Imaging Cardiovascular Calcification

Introduction

Cardiovascular disease is the leading cause of morbidity and mortality worldwide. Atherosclerosis is a complex and multifactorial process, characterized by early asymptomatic formation of plaque in the arterial walls, silent plaque progression that poses flow limitation, and risk of sudden rupture and subsequent thrombotic occlusion.1 The early assessment and predictive value of atherosclerotic plaque development and rupture are critically important to prevent clinical consequences.

Inflammation and calcification play important roles in the pathophysiological progression of atherosclerosis. In the early stage, inflammation is the predominant process involved that promotes plaque progression and calcification.2 The repeated cycles of inflammatory damage and repair ultimately lead to calcification of the vessel walls,3 whose extent provides an important estimate of atherosclerotic progression and clinical prognosis. Macrocalcification is considered the stable stage of vascular disease, whereas the dynamic process of microcalcification may increase the risk of plaque rupture and adverse clinical events.4 Therefore, imaging early calcification is important to identify individuals with enhanced risk of cardiovascular events. Furthermore, many studies have demonstrated coronary artery calcification as a marker for the extent of atherosclerosis and overall plaque burden, and as an important independent predictor of cardiovascular events, mortality, and morbidity.5, 6, 7, 8 Over the past few decades, advanced imaging methods have improved our assessment of atherosclerotic calcification and have provided insights into cardiovascular risk and have contributed to the understanding of the underlying pathophysiologic mechanisms of atherosclerosis.

This review aims to summarize current imaging techniques for vascular calcification (Table) and their applications in clinical research.

Table 1.

Current Imaging Techniques for Vascular Calcification

Approach	Advantages	Disadvantages	Macrocalcification	Microcalcification	Invasive
IVUS	Directly image the vessel wall. Ability to detect hyperechoic calcified plaque	Limited axial resolution. Limited ability in accurate assessment of plaque composition. Inability to detect microcalcification	✓		✓
CACS	Provide a simple, rapid and reliable quantification of macroscopic calcium, and the total coronary calcification burden	Lower resolution and poor tissue contrast. Inability to provide information on plaque morphology or subtype	✓
CCTA	High spatial and temporal resolution. Ability to estimate the severity of coronary stenosis and provide detailed information on plaque morphology	Requires the administration of contrast agent. Difficulty in identifying and quantifying calcium in the presence of iodine in contrast media	✓
MRI	Superior soft‐tissue resolution. Lack of ionizing radiation. Multipulse sequences can depict the location and volume of vascular calcification	Prolonged acquisition time. Motion artifacts during the cardiac contractions and respiration. Inability to identify microcalcification	✓
OCT	High spatial resolution. Capable of assessing fibrous cap thickness, macrophage filtration, the thickness and border of vascular calcification	Difficult to differentiate calcium and lipid pool. Limited tissue penetration	✓		✓
Invasive angiography	Superior spatial resolution and temporal resolution. The criterion standard for coronary atherosclerosis imaging	Inability to provide direct imaging of calcification or the atherosclerotic plaque itself			✓
18F‐NaF PET	Exquisite sensitivity. Capable of identifying microcalcification and the vulnerable plaque	Relatively lower spatial resolution. Continuous cardiac motion prevents accurate quantification of coronary artery microcalcification		✓
18F‐FDG PET/CT	Attenuated‐CT can visualize the bulky macrocalcification	Background myocardial uptake limited its assessment of coronary arteries	✓

CACS indicates coronary artery calcium score; CCTA, coronary computed tomographic angiography; CT, computed tomography; FDG, 2‐deoxy‐2‐fluoro‐d‐glucose; IVUS, intravascular ultrasound; MRI, magnetic resonance imaging; OCT, optical coherence tomography; PET, positron emission tomography.

Atherosclerosis Development, Calcification, and Progression

Atherosclerosis exists as a pathological continuum. The persistence of vascular risk factors promotes dysfunction of vascular endothelium and increases its permeability, allowing apolipoproteins to enter the intima and become oxidized. The formation of oxidized apolipoprotein and the expression of cellular adhesion molecules, monocyte chemotactic protein 1, and other chemokines from the endothelium induce monocytes to migrate into the intima and differentiate into macrophages. These macrophages secrete additional inflammatory cytokines and extracellular matrix molecules, while engulfing oxidized lipoproteins to form lipid‐laden foam cells. When the amount of oxidized lipoproteins engulfed by macrophages exceeds their clearance capacities, the foam cells undergo apoptosis and release lipoproteins and other cellular contents, which lead to the formation of an extracellular lipid core and contribute to lipid‐rich plaque accumulation within the arterial wall.9 Over time, these atheromatous plaques undergo a series of processes (eg, hypoxia, neovascularization, and microcalcification).

Vascular calcification is a complex, organized, regulated, and active progress, much like the formation of bone. In fact, macrophages in atherosclerotic plaques promote osteogenic differentiation by releasing pro‐inflammatory cytokines (eg, IL‐1, IL‐6, IL‐8, and tumor necrosis factor‐α). The resulting microcalcification crystals initiate a positive feedback loop by further stimulating the pro‐inflammatory response of macrophages, thereby propagating the pro‐calcific stimulus in the vascular wall.10

Vascular calcification can be classified into 2 distinct forms, depending on its location within the intima (intimal calcification) or in the vascular medial (medial calcification) layer. Calcification of the arterial vessel's intimal and medial layer is presumed to have a different pathogenesis and clinical consequences.11 The intimal layer consists of endothelial cells that undergo a series of processes and eventually form atheromatous plaques that can cause plaque rupture and subsequent thromboembolic events,12 whereas the medial layer consists of smooth muscle cells and elastic fibers that can regulate blood flow and arterial pressure. Calcification of the media is thought to cause arterial stiffening, reduce compliance, and limit distensibility. Presently, ex vivo histological analysis is the criterion standard to distinguish intimal and medial calcification.11

Vascular calcification occurs in a 2‐phase process: an initial stage of microcalcification and the subsequent stage of macroscopic calcium formation (macrocalcification).13 Although there is no conventional standard of size, there is general consensus that categorizes microcalcification and macrocalcification based on nodules of <50 and ≥50 μm, respectively.14

Microcalcification, a clinically more significant manifestation of vascular mineralization, represents the early stages of intimal calcium formation and greatly amplifies mechanical stresses on the surface of the fibrous plaque that may directly contribute to its rupture.4, 14 Detection of microcalcification is not possible with current clinical computed tomography (CT) systems, which are only able to identify large areas of macrocalcification of 200 to 500 μm in diameter.15 However, these structures can now be detected noninvasively using molecular imaging. Indeed, because of preferential adsorption of fluoride to areas of microcalcification, ¹⁸F‐NaF is the only currently available clinical imaging platform that can noninvasively detect microcalcification in active unstable atheroma. The preferential binding of ¹⁸F‐NaF to microcalcification is because of the high surface area of hydroxyapatite in these nanocrystalline areas.16 Irkle et al17 used autoradiography, whereby dispersion in large tissues and the lack of physical barriers increased adsorption of ¹⁸F‐NaF to microcalcifications.

With progressive calcification, plaque inflammation becomes pacified with its necrotic core walled off from the blood pool. The latter stages of macrocalcification are, therefore, associated with plaque stability and lower risk of plaque rupture. Macrocalcifications have a large volume but small surface area; much of the hydroxyapatite is internalized and not available for binding with ¹⁸F‐fluoride.16 Moreover, ¹⁸F‐NaF cannot penetrate into its deeper layers.14, 17 ¹⁸F‐NaF only bound to the outer surface of the macrocalcifications. Thus, ¹⁸F‐NaF cannot characterize macrocalcification exactly. However, morphologic calcific imaging techniques such as CT, ultrasound, and magnetic resonance imaging (MRI) are readily available to identify the macrocalcification because of their superb spatial resolution.

An illustration of this pathophysiologic progression and current imaging techniques for vascular calcification are portrayed in Figures 1 and 2.16, 18, 19, 20, 21, 22, 23

Figure 1.

The link between inflammation, microcalcification, and macrocalcification. A large necrotic core, a thin fibrous cap, and an intense inflammation are key precipitants of acute plaque rupture and myocardial infarction. Intimal calcification is thought to occur as a healing response to this intense necrotic inflammation. However, the early stages of microcalcification (detected by ¹⁸F‐fluoride positron emission tomography) are conversely associated with an increased risk of rupture. In part, this is because of residual plaque inflammation and in part because microcalcification itself increases mechanical stress in the fibrous cap, further increasing propensity to rupture. With progressive calcification, plaque inflammation becomes pacified and the necrotic core is walled off from the blood pool. The later stages of macrocalcification (detected by noninvasive imaging techniques such as computed tomography [CT] and magnetic resonance imaging [MRI], and by invasive imaging techniques such as intravascular ultrasound and optical coherence tomography) are, therefore, associated with plaque stability and a lower risk of that plaque rupturing (Illustration credit: Ben Smith). Reprinted from Dweck et al16 with permission. Copyright ©2016, Wolters Kluwer Health, Inc. Promotional and commercial use of the material in print, digital or mobile device format is prohibited without the permission from the publisher Wolters Kluwer. Please contact permissions@lww.com for further information.

Figure 2.

Multimodality imaging of cardiovascular calcification. Representative illustration of current and emerging calcification imaging multimodalities. Each modality offers unique measurements of calcification. Together, they offer the molecular, anatomical, and functional imaging of calcification, which can be used to make sense of current calcific activity, the procession of atherosclerosis, and overall disease burden in patients. A, Grayscale intravascular ultrasound (IVUS) image demonstrating a heavily calcified plaque. B, Integrated backscatter‐IVUS image demonstrating 2‐dimensional color‐coded map (red: calcification, yellow: dense fibrosis, green: fibrosis, blue and purple: lipid pool). C, Virtual histology IVUS image demonstrating coronary plaque with dense calcifications (white color) with corresponding grayscale IVUS image. D, Optical coherence tomography (OCT) image demonstrating coronary arterial calcification (arrows indicate well‐demarcated calcification). E, Pathogenic processes demonstrating the atherosclerotic plaque, including lipid core and calcification. Two forms of calcification can be seen: microcalcification and macrocalcification. Each form of calcification is linked with a related visual imaging modality. Green arrows link macrocalcification with many imaging techniques (including computed tomography [CT], magnetic resonance imaging [MRI], [IVUS], and [OCT]), which can visualize the macrocalcification in plaque, whereas the red arrow links microcalcification with the ¹⁸F‐NaF positron emission tomography (PET) imaging technique, which can visualize the microcalcification in plaque. F, ¹⁸F‐NaF (PET)–CT image demonstrating high tracer uptake (red arrow) in the left anterior descending artery culprit lesion revealing active plaque microcalcification and no tracer uptake (white arrow) in the nonculprit lesion. G, Three‐dimensional (3D) Isotropic‐Resolution Black‐Blood MRI (3D‐MERGE) image demonstrating clearly calcification of right carotid plaque (arrow). H, Transverse contrast‐enhanced coronary CT angiography image demonstrating an area of calcium (white arrow) in the left anterior descending coronary artery. I, Non‐contrast‐enhanced calcium scoring image demonstrating an area of calcium (white arrow) in the left anterior descending coronary artery. A and C, Reprinted from van Velzen et al18 with permission. Copyright ©2011, Wiley. B, Reprinted from Kawasaki et al19 with permission. Copyright ©2015, MDPI AG, Basel, Switzerland. D, Reprinted from Batty et al20 with permission. Copyright ©2016, Wiley. E, Reprinted from Tarkin et al21 with permission. Copyright ©2014, Wiley. F, Reprinted from Joshi et al22 with permission. Copyright ©2014, Elsevier Inc. G, Reprinted from Balu et al23 with permission Copyright ©2010, Wiley.

Multimodality Imaging Techniques

Computed Tomography

CT is the most established noninvasive tool to detect coronary artery calcium (CAC). It can provide a simple, rapid, and reliable quantification of macroscopic calcium24 and generate a clinical CAC score (CACS). In the 1990s, quantification of CAC on non‐contrast‐enhanced ECG gated CT examination became recognized as an important noninvasive imaging technique for identifying coronary atherosclerosis.25

CACS relies upon the high attenuation coefficient of calcium to detect calcified plaques.26 The standard method of scoring is the Agatston method.25 The Agatston score considers total calcified area and maximum density of calcification (>130 Hounsfield units [HU]) to provide a summed score of all calcified lesions. Although other methods of scoring have been used, the Agatston score remains the criterion standard because of its simplicity.

CACS provides quantification of the total coronary calcification burden and has become a useful clinical tool for risk stratification in asymptomatic patients with low‐to‐intermediate and intermediate risk of cardiovascular events.27, 28 It predicts future cardiovascular risk beyond the traditional Framingham risk score alone.29, 30, 31, 32 An elevated CACS may portend an increased risk of cardiovascular events in some patients.33 Specifically, the MESA study (Multi‐Ethnic Study of Atherosclerosis) with 6722 patients followed for an average of 3.8 years showed that those with CACS >300 have a nearly 10‐fold increased risk of coronary event across ethnic groups.34 Shah et al35 investigated the prognostic significance of calcified plaque among symptomatic patients with nonobstructive coronary artery disease. Their results also showed that among patients with detectable mild luminal stenosis, 4‐year mortality rates ranged from 0.8% to 9.8% for CAC scores of 0 to ≥400, but among patients with no luminal stenosis, CAC was not predictive of all‐cause mortality. Moreover, Mori et al36 demonstrated that CAC associates with cardiovascular events in a stepwise fashion, providing meaningful risk stratification for future coronary events with the additional knowledge that a low CACS is associated with low risk for cardiovascular mortality, which was also supported by the study of Sarwar et al37

However, Puri et al38 observed that dense calcium deposition on CT imaging seems to indicate plaque stability, and it progresses with age. As such, Nakahara et al39 regard that dense calcification (>400 HU) is usually associated with stable plaques, although coronary calcification is a marker of coronary atheroma. Thus, a high CACS is perhaps more useful as a marker of vascular disease burden than as a predictor of the likelihood of a cardiovascular event from an individual plaque, which was reinforced by studies40, 41 showing that patients with stable disease were more heavily calcified than those with unstable disease. Furthermore, a recent investigation of 4425 patients evaluated composition of the plaque and showed that 1021 had only calcified plaques,183 had only noncalcified plaques, and 685 had both calcified and noncalcified plaques. Among these groups, the incidences of cardiovascular disease (CVD) events over a median follow‐up of 3 years were 5.5%, 22.7%, and 37.7%, respectively.8

Furthermore, statin is the treatment choice to lower total cholesterol and low‐density lipoprotein and has been extremely successful in primary and secondary prevention of CVD.42 In spite of achieving the target reduction in low‐density lipoprotein, many clinical trials43, 44, 45, 46, 47 and a meta‐analysis48 have shown that statin therapy increased the progression of CAC compared with placebo in the absence of a higher event rate.43 In both primate49 and swine50 models, anti‐atherosclerotic interventions are associated with an increase in vascular fibrous tissue and calcification. Calcium deposition continues during the initial phase of plaque regression because of the death of foam cells and an increase in necrotic tissue. Thus, vascular calcification may have a key role in the initial stabilization of atherosclerotic plaques. Many researchers51, 52, 53, 54 now agree that statins likely have their salutary effects on CVD risk by reducing the lipid core in unstable plaques and activating plaque repair and healing by increasing the replacement of the lipid core with fibrosis and calcification. Eventually this process increases calcium density in such plaques and yields a reduction in CVD events through plaque stabilization.

Several hypotheses have arisen in an effort to explain these seemingly divergent findings. Specifically, Pugliese et al55 hypothesized that the discrepancy between CACS and cardiovascular risk might be related to the fact that CT imaging does not have the ability to differentiate between quiescent and active calcification. Cowell et al56 hypothesized that the morphologic imaging of CT better characterizes the structural aspects of calcification in atherosclerotic plaques than the pathological process of calcification. Other studies of culprit plaques4, 57 have also concluded that CT imaging is only able to identify macrocalcification and not the higher‐risk microcalcification that occurs at earlier stages of disease.

Additional work has refined the clinical implications of CAC imaging findings. Criqui et al58 reported that the use of the Agatston score in assessing CAC progression is problematic since an increase in CAC could be caused by an increase in volume, an increase in density, or both. So despite the studies suggesting a strong predictive value of CAC for CVD, there has been little rigorous comparison of what specific measure of CAC is most predictive.59 In order to clarify the clinical significance of these factors, Criqui et al58 conducted a multicenter, prospective observational study of the MESA cohort to determine the independent associations of CAC volume and CAC density with incident CVD events. Their results demonstrated that CAC density was inversely related to CVD events for a given CAC volume and that CAC volume was more predictive and was positively and independently associated with coronary artery disease and CVD risk when adjusted for CAC density. Thus, CAC may not be a monolithic unit, as is commonly conceived, and patterns of CAC (eg, spotty calcification versus more coalesced calcification) or its density may have different meanings than the lone number (the Agatston CAC score).60

Relative advantages of CACS performed by non‐contrast‐enhanced CT include ease of performance and interpretation, high reproducibility, low cost, and a lack of need for an intravenous contrast agent.61 However, there are important limitations of CACS including that it generally only represents ≈20% of total plaque volume, it cannot be used to determine if there is flow‐limiting stenosis, and it does not provide information on plaque morphology or subtype because of its lower resolution and poor tissue contrast.62 CACS CT imaging is also unable to detect calcification accurately at a molecular level33 and does not discriminate between microcalcification and macrocalcification, potential markers of vulnerable and stable atherosclerotic plaques, respectively.63, 64

Contrast‐enhanced coronary CT angiography scans offer improved resolution and tissue contrast, which facilitate an improved evaluation of coronary anatomy and allow an assessment of the severity of luminal stenoses. Moreover, this technique provides information about the morphology and composition of plaques, including high‐risk features such as low‐attenuation plaque, spotty calcification, and “napkin‐ring” sign.65, 66 Spotty calcification—the presence of small, scattered foci of calcium (≤3 mm in diameter) —presents a higher risk of plaque rupture in comparison to single, large deposits of the same total cross‐sectional area,67 and is more frequently observed in the culprit lesions of patients with acute coronary syndrome.57, 65 However, contrast‐enhanced coronary CT angiography is limited by the presence of iodine in contrast media, which makes it difficult to precisely identify and quantify CACS.61

In summary, multidetector CT is a useful first‐line diagnostic technique to estimate the macrocalcification burden. Nonetheless, detection of microcalcification is not possible with clinical CT systems because of their low sensitivity and resolution. Nevertheless, current clinical guidelines suggest using noncontrast CT imaging for CACS to provide clinical risk assessment in appropriate asymptomatic patients and to support the use of CAC screening to guide statin treatment decisions in some patients.68

Magnetic Resonance Imaging

Since 1985, magnetic resonance angiography has been used in clinics.69 Cardiac magnetic resonance angiography allows for a noninvasive assessment of the coronary anatomy without exposing patients to radiation, with excellent soft tissue resolution, and is superior to contrast‐enhanced coronary CT angiography for the evaluation of luminal narrowing in heavily calcified coronary segments.70

However, vascular calcification is diamagnetic and with only a few protons present, visualization is poor with conventional MRI sequences.71, 72 In past decades, there has been considerable progress in MR scanning technology and parameters, including the development of multicontrast MR protocols.73, 74, 75 Furthermore, image acquisition time, success rate, the demonstration of calcification, and the diagnostic accuracy of cardiac magnetic resonance angiography have all steadily improved.76, 77, 78, 79, 80

However, assessment of the coronary arteries by cardiac magnetic resonance is still challenging, because of the small size of the vessels, prolonged acquisition time, and complex motions caused by cardiac contractions and respiration.81 The carotid artery is large, superficial, and clinically important as a major source of ischemic stroke. Moreover, the carotid arteries can be imaged using phased‐array coils and well‐tested multicontrast imaging protocols such as bright‐ and black‐blood techniques.82 All of these make the carotid artery the most commonly assessed vessel in MRI studies of atherosclerosis.23, 83, 84, 85

Fabiano et al86 investigated the utility of multicontrast MRI in identifying calcified carotid plaque with an accuracy and specificity of 98% and 99%, respectively. Moreover, Mujaj et al87 compared CT‐ and cardiac magnetic resonance–based volumes of carotid artery calcification and established that they were highly correlated with each other. However, cardiac magnetic resonance–based calcification is systematically smaller than those obtained by CT. Despite this difference, both modalities provided comparable clinical information about a history of stroke. The study of Yang et al88 also demonstrated that the calcification seen in multidetetor computed tomography (MDCT) is clearly depicted in MRI images at the same location and with roughly the same size in the femoral artery. Baheza et al89 thought that cardiac magnetic resonance underestimated the amount of calcification, because a certain amount of calcification is required before the MR signal disappears, and possible microcalcification in the atherosclerotic plaque may be missed.

All of the morphologic calcific imaging techniques described above provide an assessment of the information on the extent, density, and its spatial distribution of calcification in plaque, but are unable to identify the active process of calcification in plaque.90

Positron Emission Tomography

Feasibility of Evaluating Vascular Calcification Using ¹⁸F‐NaF Positron Emission Tomography

¹⁸F‐sodium fluoride (¹⁸F‐NaF) is a bone tracer that has been used to detect novel areas of bone formation and remodeling since the 1960s.91, 92 The radiotracer binds and incorporates onto the surface of hydroxyapatite crystals by exchanging with hydroxyl ions to form the fluoroapatite.64, 93 Although the most commonly implemented clinical nuclear bone tracer is a technetium‐based radiotracer compound that is imaged with single photon emission computed tomography, ¹⁸F‐NaF has superior characteristics for bone imaging because of the improved resolution of positron emission tomography (PET) in comparison to single photon emission computed tomography.

The biology of the formation and progression of vascular calcification is similar to that in bone formation. Hydroxyapatite crystals in the vessel wall share histological findings with ectopically formed bone (eg, the presence of osteoclast‐like and osteoblast‐like cells).94 Early pathologic studies also observed that arterial calcification resembles a focus of skeleton‐like tissue.95 Additionally, Jeziorska et al observed that the arterial calcification microenvironment has a histochemical resemblance to areas of osteogenesis and bone remodeling (eg, the foci of osteoid matrix, osteocytes, and thin bone trabeculae).96 Modern histopathologic research further confirmed this hypothesis and demonstrated that cells in areas of vascular calcification are derived from osteoclastic and osteoblastic cells.97, 98

Recently, substantial attention has been given to using ¹⁸F‐NaF to investigate vascular calcification in the coronary arteries,22, 64 the aorta,99 and the carotid arteries,100 and it has rapidly gained acceptance as a useful tool for the investigation of plaque pathobiology.

Irkle et al17 studied the mechanism of ¹⁸F‐NaF vascular uptake by performing comprehensive examinations of carotid artery plaques in vivo and in vitro. In this study, pharmacodynamic and pharmacokinetic analysis confirmed the beneficial properties of ¹⁸F‐NaF for plaque assessment (eg, high affinity for calcification, stability before scanning, low plasma activity at the time of scanning with minimal myocardial uptake) and showed that the tracer is more specific to microcalcification than macrocalcification.

Several other studies have reported similar findings and have led to growing support for the use of ¹⁸F‐NaF for imaging the process of microcalcification and visualizing ongoing mineral deposition within the atherosclerotic plaque.64, 99, 100, 101, 102, 103

Divergence Between ¹⁸F‐NaF Uptake and Calcification Visible on CT

A key realization came from the observation that ¹⁸F‐NaF PET imaging and CT provided divergent information about atherosclerotic plaques. Dweck et al64 observed that in spite of a strong correlation between ¹⁸F‐NaF uptake and CACS (r=0.652, P<0.001), there was no corresponding ¹⁸F‐NaF uptake in many of the densely calcified regions on CT. In fact, 41% of patients with CACS >1000 had no significant ¹⁸F‐NaF uptake, and often ¹⁸F‐NaF uptake was found in areas adjacent to and remote from existing coronary calcification. In a separate study, visible arterial calcium was not associated with higher ¹⁸F‐NaF uptake, and ¹⁸F‐NaF uptake was similar in the arterial segments with or without visible calcification.100 Further investigation into the relationship between ¹⁸F‐NaF uptake, visible calcification on CT, and the Framingham risk score for CVD showed that ¹⁸F‐NaF uptake was significantly associated with many cardiovascular risk factors, while visible calcification correlated with age but not other determinants of cardiovascular risk.104

Furthermore, an inverse correlation between the plaque calcium density and ¹⁸F‐NaF uptake was observed by Fiz et al.105 In this study, they classified plaque calcium density into tertiles: light (130–210 HU), medium (211–510 HU), and heavy (>510 HU). The accumulation of ¹⁸F‐NaF ebbed in heavier calcific concretions, to a degree that the uptake of ¹⁸F‐NaF in heavy calcified plaque did not differ from that in control segments. Conversely, the accumulation of ¹⁸F‐NaF was observed in the arterial segments without calcified plaque in most patients.

The mismatch pattern between CT calcium imaging and ¹⁸F‐NaF PET imaging occurs as a result of the dynamic and complex procession of vascular calcification.100, 101 In the earliest stage of atherosclerosis, inflammation induces the release of several bone‐forming peptides and further promotes the active deposition of hydroxyapatite matrix. The exposed hydroxyapatite crystal surface area in the arterial wall and early active vascular calcification is below the resolution of CT, but the evolving powdery microcalcification is readily identified by ¹⁸F‐NaF (Figure 3).22, 98, 106 Dweck et al106 investigated histological markers of active calcification (ie, tissue nonspecific alkaline phosphatase and osteocalcin) in aortic stenosis and found a strong correlation between ¹⁸F‐NaF uptake and these markers, thus confirming that ¹⁸F‐NaF PET imaging provides information about the activity of calcification and might differentiate biologically active calcification from stable calcification. Alizarin Red staining also confirmed ¹⁸F‐NaF uptake in areas of microcalcification17 (Figure 4). As calcium density gradually increases and becomes quiescent without further precipitation of extracellular calcium,107 it becomes visible on CT imaging,108 and its hydroxyapatite core becomes hidden from ¹⁸F‐NaF.106 These 2 imaging techniques might represent 2 different but complementary markers of atherosclerotic calcification.105

Figure 3.

¹⁸F‐NaF uptake is increased in coronary artery culprit lesions. Joshi et al evaluated ¹⁸F‐NaF uptake in the coronary arteries, using positron emission tomography/computed tomography (PET/CT) imaging in the coronary arteries in individuals who had a recent myocardial infarction. A red arrow marks the site of severe stenosis, which is the culprit lesion (left anterior descending artery), while a white arrow marks the site of severe stenosis, which is a bystander non‐culprit lesion (circumflex artery) on invasive coronary angiography for 2 patients (A and C). ¹⁸F‐NaF PET imaging in those same individuals (B and D) shows intense ¹⁸F‐NaF activity at the site of the culprit left anterior descending artery lesions, but at the site of non‐culprit lesions, the CT image demonstrates obvious calcification, whereas there is no ¹⁸F‐NaF activity at this site. Group mean data (E) demonstrate that ¹⁸F‐fluoride activity in the culprit lesions is higher than that in non‐culprit vessels. Reprinted from Joshi et al22 with permission. Copyright ©2014, Elsevier Inc.

Figure 4.

¹⁸F‐Fluoride preferentially binds microcalcification beyond the resolution of computed tomography (CT). Images are taken ex vivo of a carotid endarterectomy specimen excised from a patient who had a recent stroke. A, Histological section of the excised plaque stained for calcium with Alizarin red. B, Filled black arrow shows an area of dense macroscopic calcification that is visible on micro‐CT. By comparison, the empty black arrowhead demonstrates areas of microcalcification that are beyond the resolution of the micro‐CT but by comparison demonstrate avid binding with ¹⁸F‐NaF on both autoradiography (C) and micro–positron emission tomography imaging (D). E, A second carotid endarterectomy sample from a patient post stroke demonstrates a large macrocalcific deposit on micro‐CT. F, Autoradiography shows that although ¹⁸F‐NaF is able to bind to the surface of the plaque, it is unable to penetrate into the center. G, The amount of fluoride adsorbed to microcalcifications is significantly higher than macrocalcifications (F/Ca ratio in microcalcifications 0.59±0.23 (n=10, individual plaques) ; in macrocalcifications 0.37±0.15 (n=7, individual plaques)); *P<0.02 using an ANOVA and Tukey Kramer post hoc test. As a consequence of this effect, ¹⁸F‐NaF binds preferentially to regions of microcalcification compared with macroscopic deposits. Reprinted from Irkle et al17 with permission. Copyright ©2015, Nature Publishing Group, a division of Macmillan Publishers Limited.

Evaluation of Calcification by ¹⁸F‐NaF PET as a Predictor of Plaque Rupture and Cardiovascular Risk

Ruptured atherosclerotic plaque may lead to adverse CVD events such as acute myocardial infarction or stroke. Microcalcification increases the risk of plaque rupture by increasing mechanical wall stress and predisposing the plaque to microfractures and subsequent thrombosis.4, 13, 67

Growing evidence supports the utility of ¹⁸F‐NaF PET imaging for the evaluation of the risk associated with atherosclerotic plaques. A prospective clinical study22 provides a robust and cogent argument that ¹⁸F‐NaF uptake detects coronary artery microcalcification and identifies vulnerable and high‐risk plaques in patients with acute myocardial infarction and stable angina. In patients with myocardial infarction, the ¹⁸F‐NaF uptake in 93% of culprit plaques was significantly higher than that in nonculprit plaques. Furthermore, pathological evidence from plaques excised during carotid endarterectomy verified the fact that marked ¹⁸F‐NaF uptake occurred at the site of all carotid plaque ruptures and was associated with histological evidence of active calcification, macrophage infiltration, apoptosis, and necrosis. In patients with stable angina, 45% of patients had plaques with focal ¹⁸F‐NaF uptake that were associated with more high‐risk features on intravascular ultrasound than those without uptake, including positive remodeling, microcalcification, and necrotic core. Likewise, Vesey et al109 performed ¹⁸F‐NaF PET/CT in patients after recent transient ischemic attack or minor ischemic stroke and found that carotid ¹⁸F‐NaF uptake was increased in culprit plaques in comparison with asymptomatic contralateral plaques and correlated with CVD risk.

¹⁸F‐NaF PET imaging has also been shown to provide a useful method of predicting changes in plaque calcification. Dweck et al106 observed that baseline ¹⁸F‐NaF uptake in stenotic aortic valves provides a marker of active calcification that correlated closely with change in calcium score (r=0.66, P<0.01) and predicted the change in valvular CT calcium scores at 1 year. Jenkins et al110 also investigated the value of valvular ¹⁸F‐NaF uptake in predicting disease progression and clinical outcome in patients with aortic stenosis. The results demonstrated that baseline ¹⁸F‐fluoride uptake correlated strongly with the subsequent rate of progression in aortic valve calcium score and independently predicts cardiovascular death and aortic valve replacement after age and sex adjustments (hazard ratio: 1.55; 95% CI, 1.33–1.81; P<0.001). Furthermore, Beheshti et al98 found that ¹⁸F‐NaF PET/CT allowed the detection of early molecular and cellular calcification in atherosclerotic plaque and may provide highly relevant information about the state of calcified plaque before macroscopic calcification can be detectable by standard CT techniques. These findings suggest that this modality may allow earlier risk stratification and therapeutic intervention in patients with asymptomatic atherosclerosis.

Additional research has shown that patients with increased ¹⁸F‐NaF uptake were more likely to have a clinical diagnosis of coronary artery disease, anginal symptoms, and a higher burden of systemic atherosclerosis. Ten‐year Framingham risk scores for CVD and death are related to ¹⁸F‐NaF uptake.64 Other retrospective studies have shown positive correlations between arterial ¹⁸F‐NaF uptake and CVD risk factors, including age, sex, hypertension, hypercholesterolemia, and diabetes mellitus, further supporting the notion that ¹⁸F‐NaF PET imaging may improve CVD risk stratification.98, 100, 101, 102, 104, 111, 112, 113, 114

In summary, ¹⁸F‐NaF PET imaging has been shown to be useful for the identification of early active calcification and microcalcification.105 This factor makes it potentially useful for the identification of vulnerable plaque115 and individuals with high CVD risk.22, 64, 116

Limitations of ¹⁸F‐NaF PET Imaging

¹⁸F‐NaF PET imaging has lower spatial resolution than CT imaging and continuous cardiac motion, which may prevent accurate quantification of coronary artery microcalcification. However, Dweck et al64 have reported that the spatial resolution of ¹⁸F‐NaF PET was sufficient to allow location of culprit lesions in coronary territories, and continuous cardiac movement can be addressed by using gating. Furthermore, even if the small size of the coronary arteries may fall below the resolution of most PET scanners117 and leads to underestimation of the PET signal because of a partial volume effect, Beheshti et al102 proposed that ¹⁸F‐NaF deposition could be measured throughout the entire myocardium, including a semiquantitative analysis of the contribution from both the major vessels and the microvasculature, in order to limit the impact of smear artifacts caused by the cardiac motion and small vessel diameter. They also emphasized excluding the aortic valve from the region of interest to avoid contamination by the aortic wall or calcified aortic valve leaflets.

Alternative PET Imaging Tracers

Published studies10, 118 have evaluated the association between arterial calcification and inflammation of vascular disease and demonstrated that in early‐stage atherosclerosis, inflammation precedes calcification, and macrophages promote the proinflammatory milieu and send specific signals to vascular wall cells to initiate osteogenic differentiation. Then, both processes developed in parallel and within close proximity. Microcrystals of calcium phosphate may elicit proinflammatory responses from macrophages and build a positive‐feedback amplification loop of calcification and inflammation that drives disease progression. In addition, macrophages and smooth muscle cells may undergo apoptotic changes, providing new foci for calcium deposition. Thus, the evaluation of inflammation is also very important to better understand the process of calcification and atherosclerotic plaques.

¹⁸F‐2‐Deoxy‐2‐fluoro‐d‐glucose (¹⁸F‐FDG) is the most widely validated PET tracer for the evaluation of inflammation within atherosclerotic plaques on the basis of the glucose metabolism of macrophages.21 Li et al119 used ¹⁸F‐NaF and ¹⁸F‐FDG PET/CT to evaluate the association between osteogenesis and inflammation during the progression of calcified plaque. They observed that early atherosclerotic osteogenesis is determined by macrophage accumulation and further intensive progression of arterial calcification might be associated with reduced inflammation. Furthermore, ¹⁸F‐FDG PET/CT can provide additional diagnostic and prognostic information beyond anatomic imaging alone.120 It is becoming increasingly evident that ¹⁸F‐FDG PET/CT is capable of quantifying atherosclerotic inflammation and predicting subsequent cardiovascular events121 and providing a means of evaluating changes with treatment.122 Nevertheless, ¹⁸F‐FDG lacks cell specificity, and in contrast to ¹⁸F‐NaF, is extremely limited for the assessment of coronary arteries because of myocardial spillover in spite of significant efforts at background myocardial suppression.22

In atherosclerotic inflammation, the surface of activated macrophages overexpresses the G‐protein‐coupled receptor somatostatin receptor subtype‐2.123 Preclinical124 and retrospective125, 126 studies suggest that gallium‐68‐labeled [1,4,7,10‐tetraazacy‐clododecane‐N,N′,N″,N‴‐tetraacetic acid]‐d‐Phe1, Tyr3‐octreotate (DOTATATE), a somatostatin receptor subtype‐2‐binding PET tracer, may be superior to ¹⁸F‐FDG for imaging atherosclerotic inflammation. Tarkin et al127 validated ⁶⁸Ga‐DOTATATE PET as a novel marker of atherosclerotic inflammation and showed that ⁶⁸Ga‐DOTATATE PET offers superior coronary imaging, excellent macrophage specificity, and better power to discriminate high‐risk versus low‐risk coronary lesions than ¹⁸F‐FDG. These findings require confirmation in larger studies, and the utility of ⁶⁸Ga‐DOTATATE PET in the prediction of clinical outcomes should be tested.

Notably, neither ¹⁸F‐FDG nor ⁶⁸Ga‐DOTATATE PET is able to evaluate vascular microcalcification.

Intravascular Imaging

Invasive angiography offers superior spatial resolution (0.1–0.2 mm) and temporal resolution (10 ms) and has been the criterion standard for coronary atherosclerosis imaging for more than 50 years.128 It generates a 2‐dimensional silhouette of the arterial lumen and does not image the vessel wall; thus, it does not provide direct imaging of calcification or the atherosclerotic plaque itself.

With ultrasound techniques (eg, vascular, transesophageal, and intravascular ultrasound [IVUS]), it is possible to directly image the vessel wall and analyze atherosclerotic plaques in various arterial beds including the carotids, aorta, and coronary arteries. It also allows monitoring of the natural history of plaque formation and identification of the compositional features. Usually, atherosclerotic plaque may appear in 3 distinct textural patterns by parameters of acoustic attenuation129: hypoechoic (lipid‐rich or hemorrhagic plaque), moderately hyperechoic (fibrotic or fibrofatty plaque), and markedly hyperechoic with shadowing (calcific plaque). Hyperechoic calcified plaque is easily identified by its strongly reflective ultrasound properties with an attenuation 700% greater than that in normal wall and 300% greater than that in atherosclerotic noncalcific plaques. Alternatively, hypoechoic plaque shows poor or no reflection.130, 131

Grayscale IVUS uses miniaturized crystals and generates high‐resolution images of the vessel wall and lumen to allow quantitative measurements to the plaque area, but its limited axial resolution (150 μm) and lateral resolution (300 μm) render it unable to differentiate individual plaque components.132

Virtual histology IVUS, an advanced radiofrequency analysis of reflected ultrasound signals, allows reconstruction of a color‐coded tissue map, and provides an improved assessment of plaque composition including necrotic, fibrous, densely calcified, and fibrofatty regions with reasonable accuracy.133 Nevertheless, it remains limited by noise, artifacts, and insufficient spatial resolution.134

Integrated backscatter intravascular ultrasound (IB‐IVUS) is another promising technique. In this technique, IB values for each tissue component are calculated as the average power using a fast Fourier transform, an algorithm that samples a signal over a period of time or space and divides it into its frequency components. These components are single sinusoidal oscillations at distinct frequencies with their own amplitudes and phases that are measured as decibels of the frequency component of the backscattered signal from a small volume of tissue. In IB‐IVUS images, color‐coded maps have 4 major components based on different radiofrequency signals: fibrosis (green), dense fibrosis (yellow), lipid (blue), and calcification (red). Kawasaki et al135 demonstrated that IB‐IVUS depicts calcified lesions with high sensitivity. However, a limitation resides in the fact that very dense fibrous lesions may produce sufficient reflectivity and attenuation or acoustic shadowing to be misclassified as calcified.136 Furthermore, Okubo et al137 showed that IB‐IVUS occasionally underestimated calcified lesions and overestimated lipid pools behind calcifications as a consequence of the acoustic shadowing resulting from the calcification. Moreover, the ability to detect ultrasonically hypoechoic microcalcification with IB‐IVUS is limited.

Other intravascular coronary imaging modalities such as optical coherence tomography have been increasingly evaluated by recent studies. Optical coherence tomography has high spatial resolution (10–15 μm), allows the evaluation of variations in plaque composition at a cellular level, and distinguishes between plaque types.138 However, correct differentiation between calcium and lipid can be challenging with optical coherence tomography, and its limited tissue penetration (1–3 mm) makes the assessment of the entire plaque volume impossible. Optical frequency domain imaging, a second generation of optical coherence tomography, has shown an ability to discriminate plaque composition that relates to findings at autopsy and quantitatively evaluates the thickness and border of vascular calcification.139

Importantly, all of the invasive angiography techniques are limited in their ability to identify microcalcification.

Conclusions

Significant advances have been made in our understanding of the mechanism of atherosclerotic calcification and the value of multimodality imaging techniques for evaluating this pathological process. ¹⁸F‐NaF PET imaging provides unique information about the pathologic process of vascular calcification and identifies microcalcification, which makes it a promising tool for performing noninvasive, reproducible measurements of active plaque calcification that correlate with vulnerability and culprit plaques. The combination of ¹⁸F‐NaF PET imaging with other structural imaging modalities (eg, CT) may provide a noninvasive, qualitative, and quantitative assessment of calcification and plaque vulnerability in atherosclerotic disease that offers important diagnostic and prognostic insights for the assessment of CVD risk and the monitoring of clinical treatment.

Disclosures

None.

J Am Heart Assoc. 2018;7:e008564 DOI: 10.1161/JAHA.118.008564.