of vulnerable plaques. However, we have identified longitudinal studies that investigated the natural history of plaque features that could be indicative of vulnerability or instability. Such studies involved a baseline imaging evaluation of the morphology of coronary (19–23) or carotid (24–30) plaques and analyzed the occurrence of clinical events, imaging end points, or both in patients at follow-up. These studies provided information on whether the presence of baseline vulnerability features conferred risk for future events or how these features had evolved overtime.

For coronary artery disease, the largest study to date (19) involved 1059 patients with suspected or known disease who had computed tomography (CT) angiographic examinations and were followed for 27 months for the development of ACS. The coronary lesions were analyzed for the presence of 2 features of vulnerability: positive remodeling (>10% diameter at the plaque site compared with the reference segment) and low attenuation plaques (noncalcified plaque with density <30 Hounsfield units). Acute coronary syndrome developed in 10 of 45 (22%) patients that showed plaques with both vulnerability features, compared with 4 of 820 (0.5%) patients that showed plaques without these features. None of the 167 patients with normal angiography results developed ACS. The presence of 1- or 2-feature positive plaques was the only significant independent predictor of ACS (hazard ratio, 22.8 [95% CI, 6.9 to 75.2]) (19). Additional studies (20–21) identified vulnerable plaques with intravascular ultrasonography. In 1 study, the multiplicity of vulnerable plaques (plaques with rupture, lipid core, dissection, or thrombus) was the only independent predictor of acute events (hazard ratio, 2.2 [CI, 1.4 to 3.4]) (20), whereas in patients from the ASTEROID (A Study to Evaluate the Effect of Rosuvastatin on Intravascular Ultrasound-Derived Coronary Atheroma Burden) trial (21), attenuated plaques in nonculprit segments of the coronary tree remained stable and were not associated with clinical events during follow-up. In an angioscopic study of 552 patients, the presence of several yellow plaques at baseline was associated with an increased risk for ACS (22).

For carotid artery disease, most longitudinal studies (n = 4) have used magnetic resonance imaging for the characterization of carotid lesions. The presence of vulnerability features at baseline was associated with occurrence of index (24) and recurrent cerebrovascular events (25) or with development of new lesions with vulnerability features (26). In studies that used carotid artery ultrasonography, plaque echolucency (thought to be indicative of lipid core) (27–28) or ulcerated morphology (29) were associated with new neurologic events.

Biomarkers were used as predictors by 36 studies, all of which were cross-sectional in design (Appendix Table 4). All studies involved measurement of blood or serum biomarkers in living patients. Coronary artery disease was studied in 19 studies, with a median number of 89 patients enrolled per study (interquartile range [IQR], 49 to 172 patients). Fourteen biochemical markers were investigated, C-reactive protein and matrix metalloproteinases being the most commonly studied, and their concentrations were most commonly compared with imaging findings of plaques. Carotid artery disease was evaluated in 15 biomarker studies, with a median number of 88 patients enrolled per study (IQR, 62 to 164 patients). Among the 19 biomarkers investigated, C-reactive protein, matrix metalloproteinases, and pregnancy-associated plasma protein A were the most commonly examined. Biomarkers were compared with clinical outcomes, histopathology features, and imaging findings. One study (31) enrolled apparently healthy persons and compared the apolipoprotein (apo) B–apo A-I ratio with carotid plaque echolucency (as an imaging feature of plaque vulnerability), whereas another study (32) in elderly persons examined the association between carotid plaque echolucency and insulin-like growth factors concentrations.

All studies that evaluated the association between biomarkers and vulnerability features of plaques (assessed by imaging or histopathology) and also with clinical outcomes were cross-sectional in design. Even though the specific results in the individual studies were not reviewed in this report, the available literature cannot provide information on the predictive value of biomarkers for future events caused by vulnerable plaques.

A total of 120 studies used imaging characteristics as predictors of plaque vulnerability: 52 studies evaluated coronary artery disease, and 68 studies examined carotid artery disease. Most studies (86%) were cross-sectional in design, and none of the 11 prospective studies was designed to measure the direct effect of imaging methods on a physician's decision making or patient outcomes (19–23, 25, 27–29, 33). Also, none of the studies documented whether the identified lesions were the ones responsible for any cardiovascular events. Imaging characteristics were compared with clinical, histopathologic, and other imaging outcomes.

Among the coronary artery disease studies, 45 were cross-sectional, 6 were prospective, and 1 was retrospective in design (Appendix Table 5). The median number of patients enrolled was 58 (IQR, 30 to 140 patients). Twelve different imaging methods were examined; intravascular ultrasonography (with or without virtual histology capacities) and multidetector CT were the most commonly used. Twenty studies examined combinations of imaging features, whereas the remaining evaluated single features of vulnerable plaque. Yellow color on coronary angioscopy and measurement of the fibrous cap thickness were the most commonly examined features.

Of the 68 studies that focused on carotid artery disease, 62 were cross-sectional, 5 were prospective, and 1 was retrospective. The median number of patients enrolled was 39 (IQR, 18 to 92 patients). Twelve different imaging methods were examined; magnetic resonance imaging and carotid ultrasonography were the most commonly used. Most studies evaluated single features of vulnerable plaque; the remaining examined combinations of features. Plaque echolucency on ultrasonography, presence of intraplaque hemorrhage, and ulceration complexity of the lesion were the most commonly examined features.

As with the literature on biomarkers, imaging studies related to the concept of the vulnerable plaque are limited by the retrospective or cross-sectional design in most studies and the diverse nature of imaging methods or imaging features studied. The few prospective studies published to date could not measure the direct effect of imaging methods on physician decision making or patient outcomes, nor could they reliably document whether the identified lesions were responsible for future cardiovascular events.

Because no standard definition for vulnerable plaque exists, no therapeutic approaches have been specifically developed and tested for the treatment of vulnerable plaque. Based on the current concept of the vulnerable plaque, several treatment strategies have been evaluated for their effect on plaque features suspected to confer vulnerability (as determined by imaging or histopathologic studies) and the associated reduction in risk for future cardiovascular events. In our 2004 technical report (7), we had described proposed treatments for vulnerable plaque at the time (fish oil, statin, antioxidant, and antibiotic), along with the conceptual basis for these strategies. Since that report, we have identified 17 additional studies (34–49) that examined interventions evaluated for modifying potential features of vulnerability in atherosclerotic plaques (Appendix Table 6). All included primary studies investigated systemic interventions: statins (38–45, 47–50), several risk factor interventions (advice on smoking cessation and optimal lipid levels and metabolic control in diabetic patients) (36), omega-3 and omega-6 polyunsaturated fatty acids (35), peroxisome proliferator-activated receptor-γ agonists (37, 46), and an oral lipoprotein-associated phospholipase A2 inhibitor (darapladib) (34). Although the premise of focal treatment of vulnerable plaques has been discussed in review articles on the concept (4), we did not identify any primary studies (apart from case reports) investigating focal approaches in our literature search.

Five studies examined patients with coronary artery disease, and 12 examined patients with carotid artery disease (in 3 of these studies, patients had already received a diagnosis of coronary artery disease as well). Prospective design was used in 13, and cross-sectional design was used in 4. Surrogate imaging or histopathologic outcomes were used by 14 (most commonly plaque echolucency on carotid artery ultrasonography), and 3 studies examined clinical outcomes (symptomatic carotid disease or composite cardiovascular end point); of those 3, only 1 study was prospective (36).

The literature on therapeutic approaches to modify plaque vulnerability is limited by the lack of a standard definition of the vulnerable plaque and by the obligatory surrogate nature of the outcomes studied. Although some of the interventions examined are known to improve clinical outcomes (for example, statins), others may increase cardiovascular events (51), despite the suggestion that features of plaque vulnerability are favorably modified (for example, rosiglitazone) (37).

A total of 29 active (completed or ongoing) trials on vulnerable plaque were identified from the ClinicalTrials.gov registry search. Of those, 16 studies are interventional and 13 are observational. Most studies (69%) focus on coronary artery disease, and various interventions or predictors of outcomes are examined (including imaging methods, drugs, and invasive coronary interventions). Appendix Table 7 provides a brief description of the registered trial protocols.

The vulnerable plaque concept has gained considerable attention in the literature during the past few years. If vulnerable plaques can be detected prospectively and accurately and effective therapeutic interventions can be initiated before cardiovascular events occur (all in a cost-effective manner), many cardiovascular events can be prevented. However, for this potential promise to be realized, substantial challenges need to be addressed. These challenges are both conceptual and methodological.

Conceptually, the presence of a vulnerable plaque is, by definition, a probabilistic entity. It does not denote the occurrence of an event at present but rather a higher risk for such occurrence in the future relative to a nonvulnerable or less vulnerable plaque. As such, before it is widely adopted by clinicians, plaque vulnerability (if validated) should be able to provide incremental predictive value on top of currently available methods of risk stratification, which may be less expensive and less invasive than the methods proposed to detect vulnerable plaques. Moreover, the complex implications of such a probabilistic diagnosis are exemplified in the observation that not all plaques that rupture (the basis for the classic definition of the term) actually result in a clinical cardiovascular event. Some plaques would rupture and then become quiescent and heal without causing a myocardial infarction or stroke (so called silent plaque rupture) (52–53). Conversely, not all acute cardiovascular events are the result of plaque rupture because nonruptured plaques have been implicated as culprit lesions nearly one third of the time in autopsy series (18). All these issues conceptually limit the sensitivity and specificity of the vulnerable plaque as a predictor of a future event. Furthermore, whether features of plaque vulnerability are interchangeable among vascular beds is uncertain (54). In other words, would a high-risk marker validated in coronary artery disease be relevant in studying plaque vulnerability in carotid or cerebral arteries? Understanding such distinctions is relevant to the broad application of the vulnerable plaque concept in predicting and preventing cardiovascular events.

The literature we searched has limitations. A major potential utility of the vulnerable plaque paradigm is to identify apparently healthy persons (or people with apparently stable plaques) who are at risk for future events. To date, however, almost everything we know about what constitutes a vulnerable plaque is based on studies in patients who have already had an event. In our scope of the literature, only 3 of 242 primary studies were conducted in patients without a known history of cardiovascular disease. In addition, most studies were cross-sectional (including all histopathologic and biomarker studies and 89% of imaging studies) and were mostly done with relatively small sample sizes. Large, prospective studies including patients without previous cardiovascular events over relatively long durations of follow-up are required to validate what we know from retrospective and cross-sectional studies. Such prospective studies would also have to identify the individual vulnerable plaque that develops into a culprit lesion. In the present literature overview, none of the 11 prospective imaging studies that followed patients for clinical outcomes documented that the vulnerable plaque was the one responsible for the clinical event during follow-up. This may be challenging because plaques within a given patient may progress largely independently (55). This adds to the complexity of predicting events from plaque imaging and suggests local effects (for example, physical forces) in addition to any systemic effects influencing plaque progression. The value of the available literature is further limited by the use of imaging characteristics that have not been validated as reliable surrogates for histologic markers of plaque vulnerability (for example, echolucency and plaque deformability).

Future research is needed. Once it is prospectively validated that certain plaque features are independently predictive of a future cardiovascular event, demonstrating the incremental utility of such a concept will be required. In applying the concept in persons who are not considered high risk by current criteria (typical of a primary prevention population), the positive predictive value of plaque vulnerability will be constrained by the prevalence or pretest probability of cardiovascular disease in the screened population.

Finally, once the incremental predictive utility of detecting a vulnerable plaque is established, it will have to be demonstrated that certain treatments (novel or currently in use) in patients who would otherwise not have been candidates will improve outcomes. For example, would a statin reduce the risk for a future cardiovascular event in a patient who does not meet current treatment indications but is found to have a vulnerable plaque on imaging? Would screening for such patients followed by selective treatment be more cost-effective than unselective treatment without screening? In secondary prevention populations, should an index event prompt screening for clinically silent vulnerable plaques? Would detection of such plaques modify the risk stratification for recurrent events? Can preemptive, focal treatment of invasively detected vulnerable plaques be efficacious, or should systemic secondary prevention strategies be tailored on the detection of vulnerable plaques? Such questions will need to be answered before the vulnerable plaque becomes a paradigm routinely considered in the prevention and treatment of cardiovascular disease.