J A C C : C A R D I O V A S C U L A R I M A G I N G V O L . 9 , N O . 3 , 2 0 1 6

ª 2 0 1 6 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 1 9 3 6 - 8 7 8 X / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j c m g . 2 0 1 5 . 1 2 . 0 0 3

MR Imaging of Coronary Arteriesand Plaques

Marc R. Dweck, MD, PHD,a,b Valentina O. Puntmann, MD, PHD,c Alex T. Vesey, MD,b Zahi A. Fayad, PHD,a

Eike Nagel, MD, PHDd

ABSTRACT

Fro

Fo

Ce

DZ

res

fro

rel

Ma

Cardiac magnetic resonance offers the promise of radiation-free imaging of the coronary arteries, providing information

with respect to luminal stenosis, plaque burden, high-risk plaque characteristics, and disease activity. In combination, this

would provide a comprehensive, individualized assessment of coronary atherosclerosis that could be used to improve

patient risk stratification and to guide treatment. However, the technical challenges involved with delivering upon this

promise are considerable, requiring sophisticated approaches to both data acquisition and post-processing. In this review,

we describe the current status of this technology, its capabilities, its limitations, and what will be required in the future to

translate this technology into routine clinical practice. (J Am Coll Cardiol Img 2016;9:306–16)

© 2016 by the American College of Cardiology Foundation.

M yocardial infarction (MI) is a leading causeof death and a major health resourceburden that by 2030 is estimated to cost

the global economy more than U.S. $1 trillion peryear (1). The majority of these events occur as a conse-quence of atherosclerotic plaque rupture. Identifyingplaques at risk of rupture is challenging, however—inparticular because these lesions are frequently nonob-structive on antecedent angiography and thereforealso missed with conventional ischemia imaging.Interest has therefore developed in novel strategiesfor improving the prediction of cardiovascular risk inpatients. Measures of plaque burden, such ascomputed tomography (CT) calcium scoring, offersome improvements, particularly in low-risk popula-tions, whereas advances in scanner technology nowallow measurement of disease activity and the visual-ization of high-risk plaque characteristics. A singleimaging modality that can assess each of these param-eters of coronary atherosclerosis would potentially bea major advance. Specific imaging protocols couldthen be tailored to different patient populations and

m the aTranslation Molecular Imaging Institute, Icahn School of Medicine

undation Centre for Cardiovascular Science, University of Edinburgh, Ed

nter Goethe University Frankfurt, Germany; and the dInstitute for Expe

HK Centre for Cardiovascular Imaging, University Hospital Frankfurt, Fr

earch support from Siemens Healthcare, Bayer Healthcare, Philips Hea

m CVI42 and MEDIS; and speaker honoraria from Bayer Healthcare. A

ationships relevant to the contents of this paper to disclose.

nuscript received October 8, 2015; revised manuscript received Novembe

combined protocols designed to provide complemen-tary information and to maximize the technique’sprognostic capability. Ideally such a scan would besafe, widely available, and repeatable, allowing us totrack coronary atherosclerosis during a patient’s life-time. This review investigates the promise that car-diac magnetic resonance (CMR) holds in imagingcoronary atherosclerosis, with a particular focus onmagnetic resonance angiography (MRA), assessmentsof plaque burden and plaque characteristics alongsideevolving molecular technologies that target diseaseactivity (Central Illustration). We examine some ofthe technical challenges facing CMR, its current capa-bilities, and how in the future it may become theimaging modality of choice for investigating coronaryatherosclerosis.

RATIONALE FOR CMR IMAGING OF THE

CORONARY ARTERIES

Magnetic resonance has become a routine clinicalimaging investigation used to assess a wide range of

at Mount-Sinai, New York, New York; bBritish Heart

inburgh, United Kingdom; cCardiovascular Imaging

rimental and Translational Cardiovascular Imaging,

ankfurt am Main, Germany. Dr. Nagel has received

lthcare, TomTec, and MEDIS; educational support

ll other authors have reported that they have no

r 25, 2015, accepted December 3, 2015.

AB BR E V I A T I O N S

AND ACRONYM S

CAD = coronary artery disease

CMR = cardiac magnetic

resonance

CT = computed tomography

18F-FDG = 18F-

flurodeoxyglucose

MI = myocardial infarction

MRA = magnetic resonance

angiography

MRI = magnetic resonance

imaging

PET = positron emission

tomography

SSFP = steady-state free

precession

STIR = short tau inversion

recovery

3D = 3-dimensional

2D = 2-dimensional

USPIO = ultrasmall

superparamagnetic particles of

iron oxide

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6 Dweck et al.M A R C H 2 0 1 6 : 3 0 6 – 1 6 Coronary Imaging With CMR

307

tissues and pathophysiological conditions. Althoughits application to the heart was delayed comparedwith stationary structures, technological advances insoftware and hardware components have now lead toits widespread clinical adoption for imaging themyocardium. Applying CMR to the coronary arteriesoffers further challenges because of both their smallcaliber and complex motion. As a consequence, cor-onary CMR has to date lagged behind CT. However,CMR holds several key advantages over CT that mightultimately see it become the imaging modality ofchoice for assessing the coronary vasculature. First,CMR offers superior soft-tissue contrast, potentiallyallowing improved detection of high-risk plaquecharacteristics such as intra-plaque hemorrhage,thrombus, and positive remodeling (Figure 1). Second,CMR is not affected by the calcium blooming thathampers CT and frequently renders luminal assess-ments impossible in patients with advancedatheroma. Finally and, perhaps most importantly,CMR does not involve exposure to ionizing radiation.This makes it attractive as a screening modality, al-lows multiple vascular beds to be imaged simulta-neously, and renders serial imaging to track diseaseprogression feasible. Moreover, there is the opportu-nity to image younger patients so that we can betterunderstand the early stages of atherosclerosis andhow subjects transition from health to disease.Indeed, it is worth noting that these various advan-tages have seen CMR, not CT, emerge as the imagingmodality of choice for assessing the carotid and pe-ripheral arteries. The challenge therefore is to suc-cessfully apply these various techniques to thecoronary arteries.

TECHNICAL CHALLENGES AND SOLUTIONS. Thecoronary arteries are small tortuous structures,have variable 3-dimensional (3D) branches and cover alarge volume. A clinically useful assessment requiresvisualization of the proximal and mid-vessels, whichoften only measure 3 to 4 mm in diameter (smallervessels and side branches are less important in guidingpatient management) and is made even more difficultby the rapid movement of these arteries with thecardiac cycle and breathing. The requirements forboth high spatial resolution and complex motioncorrection are highly challenging for any imagingmodality, but in CMR create a particular tension withrespect to scanning times and data acquisition.Althoughmost strategies to improve spatial resolutiongenerate exponential increases in the scanning timerequired, current motion correction approachesdiscard large amounts of data while the heart is mov-ing. The future challenge is therefore to find fast CMR

techniques capable of high spatial resolutionwithout compromising motion correction andvice versa.

To suppress cardiac motion, data arecurrently only acquired during end-systolicor mid-diastolic standstill. These periodsvary for different coronary arteries and be-tween patients but are around 70 ms for mostpatients and can be lengthened by reducingheart rate with beta-blockade (2). To suppressrespiratory motion, images can be obtainedduring breath-holding; however, this limitsscan time to <20 s and therefore the data thatcan be acquired. As a consequence, the 3Dcoverage of these scans is limited to a thinslab with a spatial resolution of w0.7 � 0.7 �4 to 5 mm. Alternatively, more detailed im-ages can be obtained during free breathing,using navigator sequences, which measurethe position of the diaphragm and only allowdata acquisition when the diaphragm is closeto a predefined position. Such scans takemuch longer, between 10 and 15 min. Forsmall deviations of the diaphragmatic posi-

tion, corrections can be applied; however, even thenthis approach remains highly wasteful given theamount of data that is discarded. Recent advances inself-navigated techniques measure displacement ofthe heart directly and then attempt to correct the datafor motion rather than discarding it. This has thepotential to greatly improve scanner efficiency and toincrease spatial resolution (3). This coupled withparallel imaging, higher field strengths, dedicatedmultichannel receiver coils, fully digital systems, andcompressed sensing techniques promise rapidimprovements in the image quality associated withcoronary magnetic resonance.

MAGNETIC RESONANCE ANGIOGRAPHY. The stan-dard assessment of the coronary vasculature remainsthe coronary angiogram, which focuses on delineatingthe arterial anatomy and detecting focal narrowings inits lumen. Various sequences are available for per-forming coronary MRA (Figure 1). In the cerebral andperipheral arteries, MRA is now clinically established.Translating these successes into reliable coronaryMRA has been a focus of research for 25 years since thefirst reports by Edelman (4) and Manning (5). Thissection describes current state-of-the-art techniquesand some of the more clinically focused researchvalidating coronary MRA against CT and invasivecoronary angiography.

For coronary MRA, rapid sequences are essentialand most centers use either spoiled gradient echo

CENTRAL ILLUSTRATION Multiparametric Cardiac Magnetic Resonance Imaging of the Coronary Arteries

Cardiac magnetic resonance potentially allows multifaceted imaging of the coronary arteries, providing information with respect to luminal

stenosis on angiographic images, plaque burden on black-blood imaging, high-risk plaque characteristics including high-intensity plaques, and

disease activity with hybrid positron emission tomography/magnetic resonance imaging.

Dweck et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6

Coronary Imaging With CMR M A R C H 2 0 1 6 : 3 0 6 – 1 6

308

(e.g., TurboFLASH) or balanced steady-state freeprecession (SSFP) (e.g., TrueFISP) sequences. Theformer requires contrast to generate bright bloodangiograms, the latter does not. At higher fieldstrengths, however, SSFP sequences are suboptimalbecause of longer repetition times (mandated byspecific absorption rate limitations), susceptibility tooff-resonance effects, and the need for higher flipangles. At 3-T or 7-T, gradient echo imaging withcontrast is therefore frequently preferred. Additionalpreparatory techniques such as frequency selectivefat suppression, magnetization transfer contrast, or

T-2 pre-pulses may also be used to enhance contrastbetween the vessel and the surrounding fat,myocardium, or venous blood.

Chelated gadolinium-based contrast agents aremost commonly used in contrast-enhanced MRA,shortening the T1 relaxation times of the blood pool.This enhances the coronary lumen compared withsurrounding tissue, facilitating improved resolutionand discrimination. Gadolinium contrast agents arerenally excreted and redistribute into surroundingtissue rapidly so that a time frame of only 10 to 15 minis available for coronary imaging. Nevertheless, this

FIGURE 1 Coronary Magnetic Resonance Angiography and Assessments of

Plaque Burden

Contrast-enhanced magnetic resonance angiograms of the left anterior descending artery

(A and B) and of the right coronary artery (C). Black-blood image of the left anterior

descending artery demonstrating marked increases in plaque wall thickness (D). Panel D is

adapted from Fayad et al. (23).

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6 Dweck et al.M A R C H 2 0 1 6 : 3 0 6 – 1 6 Coronary Imaging With CMR

309

appears sufficient to generate anatomical images ofthe coronary arteries and to delineate luminal ste-nosis, with an accuracy approaching that of CT andinvasive coronary angiography (Figure 2). The main 2disadvantages of current contrast agents are the rapidleakage from the visual lumen into the surroundingtissue, reducing signal-to-noise and the concomitantsignal increase in the coronary venous system thatcan cause overlap with the circumflex artery inparticular.

Alternative contrast agents are being investigated.These demonstrate longer half-lives in the circulation,thereby providing more time for navigator-basedsequences to acquire data. These include “intra-vascular” albumin-binding gadolinium compounds(e.g., gadofosveset) that are already approved for hu-man use and appear to improve steady-state imagingof the vasculature, while also providing assessmentsof endothelial permeability and angiogenesis (6,7).Similarly, there is interest in performing coronaryMRA using ultrasmall superparamagnetic particles ofiron oxide (USPIO), although concerns have beenraised about hypersensitivity reactions.Current performance in the detection of coronary arterydisease. Multiple studies have investigated the accu-racy with which coronary MRA can detect and quan-tify coronary luminal stenoses. The first major reportby Kim et al. (8) compared coronary MRA to invasivecoronary angiography in 109 patients in a multicentersetting, demonstrating an overall diagnostic accuracyfor coronary artery disease (CAD) of 72% with asensitivity and specificity of 93% and 42%, respec-tively. There was, however, variation depending onthe coronary artery being examined, with thetortuous left anterior descending and awkwardly sit-uated circumflex arteries proving most troublesome.A more recent multicenter study demonstratedslightly better results with an overall accuracy of 79%,a sensitivity of 88%, a specificity of 72%, and an areaunder the cure of 0.87, again using invasive angiog-raphy as the gold standard (9). Many similarcomparative reports have been published (10–13),culminating in a meta-analysis in 2010 demonstratingthat, although MRA had an overall sensitivity of 87%and a specificity of 70% for obstructive coronarydisease, CT performed better with values of 97% and87%, respectively (14). Several advances and de-velopments have been made since this meta-analysiswith the latest studies reporting improved MRAresults. For example Yonezawa et al. (15) used anunenhanced 3D whole-heart, free breathing, SSFPtechnique at 1.5-T and reported a sensitivity of 91%and specificity of 86%, with an area under the curveof 0.92. Although these data remain inferior to CT

coronary angiography (itself a rapidly advancingtechnology), it suggests that the gap may be narrow-ing and that ultimately it might be closed with futuretechnological advances.

Current recommendations. The most recent Appropri-ateness Criteria for coronary artery imaging re-commended coronary MRA for the assessment ofanomalies in the coronary arteries (class I) and aorto-coronary bypass grafts (class II) (16). MRA can robustlyvisualize anomalies of the coronary ostia and coronaryaneurysms, helped by the large caliber and minimalmotion of these particular abnormalities (16). Anothermajor advantage is the avoidance of ionizing radiationin the young patients affected by these conditions,often making CMR the preferred imaging modality.Although the presence and patency of bypass graftscan be determined with coronary MRA (16), artefactfrom clips/sternal wires and difficulty in assessing thedistal anastomosis mean that CT imaging is usuallypreferred.

Of note the use of coronary MRA for the assess-ment of native coronary arteries is currently notregarded as appropriate. However, niche areasare emerging for instance its use in combinationwith late gadolinium enhancement as a strategyfor differentiating ischemic and dilated cardiomy-opathies (16).

FIGURE 2 Characteristics of High-Risk Coronary Plaques and CMR Techniques for Their Detection

The image shows key characteristics of high-risk coronary plaques and the available cardiac magnetic resonance (CMR) techniques that can be

used to identify these features.

Dweck et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6

Coronary Imaging With CMR M A R C H 2 0 1 6 : 3 0 6 – 1 6

310

PLAQUE BURDEN. In addition to luminal assess-ments, CMR can provide detailed imaging of thearterial vessel wall allowing us to measure plaquethickness in 2-dimensions (2D) and plaque burden in3D (Figure 1). Classically, this technique involvesblack-blood imaging sequences, using double inver-sion recovery pre-pulses to null signal in the arteriallumen followed by fast spin echo sequences toenhance signal within the plaque. More recently,flow-independent black-blood approaches have beensuggested, which may be of particular value in thecoronary arteries (17). Both approaches aim to maxi-mize contrast between the vessel wall and the lumen,and addition of fat suppression techniques can thenimprove resolution between the outer vessel and thesurrounding epicardium (18).

Using these approaches atherosclerotic plaquethickness and 3D plaque burden can be measured inthe carotids (19), providing measures of the athero-sclerotic burden that can assess treatment responsesto novel therapies (20,21) and predict major adversecardiovascular outcomes (22). The feasibility of thisapproach to image coronary plaque was first demon-strated in 2000 with black-blood, breath-held, single-slice imaging (23). High resolution, cross-sectional

images of the coronary arteries were obtained,allowing measurement of the coronary wall thickness.These plaque thickness measurements were increasedin patients with established ischemic heart diseasecompared with healthy control subjects (23) and havesubsequently been shown to correlate with increasingnumbers of cardiovascular risk factors (24). To date,the standard approach has been to perform 2D mea-surements of plaque wall thickness on cross-sectionalimages (often in a single artery) rather than a moreglobal 3D assessment of plaque burden. These wallthickness measurements have a spatial resolution inthe region of 0.7 � 0.6 � 3 mm and scan-rescanreproducibility of w0.02 � 0.20 mm (25,26). Theyhave been validated against intravascular ultrasoundwith mixed results (27,28).

Although plaque thickness may provide a surrogateof plaque burden, they will always be prone to sam-pling error. The ongoing challenge will therefore be todevelop rapid methods for imaging the coronaryvessel wall in 3D, providing comprehensive plaqueburden estimates that can compete with thesimplicity of CT calcium scoring. Encouragingly,studies have suggested that 3D black-blood imagingof individual coronary arteries is feasible in patients

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6 Dweck et al.M A R C H 2 0 1 6 : 3 0 6 – 1 6 Coronary Imaging With CMR

311

with nonobstructive disease (29) and, further, thatwhole heart black-blood imaging of the coronaryvessels in healthy subjects is possible with a novelflow-independent i-T2prep sequence (17). Simulta-neous advances in image analysis will also berequired to accurately segment the arterial wall,allowing for rapid plaque burden quantification.

HIGH-RISK ANATOMICAL PLAQUE CHARACTERISTICS.

Histological studies have indicated that culprit pla-ques responsible for MI have certain pathologicalcharacteristics, including a large necrotic core, posi-tive remodeling, a thin fibrous cap, angiogenesis,intraplaque hemorrhage, inflammation, and micro-calcification (Figure 2). Each of these represents apotential imaging target that might be used to refineour ability to identify high-risk coronary lesions. Tobetter predict an individual person’s risk, it will beimportant to measure not just the plaque burden butalso the high-risk plaque burden. Indeed, emergingdata have suggested that identification of such char-acteristics using CT can improve current patient riskstratification models (30). In the next section, we willreview the ability of CMR to identify the specificcharacteristics that define high-risk plaques.

POSITIVE REMODELING. Individual atheroscleroticlesions with eccentric plaque formation and positiveremodeling can be identified using the black-bloodCMR measurements of lumen diameter and plaquethickness previously discussed (Figure 1). Thesemeasures of increased plaque eccentricity in the aortaand carotids have been shown to predict adversecardiovascular outcomes (22). Positive remodeling isalso detectable in the coronary arteries. In an analysisof 179 patients participating in the MESA (Multi-Ethnic Study of Atherosclerosis) study, 365 cross-sectional black-blood images were sampled. A totalof 129 of these images were not interpretable, but, inthe remainder, 2D measurements were made of thelumen, the plaque wall, and the epicardial diameter.As coronary plaque wall thickness increased, theoverall vessel size increased at a greater rate than thechange in lumen area, indicating that the coronaryvessel size enlarged to compensate for atheroscleroticchange (i.e., positive remodeling) (31). In a similarstudy, positive remodeling was observed in the rightcoronary artery of 223 asymptomatic subjects >60years of age. Again, around one-third of studies werenot interpretable, but in the remaining, wall thick-ness correlated with increased vessel area consistentwith positive remodeling and with the degree ofcoronary calcification (32). CMR studies relating pos-itive remodeling in the coronary arteries to clinicaloutcomes are awaited.

PLAQUE HEMORRHAGE AND LUMINAL THROMBUS. Met-hemoglobin is an intermediate breakdown product ofhemoglobin that is formed 12 to 72 h following hem-orrhage and therefore represents a key component ofacute thrombus. It is associated with a high signal andshort T1, allowing detection of fresh thrombus usingmagnetic resonance imaging (MRI) in a range ofconditions, including deep venous thrombosis andpulmonary embolism. More recently, this approachhas been used in atherosclerosis to visualize regionsof both endothelial thrombus related to plaquerupture/erosion and intraplaque hemorrhage. In thecarotid arteries, noncontrast T1-weighted echogradient sequences, incorporating fat and bloodsuppression, can detect high-intensity signal in theculprit plaques of patients who have suffered a recentstroke (33,34). The same authors went on to providehistological validation of this direct thrombus imag-ing approach in 63 patients who underwent MRIbefore carotid endarterectomy. They demonstratedthat such high-intensity plaques had negative andpositive predictive values of 70% and 93%, respec-tively, for the detection of complex carotid lesionswith evidence of surface rupture and intraluminal orintraplaque hemorrhage (35).

More recently, unenhanced T1-weighted 3D imag-ing sequences (this time incorporating navigatorsequences and electrocardiogram gating) have beenused to identify intracoronary thrombus. A majoradvantage of this technique is that it can be used toacquire information across the entire volume of theheart, ensuring that sampling error is not an issue.However, T1-weighted imaging of the coronary arteriesdoes not provide detailed anatomical information andso has to be fused with coronary MRA scans, similar tothe approach employed for hybrid positron emissiontomography (PET) imaging (Figure 3). This techniqueappears effective in the coronary vasculature, with arecent study suggesting a sensitivity and specificity ofw90% for thrombus at the site of culprit plaque rupturein patients after MI (36). Similarly, in a multimodalityimaging study of patients with angina, high-intensityplaques on these T1 imaging sequences were associ-ated with multiple high-risk imaging featuresincluding subclinical intracoronary thrombus on opti-cal coherence tomography (37), positive remodeling onultrasound, and low attenuation plaque on CT (38).

Unenhanced T1-weighted imaging also appears toprovide important prognostic information. The pres-ence of high-intensity plaques in the carotid arteriespredicted cardiac events, outperforming carotidintimal medial thickness and traditional cardiovas-cular risk factors (39). This finding has since beencorroborated in a recent meta-analysis of 8 studies

FIGURE 3 CMR Assessment of High-Risk Plaque Characteristics and Disease Activity

(A) High-intensity plaque on T1-weighted image of the coronaries. This is believed to act as

a marker of intraplaque hemorrhage or intraluminal thrombus. (B) Fusion of T1-weighted

image with a contrast-enhanced magnetic resonance angiogram localizes this high-

intensity plaque to the left anterior descending artery. (C) Fused PET/CT image showing

uptake of 18-fluoride in the proximal left anterior descending artery as a marker of

microcalcification. (D) Fused PET/MRI image demonstrating increased uptake of 18F-FDG

in the left main stem as a marker of vascular inflammation. Panels A and B are adapted

from Noguchi et al. (42). CMR ¼ cardiac magnetic resonance; CT ¼ computed tomography;

18F-FDG ¼ 18F-flurodeoxyglucose; MRI ¼ magnetic resonance imaging; PET ¼ positron

emission tomography.

Dweck et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6

Coronary Imaging With CMR M A R C H 2 0 1 6 : 3 0 6 – 1 6

312

incorporating 689 patients, which demonstrated a6-fold increase in cerebrovascular events in patientswith high-intensity carotid plaques (40). In thecoronary arteries, a case report first described ahigh-intensity coronary plaque going on to ruptureand cause MI (41). Subsequently, a large studydemonstrated high-intensity plaques in 159 of 568patients with known or suspected CAD. Forty-oneof these subjects subsequently went on to havea coronary event with the presence of a high-intensity plaque acting as an independent predic-tor on multivariate analysis (hazard ratio: 3.96; 95%confidence interval: 1.92 to 8.17) (42). Most recently,high-intensity plaques in patients with stable CADwere predictive of periprocedural MI at the timeof percutaneous intervention (43). These emergingprognostic data, coupled with the ability to imagethe entire coronary vasculature, place noncontrast

T1-weighted imaging as perhaps the most promisingCMR approach for identifying high-risk coronaryatheroma.

INFLAMMATION AND ANGIOGENESIS. Atherosclerosisis an inflammatory condition characterized at almostevery stage by the actions of macrophages, T helpercells, and an array of pro-inflammatory cytokines. Inparticular, inflammation plays a pivotal role in theprecipitation of acute MI, with macrophages secretingmatrix metalloproteinases that weaken the fibrouscap, predisposing the plaque to rupture. Angiogenesisis another key process that is triggered by the hypoxicenvironment within the necrotic core and thereforecommonly observed in high-risk lesions. The newvessels that develop are immature, leaky, and asso-ciated with intraplaque hemorrhage, which in turncan trigger abrupt plaque growth and/or rupture.Inflammation and angiogenesis therefore bothrepresent key high-risk plaque characteristics withseveral novel CMR approaches demonstrating earlypromise in their detection.

Late gadolinium enhancement has become widelyused for detecting regions of extracellular expansionin the myocardium. Similarly, in carotid atheroma,gadolinium has been shown to accumulate in areas ofinterstitial edema, angiogenesis, and fibrosis.Although this technique can be used to image thefibrous cap, late enhancement toward the adventitiarelates to deeper regions of inflammation and angio-genesis. In particular, the rate of contrast enhance-ment on dynamic MRI has been shown to correlatestrongly with histological evidence of plaque angio-genesis and macrophage content (44), whereas in aretrospective study, such late enhancement wasobserved in the culprit lesions of patients post-stroke.Attempts have been made to move this technologyinto the coronary arteries, with late gadoliniumenhancement of these vessels correlating with theseverity of atherosclerosis (45) and with vasculiticinflammatory conditions such as systemic lupuserythematous (46).

Short-tau inversion recovery (STIR) imaging pro-vides T2-weighted images that are sensitive to theincreased water content in regions of inflammation.The feasibility of detecting coronary inflammationwas first demonstrated in a pig model of coronaryballoon injury and validated against histology (47).Subsequent studies have demonstrated an increasedSTIR signal in the culprit coronary artery of apatient sustaining MI (48) and in regions of peri-stent inflammation after percutaneous coronaryintervention (49). Further work is now required totest whether both late gadolinium enhancement and

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6 Dweck et al.M A R C H 2 0 1 6 : 3 0 6 – 1 6 Coronary Imaging With CMR

313

STIR imaging can reliably identify coronary plaqueinflammation and differentiate stable from unstableatherosclerotic lesions.

FUTURE TECHNIQUES. The lipid-rich necrotic coreconsists largely of oxidized lipid, apoptotic cells, celldebris, and associated inflammation. In the carotidarteries, efforts have been made to image the lipidwithin these regions. Both fibrous tissue and lipid-rich necrotic core have similar T2 properties andproton densities, making their differentiation diffi-cult. However, late gadolinium enhancement canimprove their distinction because, although theatherosclerotic fibrous cap enhances with gadolin-ium, the necrotic core lacks its own vasculature andappears hypointense. This approach has thereforebeen used to estimate both the carotid fibrous capthickness and the size of its necrotic core (50).Importantly, this approach can also be used to iden-tify regions of carotid plaque rupture or ulceration(51) that may or may not be clinically apparent (52).Translation of these approaches into the coronaryarteries has to date not proved feasible and presents amajor challenge given the necessary spatial resolu-tion. With further technical advances, identificationof coronary lesions with a large necrotic core or athick fibrous cap may yet still be possible, both ofwhich would be of potential clinical utility. HoweverCMR is unlikely to provide sufficient resolution toaccurately identify thin-capped coronary fibroather-oma, given that these are defined by a cap thicknessof <50 mm.

Microcalcification. Like CT, CMR can resolve regionsof macroscopic calcium that have a hypointensesignal on most CMR sequences and are generallyassociated with plaque stability. By contrast, CMRis unable to resolve the areas of microcalcificationthat are associated with increased plaque vulnera-bility, although hybrid imaging with 18F-fluoridePET holds great potential in this regard. Mechanicaland shear stress appear to have an important rolein driving both the development of atheroscleroticlesions; for example, in areas of low shear stresssuch as bifurcations, and also the propensity toplaque rupture at sites of increased mechanicalstress. Both can be modeled using finite elementanalysis on the basis of anatomical imaging pro-vided by CMR and may therefore be of use inbetter understanding the pathophysiology leadingto clinical events (53). Similarly, assessments ofcoronary flow may provide assessment of hemody-namic obstruction associated with individual coro-nary lesions and measurement of the coronary flowreserve (54,55).

DISEASE ACTIVITY

The ability to study the activity of specific patholog-ical processes is now with us, heralding the dawn of anew era of cardiovascular molecular imaging. CMRsits at the center of that capability with the avail-ability not only of specific magnetic resonancetracers, but also with the emergence of hybrid PETand MRI systems that can harness the unrivaledsensitivity and availability of PET imaging agents.

CMR-BASED TRACERS. Alongside their potential asintravascular contrast agents, the powerful T2* ef-fects of USPIOs have been used to target a wide rangeof molecular processes. In humans, USPIOs havebeen used off-label to study vascular inflammation.After intravenous administration, USPIOs becomeengulfed by activated monocytes resulting in areduced CMR signal in tissue subsequently infiltratedby these cells. This technique has been used tomonitor vascular macrophages in carotid atheroma,demonstrating increased USPIO uptake in culprit ca-rotid plaques post-stroke compared with contralat-eral lesions (56) and reduction in this signal withstatin therapy (57). Translation of these imagingagents in to coronary atheroma is keenly awaited,although it should be noted that the strong blood-pool signal provided by USPIOs may make identifi-cation of signal in the coronary wall challenging.

At the pre-clinical level, USPIOs have been incor-porated into a wide range of imaging nanoparticlesthat can then be targeted to different pathophysio-logical processes. The iron oxide then allows thesenanoparticles to be tracked in vivo using T2-weightedMRI. Similarly, other MRI contrast agents can be used.For example, lipid micelle nanoparticles targetingoxidized lipid have been labeled with gadolinium (58),iron oxide (59), and manganese (60) and used to imagemurine atheroma. We have also used USPIOs to labelmodified high-density lipoprotein particles. Theseparticles evade the immune system and naturallymigrate to lipid-rich murine atheroma with the po-tential of providing important diagnostic and mecha-nistic information (61). Moreover, the center of theseparticles can be loaded with pharmacological agents,transforming them into drug delivery systems, tar-geting the most active and inflamed plaques (62). Ifthis nanotracer technology can be translated in tohumans, it might potentially allow us to both imageand treat high-risk plaques with a level of specificitynot previously possible.

HYBRID PET/CMR IMAGING. The advent of hybridPET/CMR imaging has presented the opportunity toharness the molecular imaging available with PET

Dweck et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6

Coronary Imaging With CMR M A R C H 2 0 1 6 : 3 0 6 – 1 6

314

alongside established CMR techniques. To date, cor-onary PET studies have exclusively been fused withCT, but CMR has multiple advantages, such as motioncorrection and soft-tissue contrast that are likely tosee it increasingly used. To date, 2 tracers in partic-ular have been used to target disease activity andhigh-risk atherosclerotic plaques: 18F-flurodeox-yglucose (18F-FDG) and 18F-fluoride. 18F-FDG is aglucose analog that has been widely used as a markerof vascular inflammation in the carotid arteries on thebasis that macrophages use more glucose than sur-rounding cells. Indeed, 18F-FDG uptake in the ca-rotids correlates closely with macrophage burden,localizes to culprit plaque following stroke, and ismodifiable with statin therapy (63). Studies using thistracer in the coronary arteries have had mixed results,predominantly because glucose is also the predomi-nant energy source of the myocardium, with uptakein the heart muscle often obscuring coronary activity(64,65).

18F-fluoride has recently been shown to preferen-tially bind regions of vascular microcalcification ac-tivity beyond the resolution of CT (66). In contrast tothe macroscopic calcium deposits detected by CT thatimpart stability, microcalcification is consistentlyassociated with high-risk coronary lesions andincreased risk of rupture (67). 18F-fluoride uptake inthe coronary arteries has now been described in 2clinical trials. In the first, increased uptake could belocalized to individual coronary plaques and identi-fied high-risk patients with increased Framinghamrisk scores (68). In the second, increased activitycolocalized to individual coronary plaques in patientswith stable angina, with multiple high-risk featuresincluding a large necrotic core, microcalcification,and positive remodeling. A parallel cohort of patientspost-MI 18F-fluoride localized to the culprit coronaryplaque in 37 of 40 subjects, demonstrating that atleast retrospectively 18F-fluoride can identify the in-dividual lesions responsible for adverse coronaryevents (65) (Figure 3). The ability of this technique toprospectively identify high-risk patients at risk of MIis currently being tested in a large prospectivemulticenter study (PREFFIR [Prediction of RecurrentEvents With 18F-Fluoride]; NCT02278211).

What role then might PET/CMR play in the rapidlydeveloping field of coronary plaque imaging?Perhaps the 2 major obstacles to the widespreadadoption of these PET techniques are that of cardiacmotion and radiation exposure. PET/MRI offers po-tential solutions to both. In particular, the novelself-navigation CMR techniques described earlier canbe used to correct both MRI and PET images withoutdiscarding any data, and thereby improving the

efficiency of scanning (69). Moreover, PET/MRIpromises to reduce radiation doses by as much as75% compared with PET/CT. This would allow car-diac PET to be performed using 3 to 4 mSv, makingthis a potentially acceptable clinical tool and facili-tating serial imaging to investigate how disease ac-tivity changes with time or with novel therapeuticinterventions.

FUTURE APPLICATIONS:

IDENTIFYING THE VULNERABLE PATIENT

CAD is a complex, multifaceted, and continuouslyevolving disease process, making the identification ofpatients at high risk of future MI challenging. Accuraterisk prediction is likely to require a similarlymultifaceted approach, playing to one of the keystrengths of CMR. Indeed, in a single scan, CMR canpotentially assess luminal stenosis, plaque burden,high-risk plaque characteristics, and disease activity(Central Illustration). Data to date suggest that high-risk plaques are fairly ubiquitous in the coronaryvasculature and that the majority will not go on tocause clinical events (70). First, many of these plaqueswill heal rather than rupture and, second, even whenrupture does occur then it is usually subclinical. Thishas raised doubt as to whether the positive predictivevalue of identifying individual high-risk plaques willever be great enough to warrant intervention (71). Apatient-level approach may therefore be more effec-tive, perhaps combining markers of plaque burden,ischemia, high-risk plaque characteristics, and diseaseactivity to identify the vulnerable patient (72). Suchpatients might then benefit from systemic rather thanplaque directed intervention. Indeed, with the adventof novel, expensive, and potentially toxic treatmentssuch as Ticagrelor and PCS-K9 inhibitors, an imagingstrategy that can target these medications to patientsat the highest risk would be of real clinical utility. CMRwould appear well positioned to assume such a role inthe future if the technical challenges associated withimaging the coronary arteries can be overcome.

CONCLUSIONS

CMR has the potential to provide a multiparametricassessment of coronary atherosclerosis incorporatingkey information related to plaque burden, high-riskplaque characteristics, and disease activity alongsidemore conventional assessments of luminal stenosis.Considerable technical challenges remain in reliablytranslating these techniques in to the coronary ar-teries; however, if these can be met, then CMR holdsreal promise in improving our understanding of CAD,identifying patients at increased cardiovascular risk,

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6 Dweck et al.M A R C H 2 0 1 6 : 3 0 6 – 1 6 Coronary Imaging With CMR

315

and in delivering the promise of truly personalizedhealth care for CAD.

ACKNOWLEDGMENT The authors thank Dr. PhilipRobson at the Icahn School of Medicine at MountSinai for help in putting the images together.

REPRINT REQUESTS AND CORRESPONDENCE: Dr.Eike Nagel, Institute for Experimental and Trans-lational Cardiovascular Imaging, University HospitalFrankfurt/Main, Frankfurt, Germany. E-mail: eike.nagel@cardiac-imaging.org.

RE F E RENCE S

1. Nabel EG, Braunwald E. A tale of coronary arterydisease and myocardial infarction. N Engl J Med2012;366:54–63.

2. Jahnke C, Paetsch I, Achenbach S, et al.Coronary MR imaging: breath-hold capabilityand patterns, coronary artery rest periods,and beta-blocker use. Radiology 2006;239:71–8.

3. Piccini D, Monney P, Sierro C, et al. Respiratoryself-navigated postcontrast whole-heart coronaryMR angiography: initial experience in patients.Radiology 2014;270:378–86.

4. Edelman RR, Manning WJ, Burstein D, Paulin S.Coronary arteries: breath-hold MR angiography.Radiology 1991;181:641–3.

5. Manning WJ, Li W, Boyle NG, Edelman RR. Fat-suppressed breath-hold magnetic resonance cor-onary angiography. Circulation 1993;87:94–104.

6. Huber ME, Paetsch I, Schnackenburg B, et al.Performance of a new gadolinium-based intra-vascular contrast agent in free-breathing inver-sion-recovery 3D coronary MRA. Magn Reson Med2003;49:115–21.

7. Kelle S, Thouet T, Tangcharoen T, et al. Whole-heart coronary magnetic resonance angiographywith MS-325 (Gadofosveset). Med Sci Monit 2007;13:CR469–74.

8. Kim WY, Danias PG, Stuber M, et al. Coronarymagnetic resonance angiography for the detectionof coronary stenoses. N Engl J Med 2001;345:1863–9.

9. Kato S, Kitagawa K, Ishida N, et al. Assessmentof coronary artery disease using magnetic reso-nance coronary angiography: a national multi-center trial. Journal of the American College ofCardiology 2010;56:983–91.

10. Jahnke C, Paetsch I, Schnackenburg B, et al.Coronary MR angiography with steady-state freeprecession: individually adapted breath-holdtechnique versus free-breathing technique. Radi-ology 2004;232:669–76.

11. Jahnke C, Paetsch I, Nehrke K, et al. Rapid andcomplete coronary arterial tree visualization withmagnetic resonance imaging: feasibility and diag-nostic performance. Eur Heart J 2005;26:2313–9.

12. Sakuma H, Ichikawa Y, Suzawa N, et al.Assessment of coronary arteries with total studytime of less than 30 minutes by using whole-heartcoronary MR angiography. Radiology 2005;237:316–21.

13. Sakuma H, Ichikawa Y, Chino S, Hirano T,Makino K, Takeda K. Detection of coronary arterystenosis with whole-heart coronary magneticresonance angiography. J Am Coll Cardiol 2006;48:1946–50.

14. Schuetz GM, Zacharopoulou NM, Schlattmann P,Dewey M. Meta-analysis: noninvasive coronaryangiography using computed tomography versusmagnetic resonance imaging. Ann Intern Med 2010;152:167–77.

15. Yonezawa M, Nagata M, Kitagawa K, et al.Quantitative analysis of 1.5-T whole-heart coro-nary MR angiograms obtained with 32-channelcardiac coils: a comparison with conventionalquantitative coronary angiography. Radiology2014;271:356–64.

16. American College of Cardiology FoundationTask Force on Expert Consensus Documents,Hundley WG, Bluemke DA, et al. ACCF/ACR/AHA/NASCI/SCMR 2010 expert consensus document oncardiovascular magnetic resonance: a report of theAmerican College of Cardiology Foundation TaskForce on Expert Consensus Documents. J Am CollCardiol 2010;55:2614–62.

17. Andia ME, Henningsson M, Hussain T, et al.Flow-independent 3D whole-heart vessel wallimaging using an interleaved T2-preparationacquisition. Magn Reson Med 2013;69:150–7.

18. Watanabe Y, Nagayama M, Suga T, et al.Characterization of atherosclerotic plaque of ca-rotid arteries with histopathological correlation:vascular wall MR imaging vs. color Doppler ultra-sonography (US). J Magn Reson Imaging 2008;28:478–85.

19. Underhill HR, Kerwin WS, Hatsukami TS,Yuan C. Automated measurement of mean wallthickness in the common carotid artery by MRI: acomparison to intima-media thickness by B-modeultrasound. J Magn Reson Imaging 2006;24:379–87.

20. Corti R, Fuster V, Fayad ZA, et al. Effects ofaggressive versus conventional lipid-loweringtherapy by simvastatin on human atheroscleroticlesions: a prospective, randomized, double-blindtrial with high-resolution magnetic resonance im-aging. J Am Coll Cardiol 2005;46:106–12.

21. Corti R, Fuster V, Fayad ZA, et al. Lipidlowering by simvastatin induces regression ofhuman atherosclerotic lesions: two years’ follow-up by high-resolution noninvasive magnetic reso-nance imaging. Circulation 2002;106:2884–7.

22. Mani V, Muntner P, Gidding SS, et al. Cardio-vascular magnetic resonance parameters ofatherosclerotic plaque burden improve discrimi-nation of prior major adverse cardiovascularevents. J Cardiovasc Magn Reson 2009;11:10.

23. Fayad ZA, Fuster V, Fallon JT, et al. Noninva-sive in vivo human coronary artery lumen and wallimaging using black-blood magnetic resonanceimaging. Circulation 2000;102:506–10.

24. Macedo R, Chen S, Lai S, et al. MRI detectsincreased coronary wall thickness in asymptomaticindividuals: the Multi-Ethnic Study of Atheroscle-rosis (MESA). J Magn Reson Imaging 2008;28:1108–15.

25. Hazirolan T, Gupta SN, Mohamed MA,Bluemke DA. Reproducibility of black-blood cor-onary vessel wall MR imaging. J Magn Reson Im-aging 2005;7:409–13.

26. Desai MY, Lai S, Barmet C, Weiss RG, Stuber M.Reproducibility of 3D free-breathing magneticresonance coronary vessel wall imaging. Eur HeartJ 2005;26:2320–4.

27. He Y, Zhang Z, Dai Q, et al. Accuracy of MRI toidentify the coronary artery plaque: a comparativestudy with intravascular ultrasound. J Magn ResonImaging 2012;35:72–8.

28. Gerretsen S, Kessels AG, Nelemans PJ, et al.Detection of coronary plaques using MR coronaryvessel wall imaging: validation of findings withintravascular ultrasound. Eur Radiol 2013;23:115–24.

29. Kim WY, Stuber M, Börnert P, Kissinger KV,Manning WJ, Botnar RM. Three-dimensionalblack-blood cardiac magnetic resonance coronaryvessel wall imaging detects positive arterialremodeling in patients with nonsignificant coro-nary artery disease. Circulation 2002;106:296–9.

30. Motoyama S, Ito H, Sarai M, et al. Plaquecharacterization by coronary computed tomogra-phy angiography and the likelihood of acute cor-onary events in mid-term follow-up. J Am CollCardiol 2015;66:337–46.

31. Miao C, Chen S, Macedo R, et al. Positiveremodeling of the coronary arteries detected bymagnetic resonance imaging in an asymptomaticpopulation. J Am Coll Cardiol 2009;53:1708–15.

32. Terashima M, Nguyen PK, Rubin GD, et al.Right coronary wall CMR in the older asymptom-atic advance cohort: positive remodeling andassociations with type 2 diabetes and coronarycalcium. J Magn Reson Imaging 2010;12:75.

33. Moody AR, Allder S, Lennox G, Gladman J,Fentem P. Direct magnetic resonance imaging ofcarotid artery thrombus in acute stroke. Lancet1999;353:122–3.

34. Murphy RE. Prevalence of complicated carotidatheroma as detected by magnetic resonancedirect thrombus imaging in patients with sus-pected carotid artery stenosis and previous acutecerebral ischemia. Circulation 2003;107:3053–8.

35. Moody AR, Murphy RE, Morgan PS, et al.Characterization of complicated carotid plaquewith magnetic resonance direct thrombus imaging

Dweck et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 9 , N O . 3 , 2 0 1 6

Coronary Imaging With CMR M A R C H 2 0 1 6 : 3 0 6 – 1 6

316

in patients with cerebral ischemia. Circulation2003;107:3047–52.

36. Jansen CHP, Perera D, Makowski MR, et al.Detection of intracoronary thrombus by magneticresonance imaging in patients with acutemyocardial infarction. Circulation 2011;124:416–24.

37. Ehara S, Hasegawa T, Nakata S, et al. Hyper-intense plaque identified by magnetic resonanceimaging relates to intracoronary thrombus asdetected by optical coherence tomography in pa-tients with angina pectoris. Eur Heart J CardiovascImaging 2012;13:394–9.

38. Kawasaki T, Koga S, Koga N, et al. Character-ization of hyperintense plaque with noncontrastt1-weighted cardiac magnetic resonance coronaryplaque imaging. J Am Coll Cardiol Img 2009;2:720–8.

39. Noguchi T, Yamada N, Higashi M, Goto Y,Naito H. High-intensity signals in carotid plaqueson T1-weighted magnetic resonance imaging pre-dict coronary events in patients with coronary ar-tery disease. J Am Coll Cardiol 2011;58:416–22.

40. Saam T, Hetterich H, Hoffmann V, et al. Meta-analysis and systematic review of the predictivevalue of carotid plaque hemorrhage on cerebro-vascular events by magnetic resonance imaging.J Am Coll Cardiol 2013;62:1081–91.

41. Tanaka A, Kawasaki T, Noguchi T, et al.Hyperintense plaque with noncontrast T1-weighted magnetic resonance coronary plaqueimaging leading to acute coronary syndrome. Cir-culation 2009;120:2400–1.

42. Noguchi T, Kawasaki T, Tanaka A, et al. High-intensity signals in coronary plaques on non-contrast T1-weighted magnetic resonance imagingas a novel determinant of coronary events. J AmColl Cardiol 2014;63:989–99.

43. Asaumi Y, Noguchi T, Morita Y, et al. High-intensity plaques on noncontrast T1-weightedimaging as a predictor of periprocedural myocar-dial injury. J Am Coll Cardiol Img 2015;8:741–3.

44. Kerwin WS, O’Brien KD, Ferguson MS,Polissar N, Hatsukami TS, Yuan C. Inflammation incarotid atherosclerotic plaque: a dynamiccontrast-enhanced MR imaging study. Radiology2006;241:459–68.

45. Yeon SB, Sabir A, Clouse M, et al. Delayed-enhancement cardiovascular magnetic resonancecoronary artery wall imaging. J Am Coll Cardiol2007;50:441–7.

46. Puntmann VO, D’Cruz D, Taylor PC, et al.Contrast enhancement imaging in coronaryarteries in SLE. J Am Coll Cardiol Img 2012;5:962–4.

47. Pedersen SF, Thrysøe SA, Paaske WP, et al.Determination of edema in porcine coronary ar-teries by T2 weighted cardiovascular magneticresonance. J Magn Reson Imaging 2011;13:52.

48. Kim WY, Christiansen EH, Thrysøe SA, et al.First in vivo demonstration of coronary edema in

culprit lesion of patient with acute coronary syn-drome by cardiovascular magnetic resonance. CircCardiovasc Imaging 2011;4:344–6.

49. Holm NR, Kim WY, Maeng M, et al. Coronaryedema demonstrated by cardiovascular magneticresonance in patients with peri-stent inflammationand aneurysm formation after treatment by drug-eluting stents. Circ Cardiovasc Imaging 2013;6:352–4.

50. Cai J, Hatsukami TS, Ferguson MS, et al.In vivo quantitative measurement of intact fibrouscap and lipid-rich necrotic core size in athero-sclerotic carotid plaque: comparison ofhigh-resolution, contrast-enhanced magneticresonance imaging and histology. Circulation2005;112:3437–44.

51. Hatsukami TS, Ross R, Polissar NL, Yuan C.Visualization of fibrous cap thickness and rupturein human atherosclerotic carotid plaque in vivowith high-resolution magnetic resonance imaging.Circulation 2000;102:959–64.

52. Lovett JK, Gallagher PJ, Hands LJ, Walton J,Rothwell PM. Histological correlates of carotidplaque surface morphology on lumen contrastimaging. Circulation 2004;110:2190–7.

53. Trivedi RA, Li Z-Y, U-King-Im J, Graves MJ,Kirkpatrick PJ, Gillard JH. Identifying vulnerablecarotid plaques in vivo using high resolutionmagnetic resonance imaging-based finite elementanalysis. J Neurosurg 2007;107:536–42.

54. Hundley WG, Lange RA, Clarke GD, et al.Assessment of coronary arterial flow and flowreserve in humans with magnetic resonance im-aging. Circulation 1996;93:1502–8.

55. Nagel E, Klein C, Paetsch I, et al. Magneticresonance perfusion measurements for thenoninvasive detection of coronary artery disease.Circulation 2003;108:432–7.

56. Tang T, Howarth SPS, Miller SR, et al.Assessment of inflammatory burden contralateralto the symptomatic carotid stenosis using high-resolution ultrasmall, superparamagnetic ironoxide-enhanced MRI. Stroke 2006;37:2266–70.

57. Tang TY, Howarth SPS, Miller SR, et al. TheATHEROMA (Atorvastatin Therapy: Effects onReduction of Macrophage Activity) Study. Evalu-ation using ultrasmall superparamagnetic ironoxide-enhanced magnetic resonance imaging incarotid disease. J Am Coll Cardiol 2009;53:2039–50.

58. Briley-Saebo KC, Shaw PX, Mulder WJM, et al.Targeted molecular probes for imaging athero-sclerotic lesions with magnetic resonance usingantibodies that recognize oxidation-specific epi-topes. Circulation 2008;117:3206–15.

59. Briley-Saebo KC, Cho YS, Shaw PX, et al. Tar-geted iron oxide particles for in vivo magneticresonance detection of atherosclerotic lesionswith antibodies directed to oxidation-specificepitopes. J Am Coll Cardiol 2011;57:337–47.

60. Briley-Saebo KC, Nguyen TH, Saeboe AM,et al. In vivo detection of oxidation-specific epi-topes in atherosclerotic lesions using biocompat-ible manganese molecular magnetic imagingprobes. J Am Coll Cardiol 2012;59:616–26.

61. Cormode DP, Skajaa T, van Schooneveld MM,et al. Nanocrystal core high-density lipoproteins: amultimodality contrast agent platform. Nano Lett2008;8:3715–23.

62. Duivenvoorden REL, Tang J, Cormode DP,et al. A statin-loaded reconstituted high-density li-poprotein nanoparticle inhibits atherosclerotic pla-que inflammation. Nature Commun 2014;5:3065.

63. Rudd JHF, Narula J, Strauss HW, et al. Imagingatherosclerotic plaque inflammation by fluo-rodeoxyglucose with positron emission tomogra-phy: ready for prime time? J Am Coll Cardiol 2010;55:2527–35.

64. Rogers IS, Nasir K, Figueroa AL, et al. Feasi-bility of FDG imaging of the coronary arteries:comparison between acute coronary syndromeand stable angina. J Am Coll Cardiol Img 2010;3:388–97.

65. Joshi NV, Vesey AT, Williams MC, et al. 18F-fluoride positron emission tomography for identi-fication of ruptured and high-risk coronaryatherosclerotic plaques: a prospective clinical trial.Lancet 2014;383:705–13.

66. Irkle A, Vesey AT, Lewis DY, et al. Identifyingactive vascular microcalcification by 18F-sodiumfluoride positron emission tomography. Nat Com-mun 2015;6:1–11.

67. Adamson PD, Vesey AT, Joshi NV, Newby DE,Dweck MR. Salt in the wound: (18)F-fluoridepositron emission tomography for identification ofvulnerable coronary plaques. Cardiovasc DiagnTher 2015;5:150–5.

68. Dweck MR, Chow MWL, Joshi NV, et al. Cor-onary arterial 18F-sodium fluoride uptake. J AmColl Cardiol 2012;59:1539–48.

69. Coppo S, Piccini D, Bonanno G, et al. Free-running 4D whole-heart self-navigated goldenangle MRI: Initial results. Magn Reson Med 2015;74:1306–16.

70. Stone GW, Maehara A, Lansky AJ, et al.A prospective natural-history study of coronaryatherosclerosis. N Engl J Med 2011;364:226–35.

71. Arbab-Zadeh A, Fuster V. The myth of the“vulnerable plaque”: transitioning from a focus onindividual lesions to atherosclerotic diseaseburden for coronary artery disease risk assess-ment. J Am Coll Cardiol 2015;65:846–55.

72. Tomey MI, Narula J, Kovacic JC. Advances inthe understanding of plaque composition andtreatment options: year in review. J Am Coll Car-diol 2014;63:1604–16.

KEY WORDS coronary artery disease,magnetic resonance imaging, PET/MRI