Introduction
Atherosclerosis is the most common form of vascular disease and constitutes the major
cause of death, with 17.5 million related deaths annually (31% of global mortality).1
Atherosclerotic plaque represents the hallmark lesion of atherosclerosis. Most, but
not all,2, 3 acute cardiac events occur in the context of plaque‐related thrombus
formation, with the remainder associated with primary cardiomyopathy or arrhythmogenesis
(primary myocardial electrical instability or ischemia related). Given that advances
in technology allow for imaging, access, and localized treatment of atherosclerotic
plaques, enormous efforts have been devoted to recognizing “vulnerable plaques,” defined
as rupture‐prone (or, more teleologically relevant, event‐prone) plaques and, subsequently,
treating them.4
Despite the above, results of vulnerable plaque‐oriented approaches, in terms of both
personalized risk assessment and guided treatment effects, have generally been poor,
thus raising doubts regarding the validity and usefulness of the vulnerable plaque
concept.5, 6, 7, 8 In the present review, processes associated with decreased stability
and related to the vulnerable plaque hypothesis, along with current and experimental
methods for detecting vulnerable plaques, are discussed. Finally, potential alternative
strategies, representing an amalgamation of differing views of atherosclerosis, will
be explored.
On the Genesis of Atherosclerotic Plaque
Disruption of the endothelial cell‐mediated vessel lumen‐wall barrier is considered
the earliest prodrome lesion of atherosclerosis. The molecular basis of the barrier
lies in intercellular (tight and adherens junctions) and cell/extracellular matrix
(ECM) adhesion, both being necessary for its integrity. Both structures are linked
to the cytoskeleton, thus allowing for cellular adaptation to applied stresses.9 In
fact, these stresses and consequent induced strains (deformations) are sine qua non
stimuli for barrier maintenance, acting through mechanosensor‐mediated signal transduction
pathways (involving actin and integrins).10 Furthermore, endothelial cells lose their
polarized structure and assume a spindle‐like conformation, expressing leukocyte adhesion
molecules.11 Consequently, areas of the vasculature with inherently low transmural
and shear stresses (cause) and strains (effect), such as bifurcation points and the
inner wall of curved vessels (such as the right coronary artery), may begin developing
barrier defects early in life.12
Lipid‐Driven Mechanisms
These breaches of vascular barrier integrity allow for lipoprotein infiltration and
deposition in the intima.13 Uptake by both locally present and recruited macrophages
attempts to prevent their accumulation by means of reverse cholesterol transport.
However, cholesterol does have deleterious cellular effects, causing apoptosis, initiating
inflammation and thus preventing its removal.14 Upon further material accumulation,
an amorphous lipid core (hence the component “athero,” from the Greek ἀθήρα, meaning
gruel) is formed—the infamous necrotic core. Lipid particles residing in the core
undergo enzymatic modifications, becoming oxidized, and trigger the release of proinflammatory
mediators.15 Additionally, oxidized low‐density lipoprotein (LDL) has been found to
trigger necroptosis,16 an alternate form of programmed cell death, originally a defensive
infection mechanism, whereupon cells do not quietly implode, but rather release particles
and messengers inducing proinflammatory activation of adjacent cells.17 Thus, lipid
accumulation and modification strongly enhances both quantitative and qualitative
plaque progression. Activated macrophages and apo‐/necroptotic cells within the necrotic
core produce tissue factor (factor III), a major component of 1 coagulation pathway,
significantly enhancing its thrombogenicity.
Lipoprotein(a) (Lp(a)) is a specialized molecule thought to have arisen recently in
evolutionary history, either in the context of infection control (given its ability
to amass oxidized lipids and draw inflammatory cells to injured vessel walls—normally
caused by infectious agents, not atherosclerosis) or as a surrogate for ascorbic acid
in vascular wall repair.18 However, in the context of atherosclerosis, it assumes
4‐fold deleterious/destabilizing effects by promoting thrombosis, inflammation, cholesterol
delivery to atheromas, and smooth muscle cell (SMC) proliferation. Studies' findings
have, foreseeably, associated Lp(a) levels with vulnerable plaque phenotype.19
Inflammatory‐Immune Mechanisms
Inflammation is also a major event in atherosclerosis natural history and plaque progression,
usually related to decreased structural stability. Both components of immunity, innate
and adaptive, have been found to be activated in atherosclerosis‐related inflammation.
Postulated infectious mechanisms (Chlamydia pneumoniae) also act through this pathway,
although the process soon becomes self‐sustained regardless of pathogen presence.20
Increased local thermal activity has been considered a direct effect of inflammation,
stemming from presence of metabolically highly active immune cells, energy production
uncoupling attributed to ischemia, and exothermic chemical reactions (such as oxidation,
H2O2 dissociation).21 Moreover, given results of theoretical calculations, blood‐wall
friction as well as turbulent flow at the maximal stenosis area also contribute to
localized heating. Increased temperature is not only a potentially measurable result
of atherosclerotic plaque presence, but has also been related to higher blood viscosity
and red cell aggregation. Finally, neovascularization of advanced plaques might also
play a role, on the premise of heating the metabolically more‐inert abluminal area
and core of the plaque, although a cooling effect of blood‐flow–related convection
should be considered.22
The Role of ECM
A second dynamic equilibrium is established regarding ECM production and degradation,2
giving rise to vessel remodeling, that is, the response of the ever changing plaque
to alterations in mechanical factors.23 ECM, comprised of both cellular and acellular
(usually in the form of proteoglycans) components constitutes the initial, theoretically
temporary, container of the necrotic core, preventing its contact with the bloodstream.
Furthermore, the mechanical strength of its luminal part (cap) determines plaque resistance
to stress. SMCs and myofibroblasts are the main producers and maintainers of ECM,
with inflammatory cells and necrotic core stimulating its degradation.
ECM degradation is not limited to the luminal side of the plaque; rather, abluminal
layer involvement is evident, especially in plaques with larger necrotic cores (and,
consequently, more macrophages and inflammation), providing a mechanistic link and
interpretation of the well‐known association between positive remodeling, plaque necrotic
core size, and instability.24, 25 Furthermore, direct deleterious effects of the necrotic
core have been documented, in the form of cholesterol crystals that expand upon formation
and with changes in temperature and pH, rupturing container (foam) cells and straining
the plaque from within (not unlike ice), and pierce the fibrous cap, weakening it
and providing a thrombogenic substrate to circulating coagulation components. In fact,
the relationship between cap thinning and plaque propensity to rupture is not linear,
given that significant increases in deformation and tensile stress, for the same load,
are noted past a certain point (Young's modulus reduction >60%).26
Neoangiogenesis
Neoangiogenesis is a feature of advanced atherosclerosis, appearing late in the natural
history of plaques. The main stimulus for neovessel formation comes in the form of
hypoxia‐inducible factors secreted by hypoxic macrophages deep in the lesion's necrotic
core27 and involves endothelial cells losing polarity and sprouting tube‐like forms
as precursors to fully fledged neovessels. In fact, the anoxemia theory of atherosclerosis
postulates that it is the depletion of ATP in macrophages, attributed to the extremely
high rates of energy taxing cholesterol uptake, which leads to apoptosis, formation
of a necrotic core, and angiogenetic stimuli. The neovessel wall lacks proper structure,
in terms of elastic laminae and smooth muscle cell support, and thus they are leaky,
prone to rupture, and provide the plaque with elements crucial for its continued progression,28
in the form of cholesterol (both lipoprotein and red blood cell membrane derived),
cytokines, and leukocytes, straight to the necrotic core, thus establishing a vicious
positive feedback loop favoring destabilization and, ultimately, rupture.29 Plaque
hemorrhage may thus underlie episodes of accelerated plaque growth.30
The Many Faces of Mineralization
The final act in the course of atherosclerosis as a chronic unresolved inflammatory
process lies in the permanent and secure encasement of its source area in a highly
calcified structure (mineralization), replacing the temporary fibrotic “seal” produced
by chemokine‐attracted intima invading SMCs. The process of vascular mineralization
is now considered active, highly regulated, and akin to bone formation.31, 32 Notably,
vascular calcification, as opposed to bone formation, appears to be triggered, not
inhibited by, inflammation.33 However, in a process reminiscent of myocardial wall
rupture risk following infarction, initial phases of calcification with formation
of calcified specs and nodules (microcalcification) will actually reduce plaque mechanical
strength34 by introducing contact areas between materials (calcified nodules and fibrous
cap components) with different elastic moduli, different behavior under stress, and
thus prone to stress‐induced debonding or caveolation.35 This effect is particularly
prominent in cases of adjacent specs, interpreting central plaque rupture and high
peak stresses in theoretically nonvulnerable (cap thickness >65 μm, or even >100 μm)
plaques.36 The above could help explain the often contradictory findings regarding
calcification effects on plaque stability,2 given that older imaging modalities could
not reliably discern between spotty/nodular calcification and fused, reinforced calcified
structures.
The Divergent Course of Erosion
Plaque erosion mechanisms underlie almost one quarter of all acute coronary syndrome
(ACS) events,37 with rising prevalence, yet they remain largely unresolved.38 Eroded
plaques tend to lack features associated with rupture prone structure, exhibiting
thicker caps, large numbers of SMCs, and smaller necrotic cores, while being quiescent
in terms of inflammation.39 However, significant divergence has been documented in
terms of ECM properties. Both a shift toward collagen III (as opposed to collagen
I usually dominating plaque ECM) and hyaluronan (a proteoglycan subtype) production
has been found—in fact, hyaluronan is almost absent even at rupture sites. Given that
collagen III has more‐elastic properties, compared to its type I counterpart, and
that endothelial cell adhesion to hyaluronan is reduced (especially in larger‐vessel–derived
endothelial cells),40 detachment of cells from their basal membrane becomes possible,
even at lower stresses—or in cases of coronary spasm41 (increased membrane mobility
and weaker attachment). Furthermore, not only is hyaluronan more susceptible to neutrophil‐derived
protease cleavage, but also increased numbers of neutrophils have been detected near
erosion‐prone plaques as well and have been shown to induce endothelial apoptosis.42
Finally, hyaluronan itself has been associated with reduced proliferation and survival
of endothelial cells in cultures and possesses prothrombotic properties (increased
platelet adhesion and accelerated fibrin polymerization).43 Given preponderance of
women, smokers, diabetics, and the elderly among patients suffering plaque erosion,
it could be inferred that certain idiosyncratic features or noxious (metabolic or
exogenous) stimuli alter the usual course of events during atherosclerosis progression,
leading to formation of a less‐ruptured (increased elasticity?), yet more thrombosis‐prone
plaque. The lack of need for structural support with exogenous scaffolding and increased
thrombogenicity may interpret the findings of the recent Effective Antithrombotic
Therapy Without Stenting: Intravascular Optical Coherence Tomography‐Based Management
in Plaque Erosion (EROSION) study, where antithrombotic therapy was a viable option,
compared to stenting, in nonocclusive stenoses causing erosion‐related acute events.37
The enhanced role of neutrophils in plaque erosion—analogous to that of macrophages
in rupture—is thought to be mediated by released strands of DNA, called neutrophil
extracellular traps (NETs), originally sequestering pathogens and disarming them by
means of DNA‐drawn enzymes.44 However, in the framework of plaque erosion, the process,
aptly termed NETosis, leads to a prothrombotic state, by enhancing platelet aggregation
and coagulation and inhibiting fibrinolysis (fibrin mesh constituting another putative
pathogen trap). Indeed, myeloperoxidase‐positive cells have been found in larger quantities
in the blood of patients with eroded plaques and in the overlying thrombi.45
Effects of Plaque Environment—External Stresses
Effects of abnormal exogenous stresses continue in late stages of atherosclerosis.46
Normal laminar flow causes relatively stable transverse and shear stresses, with limited
oscillatory variations.11 Moreover, in addition to cell reorientation, exogenous stress
leads to collagen laminae reorientation, introducing anisotropy of the fibrous cap
and rendering it more prone to rupture.47 Mechanical factors may also, through mechanoreceptors,
lead to a multitude of effects on other cellular processes, such as increased reactive
oxygen species (ROS) production and inflammation.11 In fact, plaque tendency to rupture
at the shoulders, as well as cases of fissuring in other areas of the fibrous cap,
may be interpreted by integrating mechanical factors' effects on biology. As blood
flow is diverged toward the remaining lumen by the upstream plaque edge, a low‐stress
area occurs, prone to loss of cellular architecture and increased lipoprotein and
leukocyte permeability, thus continuously providing fuel for plaque growth and inflammation.
At the shoulder area, however, stresses increase considerably attributed to blood
incompressibility. Thus, not only are ROS production and inflammation stimulated in
the underlying tissue (as a result of high, rather than low, stresses in this area),
but an abrupt transition between mildly and highly stressed areas occurs, favoring
plaque rupture and neovessel formation and rupture.48 Similarly, at the downstream
shoulder, flow turbulence leads to decrease of transmural pressure attributed to vortex‐formation–related
energy dissipation, mirroring the events of the upstream region. Presence of a large
necrotic core exaggerates this unfavorable altering of stress distribution, further
straining the shoulder area.49 At intermediate plaque areas, between shoulders, highly
oscillatory stresses may occur attributed to the Bernoulli effect. Similarly, this
mechanism of blood acceleration through the stenotic lumen and reduction of transmural
pressure exerts detrimental effects on the vasa vasorum with blood content oscillations,
causing plaque expansion and dilation at every cardiac cycle, stressing and building
up fatigue in the fibrous scaffold of the plaque, and potentially causing the structurally
defective vasa vasorum to rupture. High shear stresses could conceivably play an analogous
role in enhancing endothelial cell denudation in plaque erosion. Based on the above,
it is evident that all types of abnormal stress patterns, if sustained, have discrete
adverse effects on atherosclerotic lesions: Low stresses are involved in initial stages,
whereas high and oscillatory stresses come into play later on in disease progression.
Vulnerable Plaque Hypothesis
Ultimately, plaque destabilization (a general term for complication related changes)
is a biomechanical phenomenon that can be thought of as depending on a complex interplay
between applied stresses, structural features, and biological processes that determine
mechanical strength.50, 51, 52, 53, 54, 55, 56 This approach, although not conclusive,
is useful as a framework for pursuing ways to prevent, predict, and avert rupture/erosion.
Obviously, focusing solely on plaque characteristics will ignore the effects that
exogenous forces exert on stability. On the other hand, focusing on luminal stenosis
will only assess the applied forces and disregard plaque contribution, explaining
the oft‐cited finding of far greater prevalence of nonstenotic lesions in cases of
destabilized (ruptured) plaques.57 Furthermore, not all cases of destabilization will
lead to clinical events, given that vessel occlusion, leading to peripheral necrosis
with mechanical or electrical instability, is associated with blood thrombogenicity
and the ability of counter‐regulatory mechanisms to contain its sequelae—prevalence
of silent ruptures in stable coronary artery disease patients has been reported to
reach 58%.58
Plaque rupture is the most common form of plaque destabilization,50, 59, 60 accounting
for two thirds of fatal myocardial infarctions (MIs) and sudden cardiac deaths and
refers to transmural fissuring of the fibrous cap, leading to exposal of the thrombogenic
and proinflammatory underlying necrotic core to circulating blood. The affected plaque
usually has the features of a thin cap fibroatheroma, as described previously61 (Table 1).
On the other hand, plaque erosion denotes a histologically seen thrombus superimposed
on a cell‐denuded endothelial layer of a nonruptured plaque71 with different structural
features, as observed in the alternative course of plaque progression. Finally, calcified
nodules have been found in cases of culprit lesions, as expected by their inherent
disruptive effects on plaque integrity, constituting another rupture‐prone structure,
alternative to thin cap fibroatheroma.50, 59, 60
Table 1
Thin Cap Fibroatheroma Features
Histology61
Cap thickness <65 μm (based on results by Burke et al62 acquired from ruptured coronary
plaques in male cadavers)
Large necrotic core
Increased macrophage infiltration
Virtual histology IVUS63, 64
Focal lesion, containing necrotic core (≥10% of total plaque area) in direct contact
with the lumen (cap cannot be visualized)
Percent atheroma volume ≥40%
Optical coherence tomography65, 66
Wide lipid arc (>90 degrees), suggesting increased lipid content. Lipid arc is defined
the widest arc demarcating a signal poor region with diffuse borders65, 67
Necrotic lipid pools presence (signal‐poor regions poorly delineated, underlying a
signal rich cap), quantified based on number of quadrants occupied
Superficial microcalcifications (subtending a <90‐degree arc and with a border with
lumen <100 μm thick)
Cap thickness <65 μm—although different limits have been proposed, with studies suggesting
that OCT‐derived in vivo thin cap limit should be increased (postmortem/histological
preparation—related alterations in previous studies)68, 69
When virtual histology IVUS is used concomitantly to assess deeper plaque layers,
OCT fibroatheromas are confirmed to have a high lipid content and low levels of fibrosis
and exhibit more‐expansive remodeling.70
Modality‐specific definitions are necessary, given that no single approach can accurately
visualize all aspects of histology, attributed to poor resolution for IVUS and poor
penetrance for OCT. Given complimentary features of all methods, the best possible
assessment of intrinsic instability (see Figure 1) would currently entail application
of multiple modalities on a single plaque or, alternatively, combination of modalities
into multifunctional systems. IVUS indicates intravascular ultrasound; OCT, optical
coherence tomography.
The vulnerable plaque hypothesis was developed in an effort to better describe the
unpredictability of the clinical course of atherosclerosis. Vulnerable plaques have
insightfully been defined as those plaques prone to becoming culprit plaques (Table 2),
causing acute vascular events or death, regardless of form, shape, stenosis, or destabilization.2
Consequently, all forms of rupture‐prone, erosion‐prone, hemorrhaging, or containing
calcified nodules plaques (mainly expressing the intrinsic constituent of culpability
potential), as well as significantly stenotic plaques (mainly expressing the extrinsic
constituent in the form of applied stresses), can be considered vulnerable plaques
as long as they have led to hard clinical events. However, this is a post‐hoc approach,
of uncertain clinical significance. Thus, the term “vulnerable plaque” has, by extension,
been used to denote destabilization or, even more narrowly, rupture‐prone plaques.72
Moreover, difficulties in assessing applied stresses and consequent strains attributed
to requirements of advanced in silico models23, 35, 47, 51 have led to widespread
adoption in the literature of the notion that structural plaque features alone can
define plaque vulnerability, and, consequently, most efforts have been directed toward
correlating imaging features with specific morphologies. However, this may not be
the case, as evident from the pathogenesis of atherosclerosis, and visualized in Figure 1.
Specific criteria, based on event‐causing plaques, have been proposed.2 Henceforth,
the term “vulnerable” will be used for plaques considered unstable based on current
approach, not taking into account the effect of exogenous factors, such as hydrodynamics
(dynamic pressure, pressure head), flow characteristics (laminar/turbulent), and vascular
anatomy and function (eg, bifurcation and tone, respectively). It is thought, based
on tentative studies' results (see below), that destabilization propensity is better
appreciated within the latter, rather than the former, conceptual framework.
Table 2
Culprit Plaque‐Based “Vulnerable” Plaque Characterization Criteria
Expressing rupture propensity
Active inflammation
Thin cap fibroatheroma morphology/yellow color on angioscopy
Fissured plaque
Calcified nodule presence
Intraplaque hemorrhage
Endothelial dysfunction
Positive remodeling
Expressing erosion propensity
Endothelial denudation with thrombogenic proteoglycan substrate±thrombus presence
Endothelial dysfunction
Expressing effects of extrinsic factors
Lumen stenosis over 90%
The above criteria were developed based on autopsy findings of culprit plaques; thus,
it was hypothesized that they could sufficiently and unambiguously define plaques
in high risk for destabilization/events. Based on previous work.2, 3
Figure 1
Inadequacy of modern approaches to detect plaque vulnerability. Conceptual plot visualizing
the combined effects of factors leading to intrinsic and extrinsic plaque instability.
Arbitrary “instability units” are used and the precise form of the relationship is
purely conjectural—however, it was chosen to denote the more‐than‐additive effect
of simultaneous increases in both parameters. The line drawn corresponds to plaques
with a specific internal architecture and demonstrates that their actual instability
(ie, possibility for rupture) is also critically dependent on external stresses applied
to them. Of note, solely intrinsic features may indeed determine a truly destabilization‐prone
plaque, but only at extreme values (thus clinically yielding high specificity and
low sensitivity as potential criteria). Furthermore, imaging modalities rarely assess
all features of vulnerability (in the current sense)—rather, they focus on specific
aspects, such as lipid and calcium content, thus failing in even establishing the
value in the intrinsic instability axis. Moreover, both these parameters vary with
time given the (1) tendency of plaques to alternate between different structural phenotypes
and (2) possibility for alterations in the rheological (dynamic pressure, pressure
head, and viscosity) features of circulation (eg, following removal of an upstream
lesion through successful angioplasty)—thus, a given plaque's position would not remain
fixed on the plot. Clustering could be anticipated to occur in several areas of the
plot, given that, often, intrinsic features affect extrinsic and vice versa (ie, a
thick‐capped fibrous occlusive plaque would be subject to higher external stresses
and thus usually be located in the area enclosed by the red square). Models incorporating
all mentioned aspects would likely yield a much more accurate prediction regarding
possibility for rupture and (should further parameters such as viscosity be added)
acute events (even if silent) and thus guide treatment. Color coding: blue, red: minimal
and maximal instability, respectively.
Currently, plaques are most commonly characterized as vulnerable when possessing some
of the following features: large lipid cores (also laden with tissue factor); thin
adluminal fibrous caps; large inflammatory infiltrates; positive remodeling; neoangiogenesis;
intraplaque hemorrhage; and endothelial denudation (in vivo visible only with advance
optical coherence tomography use).41, 62
Detecting the Vulnerable Plaque
Although the ability of imaging modalities to visualize atherosclerotic plaques has
improved considerably during the past few years, they primarily focus, with few exceptions,
on accurately depicting a still frame of the plaque structure, not the continuous
and variable interaction with its surroundings (plaque‐centered approach). Thus, none
of them has, beyond the context of proof‐of‐concept studies, established criteria
for assessing plaque vulnerability in the broader sense, as discussed previously.
Intravascular ultrasound (IVUS) is 1 of the 2 most widely used invasive coronary atherosclerotic
plaque imaging modalities, the other being optical coherence tomography. Essentially
a catheter‐mounted ultrasound transducer, it follows the same principles regarding
ultrasound generation, signal reception, data processing, and image presentation.63
A grayscale image is generated based on which plaques can be broadly classified as
(their echogenicity compared to that of the fibrous adventitia) soft, intermediate,
calcified, and mixed. Moreover, added modules can enhance tissue characterization
abilities of IVUS, allowing for detection and quantification of different plaque structures—this
is achieved, in broad, by analyzing, in addition to reflected signal amplitude, its
frequency and power. Virtual histology IVUS, iMAP IVUS, and integrated backscatter
IVUS are examples of this, with high reproducibility although they remain proprietary
and use different color‐coding for the same structures.73, 74, 75
Regarding vulnerable plaques, IVUS, and its modular expansions, can reliably assess
plaque burden, expansive remodeling, as well as presence and relative proportion of
necrotic core, calcifications, and neoangiogenesis (contrast enhanced; penetration
depth, 5 mm). Confirmation of greater strains in fibroatheromatous, as compared to
fibrous plaques, has also been provided by compound ultrasound strain imaging (alternatively
known as palpography/elastography).76 However, at its current form (20/40 MHz transducers),
it cannot visualize plaque caps, especially thinner ones (resolution, >100 μm).63
Most prospective studies regarding ability of vulnerable plaques to predict events
used IVUS‐based approaches in their definition of vulnerability.77, 78, 79, 80
Optical coherence tomography (OCT) is the second major coronary plaque invasive imaging
modality that uses near‐infrared light (1.3 μm in wavelength) in a manner analogous
to sonography. Light reflected form plaque structures provides image data whereas
the background effect of scattered light is negated through use of interferometric
techniques. Bright or dark areas occur as a result of constructive or destructive
interference between reflected and reference beams. Given the much smaller wavelength
of light in comparison to ultrasound, OCT resolution is 1 order of magnitude higher
(≈10 μm), allowing for proper visualization of fibrous caps, collagen content (polarization‐sensitive
OCT), macrophages (related to inflammation),81 neovessels (appearing as microchannels),
ruptures, and thrombi. OCT performance, with regard to microcalcifications at the
lower end of the spectrum (<5 μm in diameter), remains uncertain. Consequently, several,
but not all, aspects of vulnerability are visualized. However, this comes at a price,
considering that high scattering of light prevents imaging of plaque structures at
deeper layers (penetration depth, 1.5 mm) and thus proper recognition and estimation
of remodeling and necrotic core size (hence the “lipid arc” concept; Table 1). OCT
has displayed high agreement with histopathology regarding both plaque characterization
and cap thickness.67
OCT limitations include the need for blood displacement, attributed to the high scattering
effect of its components, longer image acquisition times (although newer, frequency
domain analysis‐based processing methods), frequent artifacts, inability to assess
deeper plaque structures, as well as relatively poor discrimination between calcified
areas and lipid core, given that both appear as signal poor areas (with clear and
diffuse borders, respectively).68
A newer form of OCT, micro OCT (μOCT), is currently able to offer axial and lateral
resolution in the order of 1 to 2 μm, by means of advanced frequency domain analysis
and use of broad bandwidth light, approaching histology levels. Thus, events in the
cellular and molecular level, crucial to atherosclerosis development and progression,
are visualized, including leukocyte diapedesis, fibrin strand formation, ECM production,
endothelial denudation, microcalcifications and cholesterol crystal formation, and
penetration of the fibrous cap. Given that this method may, in the future, offer real‐time
in vivo images (as well as 3‐dimensional reconstruction) of the vessel wall at subcellular
resolution, it may prove extremely useful for the classification of atherosclerotic
plaques and identification of the vulnerable ones, replacing classical histology and
scanning electron microscopy, notwithstanding all inherent classical OCT limitations.
Invasive coronary thermography is an approach aiming to detect subtle temperature
increases of the vessel wall in areas of heat production, usually accompanying inflamed
and/or ruptured atherosclerotic plaques.82 Purpose‐built catheters include hydrofoil
and balloon‐based designs, ensuring adequate apposition of the thermistor module on
the vessel wall.83, 84 The latter is furthermore able to ensure temporary lumen occlusion,
thus negating any cooling, convection‐based effects of blood flow, and accentuating
underlying gradients.84 Indeed, studies have shown both an increase in temperature
correlating with presenting disease severity (stable angina<unstable angina<acute
MI), and a correlation with systemic levels of inflammatory mediators, such as C‐reactive
protein.83 In addition, prognostic value of postangioplasty measured plaque temperature
has been demonstrated, given that a difference of ≥0.5°C between plaque and healthy
vessel wall was associated with 41% probability of adverse events (repeat infarction,
recurrent angina, and death) during a follow‐up period of just 18 months.85 Good correlation
with IVUS‐derived indices of vulnerability (positive remodeling) has been reported.82
Despite a sound pathophysiological basis, technical issues have, so far, precluded
the use of thermography for plaque vulnerability assessment.
Near‐infrared spectroscopy (NIRS) utilizes characteristic emission spectra produced
by plaque contents following interaction with photons (wavelength area, 700–2500 nm).
Low sensitivity, in terms of induced response, and high penetration, as compared with
visible light spectroscopy, render this method appropriate for assessing the lipid
content of plaques, especially in cases of positive remodeling, with large, deep‐seated,
lipid‐laden necrotic cores.86 Studies using NIRS have shown that large lipid content,
rather than plaque burden, is associated with thin cap fibroatheroma features.25,
87 Larger lipid core burden has been shown to accurately differentiate between culprit
and nonculprit lesion in ST‐elevation MI patients88 and has been associated with higher
risk for periprocedural myocardial infarction.89 Interestingly, combination with IVUS
may allow for concomitant appreciation of both plaque structure and composition,90
comparing favorably with OCT.87 Despite recent studies linking lipid‐rich, NIRS‐defined
nonculprit plaque presence with a 4‐fold risk for adverse events (all‐cause mortality,
nonfatal ACS, stroke, and unplanned revascularization—excluding those definitely related
to the initial culprit lesion)91 within the first year of follow‐up, it cannot be
inferred whether these were actually triggered by the detected vulnerable plaque (thus
being amenable to preventive stenting) or by other, not assessed lesions (NIRS was
only performed over a vessel segment). The latter could have very well been nonvulnerable
lesions at the examination time, and their progression would reflect the need to reduce
the atherosclerotic burden as a whole, rather than perform localized treatment. Results
from the COLOR and LRP (Lipid Rich Plaque) studies are expected to shed light on the
role of NIRS in guiding nonculprit vulnerable plaque stenting.92, 93
A modified form of NIRS, near‐infrared autofluorescence, involves active stimulation
of lipid components to emit detectable infrared light.94 Combination with OCT has
allowed for better visualization of lipid‐laden necrotic cores and, moreover, accurate
localization of the area within it with the densest macrophage concentration.95
Computed tomography coronary angiography (CTCA) utilizes a computed tomography (CT)
scanner in combination with iodine intravenous dye to visualize the coronary arteries.
Significant advantages stem from its very nature: noninvasiveness, ability for imaging
the whole coronary vasculature, and potential for assessing both vessel wall in addition
to the lumen.96 Advances in technology have allowed for improved image quality with
reduced scan times and radiation exposure. More specifically, multislice (or multidetector)
scanners possess a 2‐dimensional detector array allowing for simultaneous acquisition
of multiple planar images (slices), reducing examination time and possibility for
motion artifacts—those with an array breadth of 160 mm allow for imaging of the heart
in a single beat. Use of softer reconstruction kernels leads to improved soft‐tissue
visualization, however at the expense of spatial resolution.97 Prospective gating
can thus be used in order to synchronize image acquisition with diastole (increased
coronary blood flow). Moreover, advent of dual‐energy monochromatic scanners may overcome
limitations in tissue assessment caused by adjacent intense calcification, by negating
the beam‐hardening effect.98 Finally, dual‐source scanners may complete the examination
in only half a rotation. The related, albeit simpler, technique of calcium scoring
does not use an intravenous dye and focuses primarily on coronary calcium quantification
to predict the risk for subsequent events.
From a clinical standpoint, CTCA has emerged as the best noninvasive imaging modality
in terms of assessing both lumen area and plaque composition,96 throughout most of
the coronary vasculature (vessels >1.5 mm in diameter,96 where the majority of plaques
are located99). Its accuracy in detecting lesion presence is noninferior to IVUS,100
and, consequently, it exhibits an excellent rule‐out ability with a negative prognostic
value approaching 100%, even in prolonged follow‐up.101, 102 Furthermore, although
unable to quantify cap thickness, it is nevertheless able to detect presence of thin
cap fibroatheromas, comparing favorably with OCT.
Regarding vulnerable plaques, CTCA can reliably assess the presence, size, and thickness
of the necrotic core, by grading tissue in Hounsfield units (HU; plaques with large
cores will cause less attenuation and thus have lower unit values).103 Specific high‐risk
plaque criteria have been developed, such as positive remodeling (remodeling index
[RI], ≥1.1, also a surrogate for plaque composition104), (very) low (given that the
threshold for a ≥10% necrotic core is 41 HU) attenuation plaque (LAP; <30 HU), napkin‐ring
sign (a low attenuation core surrounded by a rim of relatively high attenuation—pathogenesis
is still debatable), and presence of spotty calcifications (<3 mm—in accord with the
role of calcium specs in destabilizing plaques; see “The Many Faces of Mineralization”
subsection).96
It is noteworthy that all of its derived parameters regarding plaque characterization
have emerged as significant and independent predictors of events in multiple studies.
For example, a recent coronary CT study, with a follow‐up of almost 100 months,105
confirmed that, in nonobstructive disease, positive remodeling, increased plaque burden,
low attenuation, and napkin‐ring sing presence are all associated with cardiac events.
In addition, a plaque volume of 3 mm3/mm of vessel wall has been identified as an
appropriate cutoff for prediction of adverse events.106 Presence of multiple high‐risk
features has been found to confer a more‐than‐additive risk, as reported in large
studies. More specifically, presence of both positive RI and LAP yielded a more than
22% probability of ACS over an average follow‐up of 27 months, as compared to <0.5%
probability, should both features be absent.107 Presence of 3 high‐risk features (LAP,
RI, and napkin‐ring sign) leads to a 60% survival free from ACS in the first year.108
In the latter study, the napkin‐ring sign had the highest hazard ratio (HR) for occurrence
of events (HR, 5.6).
Coronary artery calcium score (CACS) is an alternative parameter measured by means
of cardiac CT without the need for intravenous dye. Despite the fact that it only
assesses 1 aspect of atherosclerosis, as opposed to CTCA, it has been proposed that
the latter offers additional prognostic information only in patients with intermediate
to high CACS.109 In another study, a combination of traditional risk factors, CACS,
and CTCA yielded an area under the curve of 0.93 for the prediction of major adverse
cardiovascular events (cardiac death, nonfatal infarction, and revascularization)
over a follow‐up of ≈1000 days, significantly higher than that of the CACS—risk factors
combination (0.82; P<0.001).110
The noninvasive nature of CTCA renders it an attractive option for population screening
(primary prevention), and the ability to assess and monitor plaque burden essentially
in the totality of the coronary arterial bed, with incremental improvement over conventional
stratification,111 allows for the pursuit of risk‐factor–reducing strategies as an
alternative to pre‐emptive stenting of high‐risk asymptomatic lesions.112 CTCA‐based
3‐dimensional reconstruction of vessel anatomy and geometry has also been used for
the noninvasive assessment of fractional flow reserve in order to determine hemodynamic
significance of lesions113; however a potentially more‐fruitful approach could be
its integration into fully fledged computational models for the assessment of plaque
stability (see below).
As pointed out in the NIRS section, although these results encourage the use of CTCA
as a means to assess probability for future events, they cannot constitute, in any
way, a proof for the validity of stenting nonculprit vulnerable plaques, given that
they do not prove that they underlay these events.
Magnetic resonance imaging (MRI) has the best ability to visualize the soft‐tissue
component of atherosclerotic plaques, as well as neovessel formation and diffusion
properties (wall permeability). Major limitations include motion artifacts and use
in cases of cardiac implants or devices. Importantly, MRI has been used in animal
studies attempting to prospectively determine vulnerable plaque features specifically
related with (pharmacologically induced) disruption.114 These were related to plaque
remodeling (with positive remodeling and grater plaque area associated with future
rupture) and inflammation indices (markedly increased gadolinium enhancement—denoting
increased neovessel permeability and extracellular space expansion—as in intense apoptosis/necrosis).
Even so, applicability to humans, the time frame between feature presence and (nontriggered)
rupture and relation with clinical events remains unknown.
Molecular imaging involving targeting and visualizing specific components of biological
processes, has been extensively used in attempts to visualize the vulnerable plaque.
Theoretically, any modality may be used for molecular imaging as long as a proper
tracer can be developed, for example, photon‐emitting or possessing paramagnetic properties.
Any substance of interest may act as a target provided that it can be either attached
to or modified as a tracer.
Positron emission tomography (PET) and single‐photon emission computed tomography
(SPECT) are the 2 modalities mostly associated with molecular imaging, often in conjunction
with CT/MRI imaging,115 to provide anatomic background and resolution on top of the
mainly functional results of molecular imaging.
Virtually every atherosclerosis‐related process has been studied by means of molecular
imaging. More important, vulnerability‐related processes have proven amenable to imaging,
including leukocyte adhesion (through involved proteins, such as selectins and vascular
cellular adhesion molecule 1), macrophage content (osteopontin),115 collagen degradation
(labeled matrix metalloproteinase [MMP] inhibitors), cell apoptosis (use of annexin
that binds to lipids exclusively present in the outer layer of apoptotic cell membrane)
or necroptosis (radiolabeled necrostatin, a preferential inhibitor of necroptosis),16
and neoangiogenesis (use of labeled anti–vascular endothelial growth factor antibodies
as tracers). PET has also been experimentally evaluated for the study of intraplaque
hypoxia27 and assessment of macrophage content. Significance of tracer choice can
be perceived in the case of plaque calcification, where PET‐CT use of radioactive
sodium fluoride as a tracer may reveal sites of active calcification, thus differentiating
between stabilizing (encasing) and destabilizing (ongoing) calcification (see above).
This is achieved because of preferential binding of the tracer on hydroxyapatite crystals,
whose exposed surface is much larger in areas of ongoing crystal formation.116
PET is generally favored over SPECT because of lower artifact rates and improved spatial
resolution as a result of detection of 2, as opposed to a single, oppositely moving
photons.
The advent of nanotechnology has further widened the repertoire of imaging probes,
allowing the construction of complex nanostructures (nanoparticles and magnetic nanoparticles),
engineered according to the desired targets.117 For instance, nanoparticles composed
of cross‐linked iron oxide molecules and 19F perfluorocarbon are used to detect macrophages
and in MRI spectroscopy, respectively (the latter attributed to excellent background
noise suppression—no fluoride isotopes exist in tissues). Many of these probes may
double as treatment vectors (theranostics), inasmuch as under the influence of infrared
lasers from a catheter they overheat and release energy to the surrounding tissue,
leading to (localized and targeted) irreparable damage (nanobombs).118
Interestingly, different approaches to optical microscopy (Brillouin microscopy) could
allow for direct estimation of plaque stiffness without need for computational models
(see below). More specifically, they exploit differential energy shift of emitted
photons following interaction with different areas of an anisotropically deformed
(strained) structure (such as a plaque cap). Proof‐of‐applicability studies have been
undertaken in an ex vivo setting, providing encouraging results regarding potential
mounting of Brillouin scattering systems on coronary catheters.
Computational models will allow for simulation of plaque behavior under (patho)physiological
conditions. These include finite element analysis models, assessing forces applied
on a surface though numerical approximation of the solution by dividing the surface
into multiple (>10 000) elementary components with homogeneous properties,52 and fluid‐structure
interaction simulations, also integrating rheological effects on plaque structural
stress. They are the only means able to actually assess true vulnerability by integrating
the plaque structure with its surroundings; however, all of them are critically dependent
on an accurate imaging technique to provide vessel geometry and plaque structural
data as input.23 Intrinsic parameters classically associated with vulnerability (cap
thickness, necrotic core size, vessel anisotropy, and all forms of calcification)
are combined with the anticipated effects of applied stresses, fatigue attributed
to vessel motion during the cardiac cycle, and blood pressure/flow effects to yield
meaningful results regarding conditions leading to rupture/erosion. Should more parameters
be available as inputs and not need to be inferred (such as stain by use of Brillouin
microscopy), results will accordingly be more precise. A major limitation of current
computational models lies in the perceived difference of structural properties between
plaque components in vivo and purified materials used to determine inserted parameters.119
Despite the above, advances in algorithms have led to reduction of necessary time
to construct the model and obtain convergent solutions (ie, functions describing plaque
behavior) from weeks (!) to less than 2 hours.51 The latter is comparable to offline
time needed for 3‐dimensional reconstruction of IVUS and OCT data; thus, it may herald
active application of models to clinical practice in the near future so as to evaluate
the atherosclerotic plaque not as an isolated structure, but as part of a living vessel.
Although relatively scarce, some clinical data exist advocating the added prognostic
value of integrating computational models in clinical practice. More specifically,
even in the negative Prediction of Progression of Coronary Artery Disease and Clinical Outcome Using
Vascular Profiling of Shear Stress and Wall Morphology (PREDICTION) study, wall shear
stress was found predictive of a future decrease in luminal area and clinically significant
obstruction,80 whereas balloon‐induced rupture was noted to occur at the site of the
maximal structural stress in a 2001 study using IVUS to generate a finite‐element
model.53 More important, in patients presenting with ACSs, increased stress values
were noted in high‐risk areas, suggesting a potential cause‐and‐effect relation.54
Obviously, large, well‐designed prospective studies and, subsequently, randomized,
clinical trials are currently lacking and sorely needed to verify or refute the validity
of the above integrative approach.
Questioning the Vulnerable Plaque Hypothesis
Considerable issues are currently being raised with regard to the vulnerable plaque
hypothesis context, especially its clinical significance.6, 7, 120 Despite vulnerable
plaque features presence having high sensitivity in cases of acute events, prospective
studies, such as Providing Regional Observations to Study Predictors of Events in the
Coronary Tree (PROSPECT),79 VH‐IVUS in Vulnerable Atherosclerosis (VIVA),77 and the
European Collaborative Project on Inflammation and Vascular Wall Remodeling in Atherosclerosis
(ATHEROREMO‐IVUS),78 have consistently shown that their positive predictive value
is low, implying low prognostic value on an individual basis, used in addition to
classical risk‐stratification factors. In the same framework, the PREDICTION study
only demonstrated an ability for IVUS and shear stress features to predict plaque
enlargement and lumen narrowing, not acute events.80 Failure of a shear‐stress inclusive
approach to define plaque vulnerability may be related to IVUS resolution, considerably
inferior to that of OCT. Moreover, in the recently presented 2‐year follow‐up of the
COLOR registry,93 NIRS‐determined culprit plaque vulnerable phenotype was not associated
with a difference in major adverse cardiovascular events occurrence (5.4% vs 6.3%
in the nonvulnerable plaques; P=0.41), suggesting that a stented lesion has the same
possibility to cause a new clinical event regardless of its type. In fact, adverse
events during follow‐up were more commonly related to nonculprit lesions; however,
outcomes for the NIRS‐defined vulnerable nonculprit plaques have not yet been analyzed.
In conjunction, the full results from the COLOR and LRP studies (NCT02033694)92 are
expected to definitely answer whether NIRS fares any better in identifying plaques
prone to causing future events.
Recent findings appear to suggest that the notion of nonstenotic plaques underlying
ACSs is a misconception, given that, when having been assessed close to the episode
of which they were the culprit, plaques appear severely stenotic, having undergone
a period of accelerated growth.79, 121 Although vulnerable plaques may carry high
lipid burdens, masked by their expansive remodeling, 1 of the major tenets of the
vulnerable plaque hypothesis is that stenosis degree has little impact on the plaque's
fate, compared to its structural features (although nothing actually precludes it
from being severely stenotic).2 However, this controversy can be reconciled by considering
that plaques with accelerated progression are, in fact, vulnerable plaques, fueled
by repeated episodes of hemorrhage and subclinical (healed) rupture and erosion.122
Studies suggesting plaques may alternate between vulnerable and nonvulnerable phenotypes123
cast doubt on the significance of diagnosing plaque vulnerability and, conversely,
suggest that nonvulnerable plaques may transition into an unstable morphology and
suffer rupture or erosion. The stochastic (by current knowledge), rather than determinate,
nature of plaque progression is indirectly supported by the finding that extensive
nonobstructive atherosclerosis confers similar risk for MI and sudden death as obstructive,
but less‐extensive, disease.124 Additionally, occurrence of clinical events heavily
depends on blood and myocardial features (Figure 2); thus, significance of treating
rupture prone plaques becomes unclear. However, it is not inconceivable that plaque
interdependence may extend well beyond an observed domino‐like effect (a destabilized
plaque releasing mediators knocking others—even in remote vascular beds—out of balance).125
More specifically, flow alterations following a silent event or even a successful
coronary intervention could inevitably, and currently unpredictably, alter the course
of even distant lesions. Finally, the number of nonculprit, asymptomatic, potentially
type‐switching plaques may be such, especially in cases of high atherosclerotic burden5
that renders, in combination with the above, all efforts at localized “vulnerability‐guided”
treatment, without the ability to accurately assess implications futile, if not dangerous.126
Figure 2
Projected natural course of atherosclerotic plaques based on their intrinsic and extrinsic
features. This figure depicts potential evolution of plaques based on an integrative
assessment, including external and internal factors, and underlies the necessary paradigm
shift in the vulnerable plaque definition. Size and thickness of arrows implies relative
probability for a course between the 2 listed in every step. Obviously, the default
state, that is, maintenance of the status quo with plaques remaining quiescent, is
far more likely in all cases—no arrow denotes certainty for an event. As seen in the
diagram, even stable, by all accounts, plaques can rupture/erode. This can be attributed
to a phenotype/external stressor shift, an erroneous assessment, or simply an unlikely
event given that plaque behavior is, as everything in medicine, the sum of probabilistic,
not determinate, events. Additional levels of uncertainty are inserted through the,
currently unpredictable, blood and myocardium response to a destabilizing event. Nonvulnerable
blood and plaque destabilization combination may also lead to events, albeit rarely.
“Silent” events are considered, for the purposes of this figure, both those resolving
previous to thrombus formation and those with thrombosis yet remaining clinically
undetected. “High” and “low” risk structural features can only be defined relatively,
that is, how would the plaque behave in cases of applied loads that can be considered
moderate. Obviously, further research is necessary to elucidate these parameters.
Color coding: blue—initial definition; red—current perception; green—integrative approach.
Double‐headed arrows denote bidirectional processes. ACS indicates acute coronary
syndrome.
Despite the above, the apparent shortcomings of the vulnerable plaque hypothesis are
not attributed to it having been superseded by our ability to prevent, diagnose, and
treat coronary atherosclerosis. Rather, current widespread interpretation and focus
on imaging and still‐frame structural features have significantly diverged from the
original concept, with regard to which we have only covered half the necessary path.46
More specifically, renewed focus on extrinsic effects on plaques and development of
accurate and usable computational models127 will allow for improved assessment of
plaques prone to destabilization of any form. Large, prospective, clinical trials
will be mandated to acquire relevant data to be used in the development of appropriate
vulnerability criteria. Although this will only apply to plaque destabilization, not
hard clinical event prognosis, an accurate model incorporating all additional blood,
myocardium, and patient‐related parameters (indeed replacing the plaque vulnerability
concept) is not expected to be available in the foreseeable future.
Regarding treatment options, it is true that numerous vulnerable plaques usually coexist
in patients at risk for cardiovascular events.128 This, in combination with the domino
effect during an ACS, potential for plaque morphology alteration with regression of
the vulnerable phenotype and presence of further determinants of clinical outcome
(vulnerable blood—vulnerable myocardium) argue against isolating and locally treating
vulnerable plaques, and in favor for implementing systemic atherosclerosis treatment
approaches,129, 130, 131 interpreting presence of these plaques to signify the degree
of its progression in terms of natural history. Exceptions to the above may arise
in cases of nonculprit destabilized plaque detection during an acute event (potential
for domino effect, as suggested by findings of the CvLPRIT [Complete versus Lesion‐only
Primary PCI trial] trial132), presence of accelerated plaque progression with increasing
stenosis (nonetheless requiring close patient follow‐up), and establishment of clinically
applicable criteria for imminent plaque destabilization.114 Thus, in this aspect,
more judicious use of targeted localized therapies is warranted, as indeed suggested
by vulnerable plaque skeptics.
Current and Future Perspectives
In addition to the classical armamentarium, several newer conceptual and systemic
treatment approaches for reducing atherosclerosis burden and preventing its progression
are now being actively pursued.
Autophagy is a cellular process affecting both lipid accumulation and inflammation.
In its kernel, autophagy describes the process through which cells are able to digest
some of their components (ranging in size from signal molecules and lipids to organelles),133
and reclaim usable core materials, such as aminoacids.134 The biological role of autophagy,
somewhat counterintuitive, appears to lie not only in ensuring recycling of damaged
or degraded molecules, especially long‐lasting ones, but also in allowing cells to
regulate processes, such as inflammation, by degrading triggering stimuli, thus ending
their effects.135, 136 Master regulators of autophagy include the mammalian target
of rapamycin complex and AMP‐activated protein kinase.137 Additional autophagy activating
stimuli include endoplasmic reticulum stress (caused by misfolded proteins), ROS,
and modified lipids, such as oxidized LDL.138, 139 Reliance of cellular protein renewal
on autophagy and inhibitory effects of energy and nutrient abundance on the latter
have given rise to the hypothesis that autophagy status plays a crucial role in determining
vascular age (vessels being under continuous stress and in need of replacing worn
components). Research findings have confirmed the above and consequently provided
a plausible basis for the well‐known adage relating reduced caloric intake to longevity.
On the other hand, overstimulation of autophagy ultimately leads to cell death and
plaque destabilization attributed to destruction of necessary structures to ensure
survival.140
In the context of atherosclerosis, dysregulation of autophagy has been described to
affect endothelial cells, SMCs, and macrophages.135 More specifically, endothelial
cells depend critically on autophagy for maintenance of proper function, and the nitric
oxide (NO) system has been shown to rely on presence of normal autophagy levels (increases
in endothelial NO synthase).141 Contact of endothelial cells with components of the
ECM induces autophagy, and the latter may inhibit neoangiogenesis.142 Effects of autophagy
on smooth muscle include inhibition of cell death following exposure to lipid peroxides,
thus stabilizing lesions.138 The most significant effects of autophagy on atherosclerosis,
however, are related to macrophages. In addition to the established cytoplasmic cholesterol
efflux, formation of autophagosomes encircling lipid droplets allows for a second
pathway contributing to the removal of lipid material from plaques.143 Interestingly,
inflammatory and lipid mediators (interleukin‐6 and oxidized LDL), have been shown
to inhibit macrophage autophagy in a vicious circle.144 Indeed, autophagy inhibition
and stimulation, often in tandem with apoptosis stimulation and inhibition, have been
shown to promote and arrest, respectively, atherosclerotic plaque progression.145
microRNAs (miRNAs) are small, noncoding RNA molecules that act as post‐transcriptional
regulators of gene expression.146 In atherosclerotic plaques, alterations in miRNA
signatures have been found in all major involved cell types, with macrophage and endothelial‐cell–derived
miRNAs promoting plaque destabilization (through promotion of inflammation and angiogenesis,
respectively).147 Conversely, SMC miRNA profile may be modulated to increase plaque
cap integrity, and thus stability, by either switching them to a less‐synthetic, more‐contractile
phenotype, improving plaque mechanical strength, or inducing increased production
of collagen to increase fibrous cap thickness.148 Modulation involves both upregulation
(by administration of miRNA mimics) and downregulation (through use of antagomirs
and miRNA sponges). Notable, the miRNA approach is the only one able to address all
aspects of plaque destabilization attributed to miRNA involvement in all phases of
atherosclerosis, efficiency of the miRNA gene expression control mechanism, and the
potential to include multiple miRNAs in a single vector. Finally, yet other miRNAs
have been shown to act by affecting autophagy149 and be upregulated in areas of increased
endothelial shear stress. Despite several more‐promising in vitro and animal studies,150
clinical application of miRNA manipulation is currently impeded by lack of adequate
vectors and off‐target effects.
The notion of cell senescence is also being strongly implicated in atherosclerosis
in current studies. In principle, senescence is a ubiquitous biological process underlying
not only biological ageing and tissue degeneration, but tissue remodeling as well
(removal of differentiated cells with limited functional potential and replacement
by other from the progenitor cell pool).151 Thus, the established fact of (biological)
age being the strongest predictor and risk factor for vascular disease is put into
perspective. Broadly, cell senescence can be classified as replicative (associated
with telomere shortening), stress‐induced premature senescence (not associated with
telomeres per se, but exhibiting the same biochemical changes arresting cell‐cycle
progression), and oncogene induced, with only the former 2 involved in atherosclerosis
progression. In addition to changes in telomeres, chromatin structure (switch to heterochromatin—transcriptionally
quiescent), and inhibition of cell‐cycle completion promoting proteins (mainly cyclin‐dependent
kinases), senescent cells display a characteristic secretory profile, including specific
cytokines.152, 153 Although, in principle, a method to promote senescent cell clearance
by phagocytosis (efferocytosis) and replacement, long‐term persistence of this secretory
phenotype leads to disproportional proliferation, angiogenesis, endothelial dysfunction,
and inflammatory cell recruitment.151
In general, the role of senescence in atherosclerosis and plaque vulnerability appears
to be related to both exhaustion of replicative potential of cells attributed to constant
need for replenishment of disease‐related losses and premature induction of functional
handicaps (cell‐cycle arrest) not only precluding self‐renewal (autophagy and synthesis
of components), but also inducing a noxious secretory phenotype. Senescence has been
found to occur in both SMCs and endothelial cells, rendering them dysfunctional. Verily,
presence of senescent SMCs has been linked to atherogenesis and presence of unstable
plaque features, such as large necrotic core and thinner fibrous cap (potentially
mediated by senescence‐related cytokines),154 whereas senescent endothelial cells
are characterized by both NO synthase uncoupling and over‐response to inflammatory
stimuli.155
Significant advances in the understanding of inflammation have led to the notion that
inflammation abatement is an active, rather than passive, process.156 Specialized
proresolving mediators (SPMs) are a group of structurally related signaling molecules
derived from polyunsaturated fatty acids acted upon by 1 or a combination of lipoxygenase,
cyclooxygenase, and cytochrome P450 monooxygenase enzymes (thus nonselective inhibition
of lipoxygenase may not always be beneficial). Main subclasses are resolvins, maresins,
protectins, and lipoxins. More broadly, they are considered autacoids (αὑτός+ἀκω,
“self‐heal”), hormones produced, acting, and catabolized locally. It appears that
they constitute the second part of a negative feedback loop, being produced by inflammatory
cells and acting on them and other cells in the tissue vicinity to promote healing,
including efferocytosis stimulation,157 and quench proinflammatory stimuli. Moreover,
SPMs initiate a positive feedback loop for the firm induction and progression of the
reparatory process. In atherosclerotic plaques, in accord with the concept of chronic
inflammatory activation, plaque progression and assumption of a vulnerable phenotype
are both related to increased levels of proinflammatory lipid mediators (such as leukotriene
B4) and decreased levels of SPMs,158 evidencing that, in chronicity, inflammation
has gone awry. Thus, inflammation control should involve not only suppression of promoting
pathways, as is the current treatment paradigm in atherosclerosis, by means of medications
(eg, statins), but also reinforcement of those promoting healing and dampening of
inflammatory responses. Recent research has established that exogenous administration
of a cocktail of SPMs does indeed lead to plaque stabilization by halting further
engorgement of the necrotic core, leading to increases in the numbers of SMCs and
cap thickness and altering macrophage phenotype to a reparatory one (M2).158
Further pathways of interest for research and putative significance for the natural
history of atherosclerosis have been identified. Pyroptosis is among them and refers
to a form of proinflammatory cell death of immune cells, following detection of foreign
antigens within, whereupon cells commit suicide while releasing immune‐stimulatory
molecules.159 Differences from apoptosis include dependence on caspase 1 activation
and formation of the inflammasome (as opposed to caspases 8 and 9 and apoptosome,
respectively). The difference from necroptosis lies in the occurrence in immune (as
opposed to tissue) cells. Studies have shown that intracellular oxidized LDL may act
as an inductor of macrophage pyroptosis, through a mitochondrial‐DNA–dependent pathway.
Although apoptosis, necroptosis, and autophagy have been reported in atherosclerosis,
it is entirely possible that pyroptosis of immune cells is also involved (and thus
apoptotic foam cells are, at least in part, actually pyroptotic cells), more dynamically
contributing to plaque destabilization.160 On more‐conventional grounds, generalized
immunosuppressive treatment has been implemented with favorable results whereas therapies
targeted to specific components (eg, inhibition of MMPs) have yielded encouraging
results.
Conclusions
The concept of vulnerable plaque was developed with the intent of formally defining
and describing atherosclerotic plaques having caused acute clinical events (a posteriori).
Need for clinical applicability led to definition shifting toward plaques prone to
causing events and by extension prone to destabilization (rupture, erosion). Most
subsequent clinical studies further narrowed the scope by focusing on intrinsic (structural)
determinants of instability, such as quantitative and qualitative composition and
remodeling. Attributed to the neglect accorded to effects of extrinsic stressors (eg,
dynamic pressure/pressure head, turbulent/laminar flow; see Figure 1), and the routine
use of a single imaging modality in plaque assessment, this led to plaques considered
nonvulnerable, under the current concept, to be found causing events.
Despite the above, the vulnerable plaque concept has led to tremendous advances in
our understanding of pathogenesis and treatment of atherosclerosis, along with substantial
improvement of the relevant technological armamentarium. Results from newer studies
and trials appear to suggest that a change of approach to the vulnerable plaque concept
is mandated to further improve patient diagnosis and treatment. However, a return
to the fundamentals of the vulnerable plaque hypothesis appears more appropriate (Table 3).
Table 3
Vulnerable Plaque Conundrum
What it originally meant
A plaque that is the culprit lesion for an acute coronary event
What it should mean
A plaque that is prone to rupture when all intrinsic and extrinsic effects are taken
into account (regardless of structure)
How it is currently interpreted in the majority of literature
A plaque with specific morphological features—usually referring exclusively to thin
cap fibroatheromas
This table attempts to summarize the different connotations inherent in the term “vulnerable
plaque.” It can be appreciated that a teleological approach has shifted toward a utilitarian
one, allowing for easier classification of plaques as “vulnerable” or not, yet depriving
this characterization of prognostic implications.
In particular, improvement in the assessment of the complex interplay between intrinsic
and extrinsic factors mentioned above (Figure 1) by use of advanced computational
models is crucial in more accurately determining which plaques are truly vulnerable161
and thus allow for a more‐accurate estimate of their “natural future” (Figure 2).
Their detection should be interpreted as a marker of advanced atherosclerosis progression,
in the context of such systemic alterations as deranged autophagy, SPM and noncoding
RNA regulation, and treated both aggressively and systematically.
The fundamental question whether and under what circumstances localized treatment
of plaques should be pursued is further complicated by two factors: The ability of
plaques to alternate between different structural phenotypes (thus changing their
location in the Figure 1 curve) and plaque interdependence, as described above. It
could be safely said, based on large trials' evidence, that treating current (nonculprit)
vulnerable plaques does not affect disease progression and patient course. However,
whether this will hold true in the future, with integrative approaches, incorporating
advanced computational models, and changing the working definition of vulnerability,
cannot be at present determined, pending development of such tools and related prospective
trials.
The vulnerable blood and vulnerable myocardium concepts should be brought back into
focus to allow for assessment of the vulnerable patient by determining which plaque‐blood‐myocardium
combinations are prone to leading to catastrophic events (that may actually be arrhythmic
rather than ischemic). On the other hand, not all silent events are innocent: It is
conceivable that they might alter, by mechanisms discussed previously, the course
of other plaques, in the long term—although at present this is merely conjectural
and hypothesis generating.
In conclusion, although the vulnerable plaque is a valid concept in principle, holding
great promise for future research, current interpretation of (incomplete) data for
vulnerability assessment has rendered it inaccurate. Resources could be more effectively
redirected toward large‐scale efforts to reduce the epidemiological atherosclerosis
burden, for which there is solid evidence for improving life expectancy and quality,
and also toward developing more‐sophisticated models for an integrative approach to
plaque assessment. In any case, it should be kept in mind that “preventing is better
than treating.” Thus, guided localized treatment of plaques, rather than systemic
approach to patients, is, in the great majority of cases and given lack of reliable
simulation models, not advisable at present for reasons of effectiveness and safety.
Disclosures
None.